POSITRON EMISSION TOMOGRAPHY IN QUÉBEC...PET, to the exclusion of its research applica-tions. POSITRON EMISSION TOMOGRAPHY Introduced as a research tool in the mid-70s, PET differs

Report submitted to the

Québec Minister of Research, Science andTechnology

POSITRON EMISSIONTOMOGRAPHY IN QUÉBEC

Prepared by:

François-Pierre DussaultVan H. NguyenFatiha Rachet

(AÉTMIS 01-3 RE)

This assessment is an official report produced and published by the Agence d’évaluation destechnologies et des modes d’intervention en santé (AÉTMIS). It is also available in PDF formaton the Agency’s Web site.

Information concerning this report or any other report produced by AÉTMIS may be obtainedby contacting it at:

Agence d’évaluation des technologies etdes modes d’intervention en santé (AÉTMIS)2021, avenue Union, bureau 1040Montréal (Québec) H3A 2S9

Telephone: (514) 873-2563Fax: (514) 873-1369E-mail: [email protected] address: http://www.aetmis.gouv.qc.ca

How to cite this report:

Agence d’évaluation des technologies et des modes d’intervention en santé (AÉTMIS).Positron emission tomography in Québec. Report prepared by François-Pierre Dussault, VanH. Nguyen and Fatiha Rachet. (AÉTMIS 01-3 RE). Montréal: AÉTMIS , 2001, xviii-270 p.

Legal depositBibliothèque nationale du Québec, 2001National Library of Canada, 2001ISBN 2-550-37972-1

© Gouvernement du Québec, 2001

This report may be reproduced in whole or in part, provided the source is cited.

MISSION

To assist the Minister of Research, Science and Tech-nology and the policymakers in Québec’s health-caresystem, including the Ministère de la Santé et des Serv-ices sociaux, by means of health technology and healthintervention modality assessments, specifically, by as-sessing their efficacy, safety, costs and cost-effectiveness, and their ethical, social and economic im-plications.

To assist the Minister of Research, Science and Tech-nology in developing and implementing scientific policy.

THE AGENCY’S MEMBERS

Renaldo N. Battista, M.D.President and Chief Executive Officer

Jeffrey BarkunPhysician (surgery)

Marie-Dominique BeaulieuPhysician (family medicine)

Suzanne ClaveauPhysician (microbiology/infectious diseases)

Roger JacobBiomedical engineer

Denise LeclercPharmacist

Louise MontreuilAdministrator

Jean-Marie MoutquinPhysician (ob stetr ics and gyn ecolo gy)

Réginald NadeauPhysician (cardiology)

Guy RocherSociologist

Lee SöderstromEconomist

SCIENTIFIC DIRECTOR

Jean-Marie R. Lance

POS ITRON EMISSION TOM OGRAPH Y IN QUÉBEC

Positron emission tomography (PET) is a noninvasivemedical imaging technique. It provides information onthe location and metabolic activity of tissues and lesions,which distinguishes it from most other medical imagingtechniques, which provide mainly anatomical informa-tion.

Originally a research tool, PET is being used more andmore in clinical settings, with the issue of coverage nowfacing most public and private health insurance plans.

With this backdrop, the Fédération des médecins spé-cialistes du Québec and the Conseil québécois de luttecontre le cancer asked the Agence d’évaluation des tech-nologies et des modes d’intervention en santé (AÉTMIS)to examine the appropriateness of deploying PET forclinical purposes in Québec, where there are already twomainly research-oriented centres equipped with thistechnology.

AÉTMIS observed that, while the list of clinical uses isgrowing, the formal assessment reports on the efficacyand cost-effectiveness of PET are cautious about thestrength of the hard data. Nonetheless, a good number ofclinical uses of PET are recognized in oncology, neurol-ogy and cardiology. There are several other, potential ornonrecognized uses in these fields (data incomplete ornonexistent).

The hard data nonetheless seem sufficient for recom-mending the deployment of PET in Québec for certainclinical uses. A ministerial master plan should governthis deployment, taking into account the population’sclinical needs and the specialized human and physicalresources that this technology requires. Furthermore, thisdeployment should be accompanied by research, trainingand validation and take place in close collaboration withuniversity hospital centres and university institutes.

In disseminating this report, AÉTMIS wishes to providethe best possible information to the policymakers con-cerned by this topic at different levels in Québec's healthand social services system.

Renaldo N. BattistaChief Executive Officer

POSITRON EMISSION TOMOGRAPHY IN QUÉBEC

Summaryi

SUMMARY

CONTEXT AND OBJECTIVES OF THIS REPORT

This assessment report was undertaken fol-lowing a joint request from the Fédération desmédecins spécialistes du Québec (FMSQ) andthe Conseil québécois de lutte contre le cancer(CQLC) concerning the clinical efficacy of arecent medical imaging technology, positronemission tomography (PET). Over the pastdecade, the use of PET for diagnostic andtherapeutic guidance and monitoring purposeshas increased considerably. The Agenced’évaluation des technologies et des modesd’intervention en santé (AÉTMIS) undertookto: a) gather hard data on the clinical use ofPET in different fields, in particular, oncology,neurology and cardiology; and b) to make rec-ommendations concerning the possible de-ployment of PET in Québec. While carryingout this project, AÉTMIS relied on the collabo-ration of an advisory committee consisting ofrepresentatives from the FMSQ, the CQLC andthe Ministère de la Santé et des Services so-ciaux (MSSS) and other health technology as-sessment specialists. It should be noted thatthis report examines only the clinical uses ofPET, to the exclusion of its research applica-tions.

POSITRON EMISSION TOMOGRAPHY

Introduced as a research tool in the mid-70s,PET differs from other medical imaging tech-nologies in that it enables one to study themetabolic activity and blood flow in tissues.PET requires the administration of radiophar-maceuticals labelled with positron-emittingisotopes. For example, FDG (fluorodeoxyglu-cose), the substance currently used most oftenin PET scans, incorporates a radioactive iso-tope, fluorine-18. These isotopes are producedby cyclotrons. Three-dimensional images areobtained by detecting photons created during

positron emission and by interpreting themusing a complex imaging system.

Most of the isotopes used in positron emissiontomography have a very short half-life (e.g., 2minutes for oxygen-15 to 110 minutes for fluo-rine-18). Facilities that offer PET services musttherefore have a cyclotron or be close enoughto one to be able to transport isotopes within areasonable amount of time.

In Québec, there are already two mainly re-search-oriented PET facilities with a cyclotron,one at the Montreal Neurological Institute(MNI), the other at the Centre hospitalier uni-versitaire de Sherbrooke (CHUS). Québec cur-rently ranks average among the industrializedcountries in terms of the number of PET facili-ties per capita. Ontario, British Columbia and,as of 2001, Alberta have PET facilities as well.

STUDY METHODOLOGY

As part of this study, AÉTMIS did a review andsynthesis of the hard data on the clinical effi-cacy of PET. The assessment was based ondata from reports published by assessmentagencies and reports from organizations con-taining recommendations concerning PET scancoverage and on publications postdating thosereports. The publications were chosen on thebasis of criteria adapted from reliable proto-cols.

As was the case for several other medical im-aging technologies, the clinical use of PET de-veloped before its efficacy and efficiency weredemonstrated. The fields of application of PETcontinue to evolve, thanks to the contributionof research. Furthermore, the rapid pace oftechnological improvements to PET are an ob-stacle to acquiring hard data to be used for as-


Summaryii

sessment purposes. In fact, the list of useswhose clinical efficacy is recognized is grow-ing by the day. Lastly, the use of PET is devel-oping in a complementary fashion with a pano-ply of medical imaging technologies, which areevolving rapidly as well. This assessment wasnot aimed at documenting the evolution of therelated technologies.

CLINICAL UTILITY OF PET

AÉTMIS’s study confirms the clinical utility ofPET in many oncological, neurological andcardiological applications. For example, in on-cology, the PET’s utility is recognized in cer-tain specific applications in lung cancer, colo-rectal cancer, melanoma, head and neck cancer,and lymphomas. Depending on the type of can-cer, PET is used for diagnostic purposes, fordetecting metastases and for monitoring re-sponse to therapy. In neurology, PET is recog-nized to be effective in certain uses in epilepsyand brain tumors. In cardiology, its utility isrecognized in certain applications, such asmyocardial viability and myocardial perfusionstudies. Lastly, PET has promising potentialfor other uses in these areas of practice.

There are few data on the efficiency of PET,except for some partial data for non-small-celllung cancer, to cite one example. This is whymodels have been constructed exclusively forthe applications in this type of cancer and thosefor evaluating myocardial viability. Thesemodels suggest that PET is cost-effective inthese cases.

SUGGESTIONS FOR DEPLOYING PET IN

QUÉBEC

Since the list of recognized or potential uses ofPET is constantly growing, it is difficult to ac-curately determine the number of patients whocould benefit from this technology in Québec.However, based on opinions expressed withinthe advisory committee, the number of scansrequired could tentatively be estimated at

15,000 or more a year. Based on this estimate,the clinical needs seem sufficient for justifyingthe timely deployment of PET for certain usesin oncology, cardiology and neurology.

These estimated needs can only be met gradu-ally. In ordinary operating conditions, the de-ployment of PET would require about 10 to 15scanners supplied by 3 or 4 cyclotrons (in-cluding those already in operation). Based onthe implementation scenarios chosen, the over-all cost of deploying additional PET resourceswould be between several tens of millions ofdollars and more than about hundred milliondollars.

Gradual deployment is all the more necessarybecause operating a PET centre requires spe-cialized physical and human resources. Pres-ently, there are not enough human resourcesspecifically trained in PET in Québec to sustainthe proposed deployment. Training such re-sources should therefore be a priority.

A ministerial master plan should govern thedeployment of PET, taking into account theclinical needs and specialized physical andhuman resources (existing or to be developed)that this technology requires.

RECOMMENDATIONS

Deployment

• Since the clinical efficacy of PET is recog-nized in many oncological, cardiologicaland neurological applications, it would beadvisable to promote and sustain the de-ployment of PET for clinical purposes inQuébec’s public health-care system.

• PET scans should be available on a prioritybasis for the clinical applications with rec-ognized clinical efficacy. These applica-tions should be reviewed periodically asnew hard data reflecting the rapid devel-


Summaryiii

opment of relevant information becomeavailable.

Deployment Specifics

• A master PET deployment plan should beprepared by the Ministère de la Santé et desServices sociaux.

• The plan should involve quantifying thepopulation’s PET scan needs both with re-gard to optimizing the other existing tech-nologies and to PET requirements forwhich human and physical resources couldsometimes prove to be a limiting factor.The plan should therefore be prepared bothin collaboration with the existing PET cen-tres, whose expertise could be put to good

use, and in consultation with the differentplayers in tertiary intervention settings.

• The plan should take into account the factthat PET cannot be deployed for clinicalpurposes without conducting research intothe promising applications that are pres-ently recognized as having potential butwhose efficacy and cost-effectiveness havenot yet been demonstrated.

• Since the plan should involve assessing theefficiency of PET during its deployment forclinical purposes, the deployment shouldtake place in close collaboration with uni-versity hospital centres and university in-stitutes.


Acknowledgmentsv

ACKNOWLEDGMENTS

This report was prepared at the request of the Agence d’évaluation des technologies et des modesd’intervention en santé (AÉTMIS) by François-Pierre Dussault, Ph.D., molecular biologist and bio-chemist, Van H. Nguyen, pharmacist and health economist, and Dr. Fatiha Rachet, cardiologist, allthree consulting researchers for the Agency. The continuous collaboration of Jean-Marie Lance,health economist and the Agency’s scientific director, helped improve the quality of this report. TheAgency wishes to thank all of these individuals and calls special attention to Dr. Dussault’s coordi-nating role in this project.

The Agency thanks the members of the Positron Emission Tomography Advisory Committee forhaving discussed the topic and read and commented on the different versions of the working docu-ments during the committee’s five meetings. The Agency is especially grateful to the committee’schairman, Dr. Jean-Marie Moutquin, obstetrician and gynecologist and scientific director of theCentre hospitalier universitaire de Sherbrooke’s Clinical Research Centre and a member of AÉTMIS’sboard of directors, for his availability and his active contribution to the realization of this project.

• François Bénard Nuclear Physician and Medical Director PET, Centre hospitalier universitairede Sherbrooke (CHUS), Sherbrooke

• Peter Bogaty Cardiologist, Institut universitaire de cardiologie, Hôpital Laval, Sainte-Foy

• James Brophy Cardiologist, Centre hospitalier de l’Université de Montréal - Notre-DameBranch, Montréal

• Raymond Carrier Biophysicist and Assistant Director, Technological Resources, Centrehospitalier de l’Université de Montréal

• Carolyn R. Freeman Radio-oncologist, McGill University Health Centre, Montréal

• James A. Hanley Statistician, Department of Epidemiology and Biostatistics, McGill University,Montréal

• Yves Lacasse Pneumologist, Hôpital Laval, Sainte-Foy

• Jacques Laplante Hemato-oncologist, Hôpital du Sacré-Cœur, Montréal

• Jacques Lévesque Radiologist, Chief of Medical Imaging, Centre hospitalier universitaire deQuébec, Québec

• Serge Péloquin Medical biologist, Department of Technology Development and Assessment,Division of Medical and University Affairs, Ministère de la Santé et desServices sociaux, Québec City

The Agency especially wishes thank those committee members who made an additional contributionto this report. They are Dr. François Bénard, for the sections on the technical aspects and costs ofPET; James A. Hanley, Ph.D., for the sections dealing with the methodological aspects of assessingdiagnostic tests and part of the update in the field of oncology; Drs. Peter Bogaty and Yves Lacasse,for their part in constructing the economic models; and Serge Péloquin, for having made it easier toobtain Québec statistics on certain imaging devices and detailed information on certain diagnosis-related groups (APR-DRGs) for cost estimate purposes.


Acknowledgmentsvi

The Agency also thanks Renald Lemieux, medical physicist and head of radiation protection in theDepartment of Radiobiology, Centre hospitalier universitaire de Sherbrooke, for writing the appendixon radiation protection, and Daniel Blay and Janet M. Faith, research assistants affiliated withMcGill University, for their contribution to the oncology and neurology updates.

The Agency also expresses its gratitude toward different members of its research and support staff andother, contractual resources who were involved in preparing this report: Pierre Vincent, librarian, andMicheline Paquin, library technician, for their unceasing support in identifying (retrospective searchand bimonthly profiles) and obtaining documentation; and Karina Lapierre, for her unflagging con-tribution during every stage of the preparation of this report, from extracting comparative informationon the positions of different organizations to doing the oncology and neurology updates and theFrench adaptation of these updates and of other sections of this report, collating the different manu-scripts, preparing the progress reports and formatting the final version of the report. Lastly, theAgency wishes to thank Mark Wickens, translator, for this English version of the report.


Table of contentsvii

TABLE OF CONTENTSSUMMARY ........................................................................................................................................................................... I

ACKNOWLEDGMENTS .......................................................................................................................................................V

TABLE OF CONTENTS......................................................................................................................................................VII

LIST OF TABLES AND FIGURES.......................................................................................................................................X

ABBREVIATIONS, INITIALISMS AND ACRONYMS .............................................................................................. XIII

GLOSSARY ....................................................................................................................................................................XVII

1. INTRODUCTION .................................................................................................................................................................11.1 CONTEXT ..........................................................................................................................................................................11.2 THE REQUESTS THAT GAVE RISE TO THIS REPORT ...........................................................................................................11.3 INTERNAL AND EXTERNAL COLLABORATION ..................................................................................................................2

2. POSITION EMISSION TOMOGRAPHY (PET).............................................................................................................52.1 GENERAL DESCRIPTION....................................................................................................................................................52.2 DISTRIBUTION OF PET .....................................................................................................................................................5

3. OBJECTIVES AND METHODOLOGICAL STRATEGIES.........................................................................................73.1 OBJECTIVES......................................................................................................................................................................73.2 IDENTIFICATION, SELECTION AND EVALUATION OF DATA..........................................................................................7

3.2.1 Results of literature searches.....................................................................................................................................73.2.2 Assessment reports ....................................................................................................................................................83.2.3 Modification of the methodological quality scale..................................................................................................113.2.4 Assessment strategy chosen by AÉTMIS ................................................................................................................12

4. CLINICAL USES OF PET ................................................................................................................................................134.1 ONCOLOGY.....................................................................................................................................................................14

4.1.1 Lung cancer..............................................................................................................................................................144.1.2 Colorectal cancer .....................................................................................................................................................184.1.3 Melanoma ................................................................................................................................................................214.1.4 Head and neck cancer..............................................................................................................................................254.1.5 Lymphoma...............................................................................................................................................................284.1.6 Breast cancer............................................................................................................................................................304.1.7 Prostate cancer .........................................................................................................................................................324.1.8 Other cancers ...........................................................................................................................................................334.1.9 Review of the positions published before or in 2000 (oncology)..........................................................................33

4.2 NEUROLOGY...................................................................................................................................................................344.2.1 Dementia and Alzheimer's disease..........................................................................................................................344.2.2 Refractory epilepsy..................................................................................................................................................354.2.3 Brain tumors (mainly glioma).................................................................................................................................384.2.4 Other clinical uses in neurology and psychiatry.....................................................................................................414.2.5 Review of positions published in or before 2000 (neurology) ..............................................................................42

4.3 CARDIOLOGY..................................................................................................................................................................424.3.1 General data .............................................................................................................................................................424.3.2 The role of PET in cardiology.................................................................................................................................434.3.3 Previous positions and update.................................................................................................................................464.3.5 The role of PET among the currently used nuclear medicine tests .......................................................................504.3.6 Conclusion ...............................................................................................................................................................51

4.4 ECONOMIC CONSIDERATIONS ........................................................................................................................................524.4.1 Australian assessment (MSAC) ..............................................................................................................................524.4.2 1999-2000 update - Economic aspects ...................................................................................................................524.4.3 Study specific to Québec: non-small-cell lung cancer (NSCLC)..........................................................................544.4.4 Study of myocardial viability in the Québec context: An exploratory analysis of the potential impact..............60


Table of contentsviii

4.4.5 Studies concerning other uses in oncology and neurology ....................................................................................644.5 GRADING CLINICAL USES ...............................................................................................................................................64

4.5.1 Recognized clinical uses .........................................................................................................................................644.5.2 Potential clinical uses ..............................................................................................................................................654.5.3 Unrecognized clinical uses......................................................................................................................................65

5. DISCUSSION.......................................................................................................................................................................675.1. ASSESSMENT OBJECTIVES AND STRATEGY ...................................................................................................................675.2 ECONOMIC CONSIDERATIONS.........................................................................................................................................685.3 LIMITATIONS OF THE ASSESSMENT ................................................................................................................................69

6. SUGGESTIONS FOR DEPLOYING PET IN QUÉBEC ..............................................................................................716.1 CLINICAL NEEDS.............................................................................................................................................................716.2 PHYSICAL AND HUMAN RESOURCES REQUIRED FOR PET..............................................................................................71

7. CONCLUSIONS AND RECOMMENDATIONS ...........................................................................................................757.1 CONCLUSIONS ................................................................................................................................................................757.2 RECOMMENDATIONS......................................................................................................................................................76

APPENDIX 1: TECHNICAL DATA ON PET....................................................................................................................79A1.1 EQUIPMENT FOR PRODUCING POSITRON-EMITTING RADIOISOTOPES .........................................................................79

A1.1.1 Shielded conventional nuclear medicine camera equipped with a high-energy collimator (511 keV).............81A1.1.2 Collimator-less hybrid nuclear medicine camera for detecting coincidence photons........................................81A1.1.3 Dedicated PET instruments ..................................................................................................................................82

A1.2 SAFETY ........................................................................................................................................................................83A1.3 PRE-PET SCAN PATIENT PREPARATION ......................................................................................................................83

A1.3.1 Oncology studies ..................................................................................................................................................83A1.3.2 Brain studies..........................................................................................................................................................84A1.3.3 Cardiac studies......................................................................................................................................................84

A1.4 ROUTINE CLINICAL PROCEDURES USED DURING PET SCANNING ..............................................................................84A1.4.1 Oncology studies ..................................................................................................................................................84A1.4.2 Brain studies..........................................................................................................................................................84A1.4.3 Cardiac studies......................................................................................................................................................85A1.4.4 Glucose metabolism studies .................................................................................................................................85

APPENDIX 2: RADIATION PROTECTION IN THE POSITRON EMISSION TOMOGRAPHY (PET)LABORATORY............................................................................................................................................89

A2.1 INTRODUCTION............................................................................................................................................................89A2.2 REQUIRED LICENSES....................................................................................................................................................90A2.3 THE PERSONNEL REQUIRED.........................................................................................................................................90A2.4 PACKAGING FOR THE RADIOISOTOPE..........................................................................................................................91A2.5 RADIOISOTOPE DISTRIBUTION DOCUMENTS ...............................................................................................................92A2.6 PROCEDURE FOR PREPARING RADIOACTIVE PACKAGES .............................................................................................92A2.7 THE CARRIER ...............................................................................................................................................................93A2.8 EMERGENCY PROCEDURE............................................................................................................................................93

A2.8.1 General emergency procedure..............................................................................................................................93A2.8.2 Procedure carried out by a specialized carrier.....................................................................................................94

A2.9 LEVELS OF RESPONSIBILITY ........................................................................................................................................94

APPENDIX 3: GEOGRAPHIC DISTRIBUTION OF PET CENTRES..........................................................................97

APPENDIX 4: LOCATION OF IMAGING EQUIPMENT (OTHER THAN PET EQUIPMENT) ...................................103

APPENDIX 5: DESCRIPTION OF ASSESSMENT REPORTS....................................................................................109A5.1 COUNCIL OF MEDICAL IMAGING...............................................................................................................................109A5.2 MINNESOTA HEALTH TECHNOLOGY ADVISORY COMMITTEE.................................................................................109A5.3 MANAGEMENT DECISION AND RESEARCH CENTER, DEPARTMENT OF VETERANS AFFAIRS -TECHNOLOGY

ASSESSMENT PROGRAM (VA-TAP) .........................................................................................................................109


Table of contentsix

A5.4 INTERNATIONAL NETWORK OF AGENCIES FOR HEALTH TECHNOLOGY ASSESSMENT...........................................109A5.5 BLUE CROSS BLUE SHIELD ASSOCIATION TECHNOLOGY EVALUATION CENTER...................................................110A5.6 MEDICARE SERVICES ADVISORY COMMITTEE (AUSTRALIA)..................................................................................110A5.7 HEALTH CARE FINANCING ADMINISTRATION (UNITED STATES).............................................................................110A5.8 COMITÉ D’ÉVALUATION ET DE DIFFUSION DES INNOVATIONS TECHNOLOGIQUES (CÉDIT).................................111A5.9 NATIONAL HEALTH SERVICES - RESEARCH & DEVELOPMENT (UNITED KINGDOM) .............................................111A5.10 ANDALUSIA HEALTH TECHNOLOGY ASSESSMENT AGENCY (AHTAA) ...............................................................112

APPENDIX 6: SEARCH, SELECTION AND EVALUATION OF STUDIES ............................................................115A6.1 SEARCH......................................................................................................................................................................115A6.2 SELECTION.................................................................................................................................................................116A6.3 EVALUATION OF DIAGNOSTIC TESTS.........................................................................................................................116

APPENDIX 7: POSITIONS OF CURRENT AND POTENTIAL USERS, ASSESSMENT AGENCIESAND REIMBURSEMENT ORGANIZATIONS ACCORDING TO THE CLINICAL USEOF PET.........................................................................................................................................................123

A7.1 ONCOLOGY................................................................................................................................................................124A7.1.1 Lung cancer.........................................................................................................................................................124A7.1.2 Colorectal cancer ................................................................................................................................................127A7.1.3 Melanoma ...........................................................................................................................................................129A7.1.4 Head and neck cancer.........................................................................................................................................130A7.1.5 Lymphoma ..........................................................................................................................................................133A7.1.6 Breast cancer.......................................................................................................................................................134A7.1.7 Prostate cancer ....................................................................................................................................................135

A7.2 NEUROLOGY..............................................................................................................................................................137A7.2.1 Dementia and Alzheimer's disease.....................................................................................................................137A7.2.2 Refractory epilepsy.............................................................................................................................................138A7.2.3 Brain tumors (glioma) ........................................................................................................................................140

A7.3 CARDIOLOGY.............................................................................................................................................................142

APPENDIX 8: SELECTED STUDIES ON PET...............................................................................................................149A8.1 LUNG CANCER ...........................................................................................................................................................149A8.2 COLORECTAL CANCER...............................................................................................................................................156A8.3 MELANOMA...............................................................................................................................................................159A8.4 HEAD AND NECK CANCER .........................................................................................................................................163A8.5 LYMPHOMA ...............................................................................................................................................................167A8.6 BREAST CANCER........................................................................................................................................................169A8.7 PROSTATE CANCER....................................................................................................................................................171A8.8 MISCELLANEOUS USES (NOT EXAMINED IN THIS REPORT) .......................................................................................172A8.9 ALZHEIMER’S DISEASE..............................................................................................................................................175A8.10 REFRACTORY EPILEPSY...........................................................................................................................................175A8.11 BRAIN TUMORS (MAINLY GLIOMA).........................................................................................................................178A8.12 MYOCARDIAL VIABILITY ........................................................................................................................................184

APPENDIX 9: METHODOLOGICAL QUALITY OF STUDIES.................................................................................191

APPENDIX 10: PET COST COMPONENTS...................................................................................................................205

APPENDIX 11: ECONOMIC MODELS: ADDITIONAL INFORMATION...............................................................211

APPENDIX 12: LIST OF EXPERTS CONSULTED FOR THE PURPOSE OF CREATING THEDECISION TREES.....................................................................................................................................223

REFERENCES ......................................................................................................................................................................227


List of tables and figuresx

LIST OF TABLES AND FIGURES

TABLES

TABLE 1: CLASSIFICATION OF THE POSITIONS OF ASSESSMENT ORGANIZATIONS AND AGENCIES ...................................14

TABLE 2: RADIOPHARMACEUTICAL TRACERS USED MOST OFTEN FOR PET IN CARDIOLOGY..........................................43

TABLE 3: LITERATURE REVIEW ON THE COST-EFFECTIVENESS OF PET IN LUNG CANCER ...............................................52

TABLE 4: LIST OF THE VARIABLES USED IN THE LUNG CANCER MODEL............................................................................56

TABLE 5: RESULTS FOR EACH STRATEGY PER PATIENT (LUNG CANCER) ..........................................................................58

TABLE 6: LIST OF THE VARIABLES USED IN THE MYOCARDIAL VIABILITY MODEL ...........................................................62

TABLE 7: RESULTS OF THE ECONOMIC ANALYSIS FOR MYOCARDIAL VIABILITY..............................................................62

TABLE 8: CYCLOTRONS PRODUCTS....................................................................................................................................89

TABLE 9: NUCLEAR REACTIONS FOR PRODUCING RADIOISOTOPES...................................................................................89

TABLE 10: RADIOISOTOPE PRODUCTION PARAMETERS .......................................................................................................90

TABLE 11: GEOGRAPHIC DISTRIBUTION OF PET CENTRES..................................................................................................97

TABLE 12: CLINICAL USE OF PET AT CHUS .......................................................................................................................98

TABLE 13: NUMBER OF CLINICAL SCANS AND WAITING TIMES AT CHUS ..........................................................................98

TABLE 14: SOURCE OF PATIENTS AT CHUS.........................................................................................................................98

TABLE 15: EQUIPMENT OTHER THAN PET EQUIPMENT BY REGION (QUÉBEC) ................................................................ 103

TABLE 16: IMAGING EQUIPMENT IN QUÉBEC.................................................................................................................... 104

TABLE 17: STUDY SELECTION CRITERIA (INCLUSION AND EXCLUSION) .......................................................................... 116

TABLE 18: METHODOLOGICAL QUALITY OF DIAGNOSTIC ACCURACY AND DIAGNOSTIC THINKING EFFICACY STUDIES 118

TABLE 19: ASSESSING THE QUALITY OF INDIVIDUAL STUDIES:CLASSIFICATIONS OF STUDY DESIGNS AND LEVELS

OF EVIDENCE ................................................................................................................................................... 119TABLE 20: POSITIONS – LUNG CANCER............................................................................................................................ 124

TABLE 21: POSITIONS – COLORECTAL CANCER ............................................................................................................... 127

TABLE 22: POSITIONS - MELANOMA................................................................................................................................. 129

TABLE 23: POSITIONS - HEAD AND NECK CANCER .......................................................................................................... 131

TABLE 24: POSITIONS - LYMPHOMA ................................................................................................................................. 133

TABLE 25: POSITIONS – BREAST CANCER ........................................................................................................................ 134

TABLE 26: POSITIONS – PROSTATE CANCER .................................................................................................................... 135

TABLE 27: POSITIONS – DEMENTIA AND ALZHEIMER’S DISEASE.................................................................................... 137

TABLE 28: POSITIONS – REFRACTORY EPILEPSY.............................................................................................................. 138

TABLE 29: POSITIONS – NEURO-ONCOLOGY (BRAIN TUMORS, MAINLY GLIOMA)........................................................... 140

TABLE 30: POSITIONS - CARDIOLOGY............................................................................................................................... 142

TABLE 31: SELECTED STUDIES ON PET (ONCOLOGY) ...................................................................................................... 149

TABLE 32: SELECTION OF STUDIES ON PET (NEUROLOGY).............................................................................................. 175


List of tables and figuresxi

TABLE 33: SELECTION OF STUDIES USING A CLINICAL CRITERION AS AN ASSESSMENT CRITERION AND STUDIES ON

CHANGES IN MANAGEMENT DUE TO PET SCAN RESULTS (CARDIOLOGY)......................................................184

TABLE 34: SELECTION OF STUDIES OF THE USE OF PET, USING AN INTERMEDIATE CRITERION AS AN ASSESSMENT

CRITERION (CARDIOLOGY: FUNCTIONAL REVERSIBILTY OF SEGMENTAL WALL MOTION).............................186

TABLE 35: METHODOLOGICAL QUALITY OF STUDIES – LUNG CANCER ..........................................................................191

TABLE 36: METHODOLOGICAL QUALITY OF STUDIES – COLORECTAL CANCER..............................................................194

TABLE 37: METHODOLOGICAL QUALITY OF STUDIES - MELANOMA ...............................................................................194

TABLE 38: METHODOLOGICAL QUALITY OF STUDIES – HEAD AND NECK CANCER........................................................196

TABLE 39: METHODOLOGICAL QUALITY OF STUDIES – LYMPHOMA ...............................................................................197

TABLE 40: METHODOLOGICAL QUALITY OF STUDIES – BREAST CANCER.......................................................................199

TABLE 41: METHODOLOGICAL QUALITY OF STUDIES – PROSTATE CANCER...................................................................199

TABLE 42: METHODOLOGICAL QUALITY OF STUDIES – OTHER APPLICATIONS(NOT EXAMINED IN THIS REPORT) 200TABLE 43: METHODOLOGICAL QUALITY OF STUDIES – NEUROLOGY..............................................................................201

TABLE 44: PURCHASE PRICES (CYCLOTRON + PET)..........................................................................................................205

TABLE 45: PURCHASE PRICES (PET WITHOUT CYCLOTRON) ............................................................................................205

TABLE 46: OPERATING COSTS (CYCLOTRON + PET).........................................................................................................206

TABLE 47: OPERATING COSTS (PET WITHOUT CYCLOTRON)............................................................................................207

TABLE 48: VARIABLES, INTERVALS AND PREDEFINED DISTRIBUTIONS FOR THE MONTE CARLO ANALYSIS (LUNG

CANCER)............................................................................................................................................................214

TABLE 49: UNIVARIATE SENSITIVITY ANALYSIS (LUNG CANCER) ....................................................................................215

TABLE 50: MONTE CARLO ANALYSIS (LUNG CANCER): MEAN, MEDIAN AND QUARTILES ...............................................216

TABLE 51: DISTRIBUTION OF INCREMENTAL COST-EFFECTIVENESS RATIOS FOR THE 1,000 MONTE CARLO

SIMULATIONS (LUNG CANCER) .........................................................................................................................217

TABLE 52: LIST OF VARIABLES AND PREDEFINED DISTRIBUTIONS (MYOCARDIAL VIABILITY) ........................................219

FIGURES

FIGURE 1: CATEGORIES OF RADIOACTIVITY.......................................................................................................................91

FIGURE 2: HAZARDOUS MATERIALS BILL OF LADING (SPECIMEN) ....................................................................................93

FIGURE 3: GROWTH IN THE DEMAND FOR PET AT CHUS (SHERBROOKE) .......................................................................99

FIGURE 4: DECISION TREE (LUNG CANCER)......................................................................................................................211

FIGURE 5: MONTE CARLO ANALYSIS: 1,000 SIMULATIONS (LUNG CANCER)..................................................................216

FIGURE 6: DECISION TREE (MYOCARDIAL VIABILITY) .....................................................................................................218

FIGURE 7: MONTE CARLO ANALYSIS: 1,000 SIMULATIONS, DETECTION OF MYOCARDIAL VIABILITY ..........................219


Abreviations and acronymsxiii

ABBREVIATIONS, INITIALISMS AND ACRONYMS

ACRIN ...................................................American College of Radiology Imaging NetworkAÉTMIS ................................................Agence d’évaluation des technologies et des modes

d’intervention en santé du QuébecAHFMR .................................................Alberta Heritage Foundation for Medical ResearchAHRQ ....................................................Agency for Health Research and Quality (USA)AHSSS ...................................................An Act respecting health services and social services (Qué-

bec)AHTAA..................................................Andalusia Health Technology Assessment AgencyAMSMNQ..............................................Association des médecins spécialistes en médecine nucléaire

du QuébecAP-HP ....................................................Assistance publique - Hôpitaux de ParisAPR-DRG ..............................................All patients revised - diagnostic related groupsARQ .......................................................Association des radiologistes du QuébecASCO.....................................................American Society of Clinical OncologyASR........................................................Age-, sex- and race-adjusted annual mortality rate (general

population)BCBSA...................................................Blue Cross Blue Shield Association (USA)BCBSA-TEC..........................................Blue Cross Blue Shield Association - Technology Evaluation

Centre (USA)BGO .......................................................Bismuth germanium oxideCAHTA..................................................Catalan Agency for Health Technology AssessmentCEA .......................................................Carcinoembryonic antigenCÉDIT....................................................Comité d’Évaluation et de Diffusion des Innovations

Technologiques (France)CHUM....................................................Centre hospitalier de l’Université de MontréalCHUQ ....................................................Centre hospitalier universitaire de QuébecCHUS.....................................................Centre hospitalier universitaire de SherbrookeCI ...........................................................Confidence intervalCMI........................................................Council of Medical Imaging (Ontario)CMS .......................................................Centers for Medicare and Medicaid Services (new name for

the HCFA)CNS........................................................Central nervous systemCNSC .....................................................Canadian Nuclear Safety CommissionCQLC.....................................................Conseil québécois de lutte contre le cancerCRC .......................................................Centre de recherche clinique (Sherbrooke)CT ..........................................................Computerized tomographyDEALE...................................................Declining exponential approximation of life expectancyDHAC-DTB ...........................................Department of Health & Aged Care - Diagnostics & Tech-

nology Branch (Australia)


Abreviations and acronymsxiv

DSR ....................................................... Disease-related survival rateECRI...................................................... Emergency Care and Research Institute (USA)EEG ....................................................... Electroencephalography or electroencephalogramENT ....................................................... Ears, nose and throatEPC........................................................ Evidence-based Practice Centre (USA)EUS ....................................................... Endoscopic ultrasoundFDA....................................................... Food and Drug Administration (USA)FDG....................................................... FluorodeoxyglucoseFMSQ .................................................... Fédération des médecins spécialistes du QuébecGSO....................................................... Gadolinium oxyorthosilicateHCFA .................................................... Health Care Financing Administration (USA) – now called

the Centers for Medicare and Medicaid Services (CMS)ICES ...................................................... Institute of Clinical Evaluative Sciences (Ontario)INAHTA................................................ International Network of Agencies for Health Technology

Assessment (network whose secretariat is in Sweden)ITC ........................................................ Interterritorial Council (of the national health-care system of

Andalusia)keV ........................................................ Kiloelectron voltLSO ....................................................... Lutetium oxyorthosilicateLV.......................................................... Left ventricleLVEF ..................................................... Left ventricular ejection fractionMBq....................................................... MegabecquerelMBS....................................................... Medicare Benefit Schedule (USA)MET....................................................... MethionineMHTAC................................................. Minnesota Health Technology Advisory CommitteeMNI ....................................................... Montreal Neurological InstituteMR......................................................... Magnetic resonanceMRI ....................................................... Magnetic resonance imagingMSAC.................................................... Medicare Services Advisory Committee (Australia)MSSS..................................................... Ministère de la Santé et des Services sociaux (Québec)NaI ......................................................... Sodium iodideNCIC...................................................... National Cancer Institute of CanadaNHL....................................................... Non-Hodgkin’s lymphomaNHS R&D.............................................. National Health Services Research & Development (United

Kingdom)NPV....................................................... Negative predictive valueNSCLC .................................................. Non-small-cell lung cancerPET........................................................ Positron emission tomographyPPV........................................................ Positive predictive valuePSA........................................................ Prostate-specific antigen


Abreviations and acronymsxv

PTC ........................................................Positron Technology Committee (Fédération des médecinsspécialistes du Québec)

PV- .........................................................Negative predictive valuePV+ ........................................................Positive predictive valueRAMQ....................................................Régie de l’assurance-maladie du QuébecROC .......................................................Receiver operating characteristicsROI.........................................................Region of interestR.S.Q......................................................Revised Statutes of QuébecSCLC......................................................Small-cell lung cancerSD ..........................................................Standard deviationSPECT....................................................Single-photon emission computed tomographySPN ........................................................Solitary pulmonary noduleSTARD...................................................Standard for Reporting of Diagnostic AccuracySUV .......................................................Standardized uptake valueT/B .........................................................Tumor-to-background ratioTFNAB...................................................Transthoracic fine-needle aspiration biopsyTRUS .....................................................Transrectal ultrasoundUSP-DI...................................................United States Drug Pharmacopeia - Drug InformationVA-TAP .................................................Veterans Affairs - Technology Assessment Program (USA)


Glossaryxvii

GLOSSARY

Coincidence detection Characteristic of PET technology based on the si-multaneous detection of photons at the oppositepoles of a detector.

Computed tomography A radiologic examination consisting in recon-structing 2- or 3-D images from one-dimensionalprojections from several angles.

Conventional imaging A medical imaging technology that permits onlythe anatomical localizaton of lesions, such as ra-diography, computed axial tomography, ultrasono-graphy, magnetic resonance imaging and scinti-graphy.

Half-life The amount of time it takes a radioactive sub-stance to lose half of its radioactivity, that is, halfof its number of radioactive atoms.

Isotope One of two or more types of atoms of a given ele-ment whose nucleus contains the same number ofprotons but a different number of neutrons.

Magnetic resonance imaging An imaging technique that exploits the orientationproperties of the magnetic moment of hydrogennuclei in the human body under the effect of astrong magnetic field.

Metabolic imaging Visualization of biochemical changes or reactionsin the body.

Morphological imaging Visualization of the shape and structure of organsand tissues.

Positron A subatomic particle whose mass is equal to thatof an electron but which is oppositely charged, i.e.,positive.

Radioisotope An unstable isotope of an element that breaks upor disintegrates spontaneously, emitting radiationand moving to a more stable state in the process.


Glossaryxviii

Radiopharmaceutical A radionuclide-labelled pharmaceutical orchemical preparation used for diagnostic or thera-peutic purposes.

Radiotracer A radionuclide or radiolabelled chemical used toreveal the path travelled by nonradioactive sub-stances or their location.

Scintigraphy A medical imaging technique for showing both thestructure and functioning of organs and of certainpathological processes. The detection of the radia-tion emitted by a radioactive substance introducedinto the body that has a particular affinity for agiven organ or tissue.

Scintillator A substance that emits visible light when hit by asubatomic particle or an x-ray or a gamma ray.

Signal amplification and processing electronics Conversion of an optical image into an electronicimage in order to increase the luminosity and accu-racy of a radiological image.

Single-photon emission tomography (SPECT) A medical imaging technique based ondetecting, by means of a special camera, radiationemitted by a radioactive substance introduced intothe body (scintigraphy) and which provides cross-sectional images (tomography) of different organs.

Standardized uptake value (SUV) Activity of a tracer (counts/pixel/second) in the re-gions of interest (ROIs)/injected dose of tracer +patient’s body mass index (mCi/kg).

Ultrasound (or ultrasonography). Technique for visualizingcertain internal organs by studying the reflection ofa beam of ultrasounds.


Introduction1

1. INTRODUCTION

1.1 CONTEXT

Introduced in the mid-70s, positron emissiontomography (PET) has prompted a great dealof research because of its ability to analyze themetabolic activity and pathophysiology of tis-sues. For the most part, the other medical im-aging technologies (referred to as “conven-tional”) only permit the anatomical localizationof lesions (radiography, computed axial tomo-graphy, ultrasonography, magnetic resonanceimaging, etc.). Over the past decade, the clini-cal use of PET for diagnostic and therapeuticguidance and monitoring purposes has in-creased dramatically, as evidenced in theUnited States by the ever-increasing list of ap-plications covered by public and private healthinsurance plans.

In Canada, current and potential PET usersenthusiastically recommend the deployment ofthis technology for clinical purposes. In Qué-bec, there are presently two PET centres, eachequipped with scanners and a cyclotron, whichis helping to fuel this trend.

Contrary to this enthusiasm, the conclusions ofmost of the assessment organizations that haveexamined this matter and this, still quite re-cently contain a number of provisos on theclinical use of PET. These reservations stemmainly from the paucity of adequate studies ofthe efficacy and cost-effectiveness of PET inrelation to the existing medical imaging tech-nologies.

It was in this context that the Agenced’évaluation des technologies et des modesd’intervention en santé (AÉTMIS) began, inlate September 2000, to examine positronemission technology in Québec. In this report,we first explain the origin of the requests to

have this technology assessed by AÉTMIS andthe steps undertaken to meet these requests.This is followed by a description of the tech-nology, the positions of the current and poten-tial users of PET, those of third-party payersand assessment agencies, and an update onpublications that postdate the latest assess-ments. The current and foreseeable uses arelisted before discussing the avenues for thepossible deployment of this technology forclinical purposes in Québec.

1.2 THE REQUESTS THAT GAVE RISE TO THIS

REPORT

In 1999, the management of the Office of Pro-fessional Development and Health Policy ofthe Fédération des médecins spécialistes duQuébec (FMSQ) set up a committee on posi-tron technology (PTC).

The committee’s mandate was to assess, withinthe framework of the expertise of the special-ties represented on the committee and from areport submitted, in May 2000, to the FMSQby the Association des médecins spécialistes enmédecine nucléaire du Québec (AMSMNQ),the scientific and clinical value of positronemission tomography imaging technology andthe advisability of recommending its use inQuébec, but without specifying, at this point,any sites for implementation.

The specialties represented on the workingcommittee were as follows: cardiology, generalsurgery, hemato-oncology, nuclear medicine,neurosurgery, neurology, respirology, psychia-try, radiology and radio-oncology. A represen-tative from the Ministère de la Santé et des


Introduction2

Services sociaux (MSSS) also sat on this com-mittee, as an observer. In September 2000, sev-eral of the committee’s members submitted, asprivate individuals or as representatives of theirrespective specialties, a report on the potentialuse of PET.

Concurrently with these activities, given theincreasing number of potential uses of PET inoncology, the Conseil québécois de lutte contrele cancer (CQLC) was taking an active interestin this technology.

The FMSQ and the CQLC contacted AÉTMISin September 2000 for the purpose of initiatinga collaborative effort between the three organi-zations with a view to drafting an assessmentreport on PET. On the basis of this assessment,it would be possible to make recommendationsconcerning the deployment of this technology.

During a meeting held on September 26, 2000,which was attended by members of theFMSQ’s Positron Technology Committee andrepresentatives from AÉTMIS, it was agreedthat the process of evaluating the hard data anddrafting a report would involve interactionbetween the authors at AÉTMIS and an advi-sory committee consisting of specialists fromthe FMSQ and representatives from organiza-tions interested in preparing this report,namely, the CQLC and the MSSS.

The advisory committee’s main function was toassist in the drafting of the AÉTMIS report byhelping to incorporate the existing data on theclinical use of PET and the results of theevaluations of these data by various organiza-tions. The members of the advisory committeeare mentioned in the Acknowledgments.AÉTMIS invited one of the members of itsboard of directors, Dr. Jean-Marie Moutquin,to chair the committee.

1.3 INTERNAL AND EXTERNAL COLLABO-

RATION

Because of the wide array of current and po-tential uses of PET and of a tight deadline forpreparing a final report, AÉTMIS sought inter-nal and external collaboration to take advan-tage of specialized expertise and to documentthe various aspects to be covered. The in-houseteam consisted of François-Pierre Dussault,Ph.D., who wrote the entire report, except forthe sections on cardiology, which were pre-pared by Dr. Fatiha Rachet, and the section oneconomics, which was written by Van H.Nguyen. These three authors revised the entirereport at the different stages of its preparation.Other writers contributed as well. They arelisted below.

Specific contributions were requested frommembers of the advisory committee or otherexperts. Thus, James Hanley, Ph.D., collabo-rated in developing the methodological selec-tion-and-assessment strategy; Dr. François Bé-nard wrote the parts on technical data, PET-related activities at CHUS and data interpreta-tion; and Renald Lemieux wrote the section onradiation protection. Two external resourcepersons were asked to select, retrieve andevaluate recent publications in the field of on-cology (Janet Faith and Daniel Blay), whichhad been previously identified, located and ac-quired by AÉTMIS’s documentation staff (Pi-erre Vincent and Micheline Paquet).

Furthermore, a willingness on the part ofAÉTMIS and Ontario’s Institute of ClinicalEvaluative Sciences (ICES), both of whichwere assessing this technology for their re-spective provinces, to cooperate led to ex-changes of information on each other’s meth-odological process, for the purpose of draftingindependent reports. A meeting aimed at com-paring results, obtained or anticipated, tookplace, and a framework for a joint statement on


Introduction3

these two agencies’ positions regarding theclinical use of this technology was developed.ICES had begun its work much earlier thanAÉTMIS and published its report in May 2001.It is available on the Web at http://www.ices.on.ca. Both agencies’ conclusions call forguided deployment of PET for clinical pur-poses.

While this report was being drafted, there wereinformal exchanges with other assessmentagencies, especially the Comité d’évaluation etde diffusion des innovations technologiques(CÉDIT). Once again, although there was nojoint drafting, these exchanges provided an op-portunity to compare methodological ap-proaches and goals, thus preparing thegroundwork for future collaboration.


Positron emission tomography (PET)5

2. POSITION EMISSION TOMOGRAPHY (PET)

2.1 GENERAL DESCRIPTION

PET is a 3-dimensional medical imaging tech-nology. Images are obtained by administeringradiopharmaceuticals containing positron-emitting tracers. Positrons are particles whosemass is equivalent to that of electrons butwhich are oppositely charged. “Positron” is ablend of “positive” and “electron”.

Positrons leave the nuclei of isotopes and pairoff with electrons after travelling a few milli-metres to form short-lived positroniums, whichannihilate, emitting two 511-keV photons ori-ented in opposite directions (180°). The simul-taneous detection of these photons by appropri-ately aligned detectors is used, after computerprocessing, to constitute a 3-dimensional imageof the sites of this energy release.

Positrons were discovered some 60 years ago,but the use of their properties for medical im-aging purposes did not really begin until aboutthe mid-70s [Phelps et al., 1976]. Intensive re-search in various areas of application graduallyevolved to the clinical use of PET in neurologyand cardiology, then in oncology.

The brief overview provided below of the op-erating principles of PET is supplemented inAppendix 1 with additional details on theequipment (cyclotrons and scanners [SectionA1.1]). The appendix also explains the basis ofthe general opinion regarding the safety of PET(Section A1.2). Another section (A1.3) in thisappendix deals with patient preparation, yetanother (A1.4) with the examination procedure.The guiding principles and the regulations gov-erning radiation protection, in particular, thetransport of positron-emitting isotopes, are de-tailed in Appendix 2.

Basically, PET is based on identifying, in thebody, sites where there is an accumulation ofpositron-emitting isotopes. These radioisotopesreveal their presence by emitting photons whenthey enter a stable energy state. One can thuslocate isotopes incorporated into molecules thatare analogs of natural substrates or that havepharmacologic properties and follow their dis-tribution and accumulation in the body.

The use of fluorine-18 (18F) as a substitute foran oxygen atom in a glucose molecule, for ex-ample, permits localization of regions wherethe accumulation of this analog is indicative ofincreased glucose metabolism in the tissues[Phelps, 2000]. In effect, the 18F-labelled ana-log enters cells but is not subsequently me-tabolized. This feature of PET makes this anattractive technology because, depending onthe isotopes used, simple administration by in-halation or intravenous injection enables one tolocate an anatomical entity and to simultane-ously assess its functional status. This distin-guishes PET from a good number of imagingtechniques that provide information mainly onthe anatomical location of the sites of interest,although some of these techniques can providefunctional information.

2.2 DISTRIBUTION OF PET

The increase in the number of PET centresworldwide has occurred rapidly, especiallyover the past few years. In Germany, therewere six PET centres in 1988. In October 1999,there were 65. In the United States, there wereabout 55 centres in 1997, with the figure ex-ceeding 300 in February 2000. Nonetheless, the


Positron emission tomography (PET)6

implementation of PET is not uniform fromcontinent to continent or country to country, ascan be seen from Table 11 in Appendix 3. Thistable does not make a distinction between theuses of PET for clinical purposes and those forresearch purposes. However, PET alreadyseems to have become commonplace at about ahundred centres worldwide and is a researchtechnology used at a hundred or so others, anda large number of potential users would like toacquire it.

It is in oncology that most PET scans are cov-ered, accounting for 65% of all payments.Based on the results of an INAHTA survey,about 85% of clinical PET scans reimbursedworldwide were performed in Australia, Swit-zerland, Denmark and the United States. TheUnited States, Switzerland and Denmark reim-burse the largest number of PET scans per100,000 patients [Adams et al., 1999].

The uses of PET in neurology account for 25%of the clinical PET scans that are reimbursed.The health-care systems in Switzerland, Aus-tralia and Finland are those that reimburse thelargest number of PET scans in neurology.The INAHTA survey revealed that the numberof scans in each area of use varies considera-bly. For instance, scans for brain tumors andepilepsy account for 75% of all PET scans inneurology [Adams et al., 1999].

The uses of PET in cardiology account for thesmallest proportion (6%) of PET scans reim-bursed worldwide, and 60% of them are per-formed to detect viable myocardium. Fewerthan half of the public health-care systems thatparticipated in the INAHTA survey cover the

use of PET in cardiology, the largest proportionof PET cardiac studies being performed in theUnited States, Switzerland and Denmark [Ad-ams et al., 1999].

In Canada, there are PET scanners in sevencities located in four provinces. In Québec,there are two centres equipped with cyclotronsand scanners, both research-oriented and clini-cally oriented in different proportions.

The Montreal Neurological Institute (MNI)has, for about 10 years, been using PET toconduct research on multifocal epilepsy, recur-rent brain tumors, Parkinson’s disease andAlzheimer’s disease. The clinical use of PET atthe institute is limited to about a hundred pa-tients a year.

The PET unit at the Centre hospitalier univer-sitaire de Sherbrooke (CHUS) began operationin 1998 and uses PET for research and clinicalpurposes. Between May 15, 2000 and May 15,2001, 1,253 clinical scans were performed.Scans in oncology patients (lung cancer, lym-phoma, etc.) topped the list, as can be seenfrom Table 12 in Appendix 3. Table 14 showswhere the patients were from, and Table 13 in-dicates that the average waiting time is 45days. Figure 3 in that appendix shows the in-crease in demand over the past year. It wentfrom about 60 a month in 1999 to about 100 in2000 and continues to increase.

It should be noted that the breakdown of thesescans reflects the centre’s activities and cantherefore not be used immediately to extrapo-late to all of Québec’s needs.


Objectives and methodological strategies7

3. OBJECTIVES AND METHODOLOGICAL STRATEGIES

3.1 OBJECTIVES

The objectives of AÉTMIS’s assessment are:

1. To gather hard data on the clinical use ofPET in different fields, in particular, oncol-ogy, neurology and cardiology; and

2. To make recommendations concerning thepossible deployment of PET in Québec.

The methodological strategy for the first ob-jective is the subject of the present chapter. Thesecond objective is discussed in light of theconclusions regarding the clinical efficacy andcost-effectiveness of PET (Chapter 4) and theprerequisites for implementing this technology(Chapter 6).

3.2 IDENTIFICATION, SELECTION AND

EVALUATION OF DATA

The identification of the relevant data involvedseveral steps, most of which were carried outsimultaneously. First, the reports stemmingfrom the work of the Positron TechnologyCommittee (PTC) of the Fédération desmédecins spécialistes du Québec (FMSQ) wereinventoried and consulted (see the section enti-tled “Reports by current or potential users” inthe References). Next, numerous Web siteswere searched, and several reports weredownloaded (see Appendix 5 for a brief de-scription of the latest ten reports on the sub-ject). Lastly, a protocol for querying biblio-graphic databases was instituted in October2000. It included establishing profiles at 2-week intervals and a retrospective search (de-scriptors listed in Appendix 6). The databasesqueried were mainly MEDLINE (PubMed) andthe Cochrane Library. In addition, a fewsearches were done in Embase, Cancerlit and in

health technology economic assessment data-bases.

Given that the assessment reports identified areof recent publication (e.g., that of the MSAC inAustralia [Ghersi et al., March 2000, officialversion disseminated in April 2001:http://www.health.gov.au/haf/pet] and theHCFA’s report [Tunis et al., December 2000]),it was agreed that the retrospective searchwould be limited to forming a bridge betweenthe surveys of the latest available reports andrecent publications postdating those reports.

To ensure overlapping between the end of thesurveys of the recent reports and certain publi-cations predating those reports, given the in-dexing time in certain databases, such asMEDLINE, the cut-off date for the retrospec-tive search performed in November 2000 wasset at January 1999. The surveys of bimonthlyprofiles began in November 2000 and contin-ued until February 2001.

3.2.1 Results of literature searches

Even though the literature search was limitedin time, it generated several thousand individ-ual references. Of these, about 600 were con-sidered for selection, based on the criteria pre-sented below (Section 3.2.3 and Appendix 6).On the whole, the publications identified showthat PET is rapidly advancing from a techno-logical standpoint and that its clinical applica-tions are evolving quickly and in a diversifiedmanner.

Thus, we identified highly technical articles onvarious aspects, such as new crystals for de-



tectors, new attenuation calculations, newmarkers and, more generally, new-generationdevices that increase PET’s clinical perform-ance.

We also found numerous results of studiesaimed at exploring or validating various appli-cations. In oncology, the objective is mainly toenhance the diagnostic capacity of PET in or-der to differentiate between benign and malig-nant lesions, the ability to grade malignant le-sions and the ability to assess the spread of thedisease in order to guide the choice of treat-ment, then to measure its results. In neurology,particularly in refractory epilepsy, PET is usedto determine the sites of surgical intervention,while in cardiology, it is used to assess myo-cardial viability in order to optimize the thera-peutic options (medical, revascularization ortransplantation) and to evaluate coronary per-fusion for the diagnosis, prognosis and thera-peutic monitoring of coronary patients.

Studies exploring or confirming potential clini-cal applications of PET are as increasingly nu-merous in literature searches as comparativestudies are sparse, as are economic studies, es-pecially those on the cost-effectiveness of thistechnology in relation to that of other medicalimaging technologies. In this context, the re-cent assessment reports can shed some light onthe clinical performance of PET by providingorganized data on this aspect, as will be seen inthe following sections.

3.2.2 Assessment reports

PET has been in existence for a few decadesnow, and its clinical use seems to have been anoutgrowth of research. Indeed, structured, ex-haustive reports systemically assessing theclinical performance of PET are still recent andfew in number. For all practical purposes, thefirst widely disseminated report was written in

1996, with an update in 1998, by the VeteransAffairs Technology Assessment Program (VA-TAP) [Flynn and Adams, 1996; Adams andFlynn, 1998].

Between these two reports, the problem ofevaluating diagnostic tests performed by medi-cal imaging was outlined in a report publishedby the Department of Veterans Affairs [Adams,1997] that includes several criteria, which arelisted in Appendix 6. These criteria concern theinclusion or exclusion of studies that wereidentified (Table 17), the evaluation of theirmethodological quality (Table 18) and the as-sessment of the overall quality of the resultsand conclusions (Table 19).

The VA-TAP criteria, which were used to as-sess the clinical performance of PET, weresupplemented with criteria, found in reportsfrom a few other organizations, for evaluatingeconomic studies, which are discussed in Sec-tion 4.4.

Since these publications by the VA-TAP, PEThas been the subject of reports by assessmentagencies and organizations that make recom-mendations for coverage by public and privatehealth-care systems. Some of these reportswere often used as guidelines when drafting thepresent report (e.g., the report of the AndalusiaHealth Technology Assessment Agency[AHTAA]; those of the Medicare Services Ad-visory Committee (MSAC) in Australia; andthose of the Blue Cross Blue Shield Associa-tion Technology Evaluation Center [BCBSA-TEC] and the Health Care Financing Admini-stration [HCFA] in the United States).

Although the raison d’être of the various orga-nizations concerned with the coverage of clini-cal examinations and that of assessment agen-cies are sometimes different, the assessmentmethodologies used for PET were similar inseveral cases. Thus, in the United States, the



HCFA turned to the Agency for Health Re-search and Quality (AHRQ) for evaluating arequest for broad PET coverage submitted by agroup of promoters of this technology, and theBCBSA submitted questions to its own tech-nology evaluation centre (TEC). For its part,the MSAC, which is an assessment organiza-tion, acted on a request from Australia’s Min-ister of Health and Aged Care (MHAC). Inshort, assessment committees drafted or col-laborated closely in the drafting of these orga-nizations’ or agencies’ conclusions.

Furthermore, these PET assessment reports areclosely linked, not only by the criteria used(most of which are those in the VA-TAP re-port, with slight modifications, depending onthe agency or organization), but also by abun-dant mutual citations. For example, the MSACrefers to the INAHTA, VA-TAP and NHSR&D reports (see Appendix 5 for a brief de-scription of these reports), pointing out that itsconclusions are largely consistent with those ofthose reports. In turn, the HCFA based, in part,its assessment on those of the BCBSA-TECand the MSAC, in addition to the results of itsown study review.

From this standpoint, the MSAC and HCFAreports were a major source of information forthis report. That information is described ingreater detail below.

3.2.2.1 Report of Australia’s Medicare Serv-ices Advisory Committee (MSAC)

The MSAC is an independent committee thatwas set up to advise the Commonwealth Min-ister for Health and Aged Care on the con-vincingness of safety, efficacy and cost-effectiveness data on new technologies andnew procedures.

There are three main PET centres in Australia.The first one was created in 1992. Gradually,

and as the other units were put into operation,the number of main clinical uses increased inoncology (lung cancer, melanoma, gastrointes-tinal cancer and a few other uses, namely, braintumors, breast cancer, and head and neck can-cer). In cardiology, the number of uses in-creased until 1994, then decreased.

The issue of PET’s diagnostic efficacy quicklyarose. In 1993, a 5-year assessment project be-gan at two centres. In 1997, the project wasmodified to include partial coverage for certainexaminations in oncology, neurology and car-diology. Afterwards, other centres put into op-eration in 1996 and 1998 made a request to re-ceive such coverage.

The evaluation of the merits of coverage wasrevived. The matter was submitted to theMSAC, and its report on PET was circulatedon a limited basis in March 2000 [Ghersi et al.,2000] for consultation before official approvalof the interim-coverage recommendations. Thelatter were ratified by the Department of Health& Aged Care - Diagnostics & TechnologyBranch (DHAC-DTB) in a report that becameavailable in August 2000 and which was offi-c ia l ly publ ished in Apri l 2001(www.health.gov.au/haf/pet).

As regards methodology, a team of five experi-enced evaluators systematically reviewed thepublications identified between 1966 and Feb-ruary 2000 on the different uses of PET. Theliterature search was carried out using severalbibliographic databases (MEDLINE, the Coch-rane Library and its related databases, assess-ment agency Web sites, health technology eco-nomic evaluation databases, etc.).

The selection criteria were the same as those ofthe VA-TAP and are presented in Table 17 ofAppendix 6. The articles chosen were sub-jected to the following questions [Jaeschke etal., 1994]:



1 . Was there an independent, blindcomparison with a reference stan-dard?

2. Did the patient sample include anappropriate spectrum of patients towhom the diagnostic test will beapplied in clinical practice?

3 . Did the results of the test beingevaluated influence the decision toperform the reference standard?

4. Were the methods for performingthe test described in sufficient de-tail to permit replication?

The articles were then grouped according to thedifferent uses of PET: diagnostic (staging,spread, recurrence), therapeutic impact (choiceof treatment and therapeutic monitoring), pa-tient outcomes and cost-effectiveness.

The application of the said criteria was sup-plemented by opinions of experts in nuclearmedicine, radiology, cancer surgery, neurologyand cardiology.

As an illustration, in the MSAC’s literature re-view of the diagnostic use of PET in non-small-cell lung cancer, 42 papers were selectedfor assessment. A partial breakdown of thesearticles revealed 14 on the diagnostic accuracyof PET, six of which were included in the VA-TAP report. Three papers compared computedtomography alone and computed tomographyfollowed by PET. Five studies concerned thedetection of distant metastases.

It will be noted that the MSAC’s conclusionsconcerning the clinical uses of PET and its rec-ommendations regarding its interim coveragewere ratified by the DHAC-DTB. They aresummarized in Chapter 4 and are spelled out ingreater detail in Appendix 7.

3.2.2.2 Report of the U.S. Health Care Fi-nancing Administration (HCFA)

The HCFA is a federal agency responsible forthe Medicare and Medicaid programs in theUnited States. It also administers the Chil-dren’s Health Insurance program jointly withthe Health Resources and Services Admini-stration. During the last few weeks when thisreport was being drafted, the administration’sname was changed to “Centers for Medicareand Medicaid Services” (CMS). The name“HCFA” was nonetheless kept in this report.

Already in 1995, the HCFA agreed to coverPET scans using rubidium-82 for coronary per-fusion imaging and the management of coro-nary patients. In 1998, it agreed to cover PETscans using FDG (FDG-PET) for staging non-small-cell lung cancer (NSCLC) and evaluatingsolitary pulmonary nodules (SPNs). In 1999,the list was expanded to include the localiza-tion of recurrent colorectal cancer in the con-text of an elevated carcinoembryonic antigen(CEA) level, the staging of lymphoma (in placeof gallium imaging) and the preoperativeevaluation of recurrent melanoma (in place ofgallium imaging).

In July 2000, the HCFA received a request forbroad coverage of FDG-PET scans. The articlein support of this request was published in May2001 [Gambhir et al., 2001]. The request wasaccompanied by 419 articles and abstracts on22 diseases. Because of the large amount ofdata submitted, the HCFA requested assistancefrom the Agency for Health Research andQuality (AHRQ). The AHRQ had the Evi-dence-based Practice Center (EPC) perform avalidation check on the PET submission. TheEPC concluded that the request had not beensubmitted as a standard systematic literaturereview. It was, instead, a vast fresco of PETuses. Questions were at once raised about theinclusion criteria in the cited studies. TheHCFA concluded that systematic reviews sepa-



rate from the compilations accompanying therequest were necessary in order to substantiatea coverage policy in line with its mandates.

The HCFA completed its task using the BlueCross Blue Shield Association’s assessments,those of the Commonwealth Review of Posi-tron Emission Tomography (DHAC-DTB andMSAC), additional analyses of results on lungand esophageal cancer using material from therequester’s submission, and a literature search.The HCFA considered other sources of evi-dence as well, including extensive consulta-tions with clinical experts in oncology, nuclearmedicine, cardiology, neurology and otherrelevant clinical disciplines.

The literature search was performed by theEvidence-based Practice Center (EPC) inMEDLINE and BIOSIS for each of the usesmentioned in the request. For the 10-year pe-riod from 1990 to 2000, the EPC identifiedmore than 500 articles, which were selectedand reviewed according to criteria adaptedfrom those in the 1997 VA-TAP report (Tables17, 18 and 19 in Appendix 6).

In November 2000, the Medicare CoverageAdvisory Committee (MCAC) establishedguidelines for evaluating diagnostic tests andfocused on the following questions:

1 . Is the evidence sufficient for deter-mining if the use of the diagnostic testprovides more-accurate diagnostic in-formation?

Step 1: Assess the quality of thestudies of the test’s performance.

Step 2: Evaluate the possibility thatthe two tests are complementary.

2. If the test improves diagnostic accu-racy, is the evidence adequate to con-clude that the improved accuracywould lead to better health outcomes?

Step 1: Calculate the posttest prob-ability of disease.

Step 2: Evaluate the potential impacton patient management when twotests differ in terms of the posttestprobability of disease.

Over and above the initial request, the follow-ing diseases were chosen for assessment pur-poses: lung cancer, colorectal cancer, lym-phoma, cancer of the esophagus, melanoma,head and neck cancer, coronary artery diseaseand epilepsy. The HCFA concluded that theevidence was sufficient to justify adding cer-tain examinations to the Medicare PET cover-age policy. However, the quality of the evi-dence was not consistent with the standards ofpractice for assessing diagnostic tests. Severalof the studies reviewed had serious methodo-logical biases, which made it difficult to draw afirm conclusion as to the benefits of FDG-PETin clinical practice. The HCFA concludes bystating that, when studies of higher quality areavailable, it would advisable to reconsider thecoverage decision concerning PET scans in or-der to reflect the advancing state of knowledge[HCFA: Tunis et al., 2000].

3.2.3 Modification of the methodologicalquality scale

In general, all assessment reports raise theproblem of assessing diagnostic tests. These is-sues are discussed in greater detail in Appendix6. Not all the criteria proposed by the VA-TAP[Adams, 1997] have been used by the organi-zations and agencies mentioned thus far, be-cause some of the criteria are inapplicable orbecause of the nature of studies aimed atdocumenting the efficacy of PET. These crite-ria are detailed in Tables 17, 18 and 19 in Ap-pendix 6.



The published criteria do not generally seem toexplicitly take into account the simultaneouscomparison of two test modalities in a givenstudy. It is recognized [Ransohoff and Fein-stein, 1978] that patient characteristics cangreatly influence the performance of a givendiagnostic modality. This notion is implicit inthe phrases “broad generalizability to a varietyof patients” and “a narrower spectrum of gen-eralizability” presented in the definitions ofgrades A and B (Table 18, Appendix 6).

For purposes of this report, the comparisoncriterion was considered explicitly, for the bestway to avoid the likelihood of “partial” com-parisons is to insist that the same sample ofcases be evaluated for each of the modalitiesbeing compared. Otherwise, the accuracy ofPET could be assessed with an “easy” sampleat one facility and the accuracy of conventionalimaging assessed in a more “difficult” sample(or vice versa) at another.

Comparative studies were rated A or B whenthey included all or some of the criteria (Table18, Appendix 6). Any study that did not in-clude an explicit comparison between compet-ing technologies or between PET used in con-junction with conventional imaging and con-ventional imaging used alone was rated C or D.However, in some cases, when there is no otheruseful test, the sensitivity or specificity of PETmust be considered in absolute terms.

3.2.4 Assessment strategy chosen byAÉTMIS

AÉTMIS chose, as its point of departure, theexpectations of clinicians as expressed in nar-rative reviews, with no explicit methodologyfor assessing this technology. This group wasdesignated as follows:

� Current or potential users (opin-ions of specialists, based on theirclinical experience and exper-tise).

Second, AÉTMIS used, as a guideline, the con-clusions of assessment reports based on proto-cols that explicitly state criteria for selectingand evaluating studies of the clinical perform-ance of PET. It will be noted that, while theconclusions of these evaluations serve as aspringboard for AÉTMIS’s assessment, the de-cisions of other organizations regarding theconditions for using PET (eligibility of centres,conditions for coverage, etc.) cannot be imme-diately transposed to the situation in Québec.This option led to the grouping of these reportsunder the following headings:

� Assessment agencies (assessments, usingan explicit methodology, of the clinical ef-ficacy of PET).

� Public and private reimbursement organi-zations (assessments of PET, using an ex-plicit methodology, for the purpose ofmaking PET scan coverage decisions).

Third, the VA-TAP’s study selection andevaluation protocols were applied to the publi-cations that were identified from 1999 to Feb-ruary 2001 which had not been included in thepreviously published assessment reports [Ad-ams, 1997], with the addition of the explicit re-quirement of comparing the technologies beingexamined, as described above. The ratingsgiven to the articles selected thus reflect themethodological quality of these studies inAÉTMIS's updates (Tables 35 to 43 in Appen-dix 9) and are based on the criteria in Table 18in Appendix 6. However, these ratings are notbased on the overall study evaluation criteria,since they are usually inapplicable (Table 19,Appendix VI).


Clinical uses of PET13

4. CLINICAL USES OF PET

Which uses of PET are clinically effective?The answer to this question is provided by epi-demiologic data on the diseases of interest (es-pecially for Québec and Canada), on the con-ventional diagnostic modalities and on the roleof PET in each of the uses examined byAÉTMIS in oncology, neurology and cardiol-ogy.

This order of presentation is not indicative ofthe relative importance that AÉTMIS attachesto these areas of application. Furthermore,based on the INAHTA survey [Adams, 1999],the breakdown of the reimbursements for allthe clinical uses was 65% in oncology, 25% inneurology and 6% in cardiology. Oncology ispresented first, as in a number of other assess-ment reports, no doubt to follow their example.Furthermore, data in oncology and neurologyare not always mutually exclusive. As a generalrule, the uses in oncology concern tumors otherthan those of the central nervous system, andpart of the section on neurology concernsneuro-oncology (brain tumors), hence the closeproximity of these two areas of application inthe order of presentation, since there is some-times some overlapping.

The list of uses examined by AÉTMIS is basedon information from the literature surveys andexpectations expressed by current or potentialusers. Thus, for each use examined, there is asummary table of the positions expressed byassessment organizations and agencies. Thesetables provide, in a condensed form, the con-clusions based on structured assessments and abasis for comparison with the results of the up-dates (1999 to February 2001), which confirmor invalidate these conclusions.

Detailed compilations of the positions of cur-rent and potential users, reimbursement organi-zations and assessment agencies are presented

in Appendix 7 for each use examined. Thesepositions are expressed in condensed form forthe purposes of this section and are ranked on ascale of 1 to 6 (see Table 1 on the followingpage).

Next in order are the bases for the MSAC's andHCFA's conclusions. In accordance with theexplanations given in Section 3.2.4, theMSAC's and HCFA's assessments were used asguidelines for recognizing the clinical efficacyof PET in the uses examined here. Failing anassessment by the MSAC or HCFA, the con-clusions of other organizations or agencieswere taken into account, if necessary. In accor-dance with the methodological strategy used byAÉTMIS, updates aimed at identifying publica-tions subsequent to the assessment reports arecompared with the MSAC's and HCFA's con-clusions.

As can be seen from Section 4.4, which con-cerns the economic aspects of PET, there is apaucity of hard data on the cost-effectivenessof the various uses of PET. The cost-effectiveness of the various uses of PET hasnot yet been determined, except for the char-acterization of the solitary pulmonary nodule.This is why these data need to be supplementedby models.

The conclusions of the main organizations oragencies, in particular, those of the MSAC, arenot definitive and involve many nuances re-garding the interpretation of the significance ofthe existing hard data. Consequently, the over-all conclusions (conclusions of agencies andorganizations and conclusions of updates) foreach clinical use were classified as follows:

� A clinical use is said to be recognized ifthe hard data are acceptable (in terms ofclinical efficacy).



� It is considered potential if the hard dataare acceptable but incomplete.

� It is considered nonrecognized if the dataare insufficient to a make a decision, if

they show that it performs poorly or ifthere are no data in the literature surveyscarried out for this report.

.

Table 1: Classification of the positions of assessment organizations and agencies

1 � Recognized use.2 �/� Use recognized by some and considered potential by others.3 � Potential use (but further research is required).4 � Debated (use recognized by some, potential or not recognized by others).5 �/� Use considered potential by some and not supported by the hard data for others.6 � Evidence insufficient to support this use.7 NM Not mentioned.

4.1 ONCOLOGY

In oncology, the following uses were exam-ined: lung cancer, colorectal cancer, melanoma,head and neck cancer, Hodgkin's and non-Hodgkin's lymphoma, breast cancer and pros-tate cancer.

There are probably several important uses ofPET that were not examined in this report. Itshould therefore not be assumed that PET doesnot play a role in the uses that are not men-tioned here. Based on the MSAC's consulta-tions with PET suppliers, the uses examined inits March 2000 report (lung cancer, colorectalcancer, melanoma, glioma, epilepsy and coro-nary revascularization) account for only 40%of the PET scans performed in clinical settings.

4.1.1 Lung cancer

4.1.1.1 General data

In Canada, the age-standardized incidence ratesfor lung cancer have been estimated at 80 per100,000 males and at 47 per 100,000 females

in 2001. For Québec, they are 108 per 100,000males and 51 per 100,000 females, whichtranslates into 6,500 new cases of lung cancerand 5,600 deaths due it (NCIC, Canadian Can-cer Statistics, 2001].

The presence of a radiographically detectedpulmonary lesion (nodule or mass) is reason tosuspect lung cancer. To determine how a pa-tient with lung cancer should be managed andwhat his or her prognosis is, it is essential todifferentiate, histopathologically, betweensmall-cell lung cancer and non-small-cell lungcancer (NSCLC) [Lacasse, 2000].

Small-cell lung cancer (SCLC) accounts forabout 20% of all primary lung cancers. Patientswith this type of cancer have a very poor prog-nosis, with 5-year survival rates of less than1%. Non-small-cell lung cancer (NSCLC) isoften limited to the chest upon initial staging,and 5-year survival rates for these patients are60 to 70% for stage I∗ , 30 to 50% for stage II, 5to 30% for stage III and less than 2% for stage

∗ Staging in accordance with 1997 UICC recommenda-

tions.



IV cancers [Klemenz and Taaleb, 2000].

Despite the intention to treat NSCLC withcurative-intent surgery, the lesions are not re-sectable in 5 to 7% of operated patients, and14% of patients die during the first year fol-lowing curative-intent surgery. These resultsunderscore the major need to improve the strat-egy for diagnosing lung cancer [Klemenz andTaaleb, 2000].

The conventional methods used to characterizethe pulmonary nodule (bronchoscopy, transtho-racic needle biopsy) and to determine the ex-tent of the disease when a diagnosis of canceris made (axial CT scan of the chest, bone scan,abdominal ultrasound and mediastinoscopy)are imperfect and often invasive. Furthermore,it is common for these conventional methods tounderestimate the extent of the disease, whichresults in inadequate patient management [La-casse, 2000].

4.1.1.2 The role of PET

In NSCLC, the following uses were examined:characterizing the solitary pulmonary nodule(SPN), initial staging (detecting distant andmediastinal metastases), monitoring responseto therapy and detecting recurrence or residualtumors following therapy.

On the whole, PET appears to perform betterthan the conventional invasive investigativemethods (e.g., CT-guided transthoracic bi-opsy). Furthermore, it does not involve anycomplications.

4.1.1.3 Previous positions and updates

The list below is based on the positions (con-clusions and decisions) of organizations thatused explicit assessment criteria. In practice, ofthe more recent reports, those of the MSACand HCFA are cited first. Expert opinions areincluded to shed additional light on the subject,if need be. The complete list of positions is

presented in Appendix 7. The studies selectedand evaluated for the updates using the criteriaset out in Chapter 3 and in Appendix 6 are pre-sented in Appendix 8.

� Characterizing the solitary pulmonary nodule(SPN)

� Use recognized by the HCFA, MSACand BCBSA and the assessment agen-cies MHTAC, AHTAA and INAHTA.The VA-TAP recognizes PET’s poten-tial for this use (�).

Following the HCFA’s 1998 decision, Medi-care covers PET scans in patients with an un-determined solitary pulmonary nodule (SPN)(as a substitute for transthoracic biopsies andsurgical excisions) [Tunis et al., 2000].

The MSAC cites three studies concerning thediagnosis of solitary pulmonary lesions [Loweet al., 1998; Prauer et al., 1998; Dewan et al.,1997]. Lowe et al. obtained a sensitivity of92% and a specificity of 90% for PET in suchcases, results which were supported by those ofPrauer et al. (sensitivity of 90% and specificityof 83%). The MSAC concluded that "[t]he po-tential value for PET in this indication is in theavoidance of biopsy in negative lesions. How-ever, since FNAB is still a reasonably low-riskprocedure, PET would mainly be of value forlesions considered to be unsuitable for FNAB[because of severe lung disease or because thelesion is in an unfavourable location] or forthose with a very low post-test probability ofmalignancy" [Ghersi et al., 2000].

Up to February 2001, no new studies on thecharacterization of the SPN were identified.Given that there is no noninvasive examinationfor differentiating between a benign nodule anda malignant one at this time, that the preva-lence of malignant nodules is high in Québecand that most pulmonary nodules are malignant



(personal communication from the advisorycommittee, 2001), PET would be useful in suchcases. The MSAC's and HCFA's conclusionsare repeated here: The characterization of theSPN by PET is considered a recognized use.

� Staging: distant and mediastinal metastases

� Use recognized by the HCFA, BCBSA-TEC, MSAC, MHTAC and INAHTA.According to the VA-TAP, PET haspotential for detecting mediastinal me-tastases.

Following the HCFA’s 1998 decision, Medi-care covers PET scans for the initial staging oflung cancer in patients diagnosed with NSCLC[Tunis et al., 2000].

The MSAC report reviews 14 relevant studiespublished before the end of 1999. It states thatPET is more accurate than conventional imag-ing in detecting distant and mediastinal metas-tases and that the evidence is significant withregard to a change in management before sur-gery or radiation therapy, although there is noevidence at this time to quantify the impact ofthese changes on health outcomes. It concludesthat "[d]espite the methodological limitationsof some of these studies, it is reasonable toconclude that PET significantly improves thediagnostic accuracy of staging the mediastinumover conventional imaging, particularly whenPET is considered in additionto CT" [Ghersi et al., 2000].

Our searches (Appendix 8) identified six newstudies on PET for staging NSCLC. Data onthe sensitivity and specificity of PET are pre-sented in five of them, although only fourcompared it with computed tomography. Thefour studies in question are those by Gupta etal. [2000] (25 patients), Vanuytsel et al. [2000](105 patients), Pieterman et al. [2000] (102 pa-tients) and Hara et al. [2000] (29 patients) (seeAppendix 9 for the methodological quality of

these studies). In these four studies, the sensi-tivity of PET was superior to and its specificitygreater than or equal to those of computed to-mography. In Vanuytsel's study, better sensi-tivity and equal specificity were achieved withPET + CT than CT alone. Only one study, thatby Gupta et al., reports that the sensitivity ofPET depends on the size of the lesions, statingthat it is less (82%) for nodules of < 1 cm.

The improvement achieved with PET in thesefour studies was similar to that reported in thestudies evaluated by the Australian authors:sensitivity is improved by 10% or more, whilethe improvement in specificity is between 0and 25%.

A meta-analysis performed by Dwamena[1999] compares PET for the mediastinalstaging of non-small-cell lung cancer in 14studies and computed tomography in 29 butdoes not provide a matched comparison be-tween the simultaneous, comparative studies.In addition, five of the 14 studies of PET hadalready been reviewed in the Australian report.This meta-analysis concludes that PET is supe-rior to computed tomography.

Another recently published meta-analysis[Gould et al., 2001] does not compare PET toany other imaging method for diagnosing pul-monary nodules and lesions. However, theauthors conclude that FDG-PET is a noninva-sive test with acceptable efficacy, althoughthere are few data for nodules of < 1 cm. Theystate that, in clinical practice, FDG-PET hashigh sensitivity but intermediate specificity indetermining malignancy of a lesion.

With surgical intervention avoided in some pa-tients, PET could also be cost-effective[Gambhir et al.,1998]. Section 4.4 examinesthis in greater detail and provides a decisionmodel for NSCLC.



This update shows that the clinical utility ofPET in staging NSCLC is supported by newdata demonstrating superior sensitivity andequal or superior specificity, which facilitatespatient management in the immediate term.

� Monitoring response to therapy

� The MSAC and VA-TAP recognize thatPET has potential for this use. TheHCFA does not cover PET scans formonitoring response to therapy.

In its report, the MSAC mentions the potentialrole of PET in monitoring response to therapy(chemotherapy or radiation therapy), but thisuse was not thoroughly examined in its March2000 assessment. It does, however, mentiontwo studies [Kiffer et al., 1998; Nestle et al.,1999] that report a change in the radiation fieldin close to one third of cases after a PET scan.However, Kiffer et al.'s study was a retrospec-tive analysis, and that of Nestle et al. involveda series of selected cases, with the result thatthe real impact of PET on the change in man-agement cannot be assessed. The MSAC be-lieves that "[f]or patients whose disease is po-tentially curable but who present a poor surgi-cal risk, radical radiotherapy may be given. Insuch cases, the role and value of PET in theinitial staging and in determining whether ag-gressive or palliative radiotherapy is givenshould be similar to its role in preoperativesurgical imaging."

No new studies on monitoring response totherapy were identified. The MSAC's conclu-sion is repeated here: PET could play a role intherapeutic response monitoring in NSCLC,although hard data are needed in order to con-clude that it is clinically effective.

� Detecting residual or recurrent tumors

� /� The HCFA recommends that this use becovered, and the MHTAC (Minnesota)recognizes this use. The MSAC and

VA-TAP recognize PET’s potential forthis application.

In its 2000 decision, the HCFA made a rec-ommendation to Medicare that it extend itscoverage of PET to the detection of residual orrecurrent NSCLC tumors, the main rationalebeing the study by Bury et al. [1999], whichinvolved 126 patients. Although the PET re-sults were not analyzed in a blinded fashion,the HCFA calls attention to the study's rigorousdesign and the large patient sample. The com-parisons showed that PET was superior tocomputed tomography in detecting residual orrecurrent tumors, with a sensitivity of 100%(compared to 72% for computed tomography)and equivalent specificity (90%) [Tunis et al.,2000].

No new studies on the detection of recurrent orresidual NSCLC tumors were identified. Al-though the HCFA recommends coverage ofPET for detecting recurrent or residual tumors,this recommendation is based on just onestudy. This use of PET is therefore a potentialapplication, according to the MSAC's conclu-sions.

4.1.1.4 Conclusions

In non-small-cell lung cancer, the clinical util-ity of PET is

recognized for the following uses:� characterizing the solitary pulmonary nod-

ule;� initial staging when a diagnosis of non--

small-cell lung cancer is made, including:� the detection of mediastinal lymph

node metastases; and� the detection of distant metastases;

and has potential for:� monitoring response to therapy; and� detecting recurrent or residual tumors.



4.1.2 Colorectal cancer


It is estimated that there will be 17,200 newcases of colorectal cancer and 6,400 deaths dueto colorectal cancer in Canada in 2001. InQuébec, for the same year, the number of newcases of colon and rectal cancer is estimated at4,300 and the number of deaths at 2,100[NCIC, Canadian Cancer Statistics, 2001].

The examination of choice for initially diag-nosing colon cancer and biopsying suspiciouslesions (polyps) is the colonoscopy [Laplante,2000]. Barium enemas, a complete bloodcount, a carcinoembryonic antigen (CEA) as-say and other tests (electrolytes, enzymes, etc.)also aid in the diagnosis. Treatment is essen-tially surgical in nature and consists in excisingthe tumors. This will be the only interventionin cases that are diagnosed early. In more ad-vanced cases, radiation therapy and chemother-apy may be added.

The frequent sites of metastasis during recur-rence are the pelvic region, liver and lungs. Re-currences occur in 75% of patients, and theirdiagnosis is somewhat problematic, especiallyin the case of recurrences in the region of theanastomosis, local recurrence, lymph nodemetastases and distant metastases. The thera-peutic outcomes in these patients mainly de-pend on the recurrence being diagnosed early.The usual follow-up consists of a postoperativeexamination with tumor markers, mainly CEA[Dimitrakopoulou-Strauss, 2000; Laplante,2000].

In the presence of an elevated CEA, the bestmethods for diagnosing recurrent colorectalcancer include the double-contrast barium en-ema, rectosigmoidoscopy, endoscopic ultra-sound (EUS), computed tomography and mag-netic resonance imaging. The efficacy of EUSis superior to that of computed tomography and

magnetic resonance imaging. The latter twotechniques are useful for detecting an increasein lymph node size and detecting distant me-tastases, although their efficacy in detecting alocal recurrence is not entirely satisfactory[Dimitrakopoulou-Strauss, 2000].

Another method for detecting local recurrenceis the CT-guided biopsy, an invasive procedurethat cannot be used as a diagnostic standard be-cause, although a positive result is clearly in-dicative of tumor recurrence, a negative resultcannot rule out the presence of a tumor. Immu-noscintigraphy is another method that can beused to detect recurrence, but its sensitivity de-pends on several factors, including the physicalproperties of the radionuclide, the labellingtechnique, the location and size of the lesions,and the protocol used during the examination[Dimitrakopoulou-Strauss, 2000].


In colorectal cancer, the following uses wereexamined: diagnosing the primary lesion;monitoring response to therapy; evaluating re-gional lymph nodes; detecting hepatic and ex-trahepatic metastases in cases of suspected lo-cal recurrence; detecting recurrence; differenti-ating between recurrence and postoperativescar; and determining the location of recurrenttumors in the context of a rising CEA level.

It is for detecting metastases that PET appearsto be most beneficial in colon cancer, as it en-ables one to plan the simultaneous resection ofthe metastases or to avoid a major operation inthe presence of generalized disease. For de-tecting hepatic or extrahepatic metastases anddifferentiating postoperative scar from local re-currence, PET seems to be more sensitive thencomputed tomography. However, it should benoted that PET can yield false-positive resultsduring the six months following radiation ther-apy because of the inflammatory reaction



[CMI: Beanlands et al., 1999; Laplante, 2000;Létourneau, 2000].


The list below is based on the positions (con-clusions and decisions) of organizations thatused explicit assessment criteria. In practice, ofthe more recent reports, those of the MSACand HCFA are cited first. Expert opinions areincluded to shed further light on the subject, ifneed be. A complete list of the positions is pro-vided in Appendix 7. The studies selected andevaluated for the updates using the criteria setout in Chapter 3 and in Appendix 6 are pre-sented in Appendix 8.

� Diagnosing the primary lesion

�/� Only the MHTAC mentions PET inconnection with diagnosing colorectalcancer. However, the members of theFMSQ's committee consider that there isno evidence to support this use.

No new studies on diagnosing the primary le-sion in colorectal cancer by FDG-PET wereidentified up to February 2001.


� Only the MHTAC and the AHTAAmention PET’s potential for this use.

No new studies on monitoring response totherapy with FDG-PET were identified up toFebruary 2001.

� Detecting hepatic and extrahepatic metastases incases of suspected local recurrence

� /� The HCFA, BCBSA-TEC and MSACsupport this use. The MHTAC and VA-TAP recognize PET’S potential for thisapplication

The conclusions of the Australian report and

those in the HCFA's coverage decision arebased on two high-quality studies with a suffi-ciently large number of patients [Flamen et al.,1999 (103 patients) and Valk et al., 1999 (155patients)]. In addition, similar results had beenobtained in two smaller studies previouslypublished by Ogunbiyi et al. [1997, 47patients] and Delbeke et al. [1997, 45 patients].Valk et al. obtained a sensitivity of 95% withPET for the detection of hepatic metastasesversus 84% with computed tomography (speci-ficity: 100% for PET vs. 95% for computedtomography) and a sensitivity of 90% withPET versus 58% with computed tomographyfor extrahepatic metastases. Flamen et al. ob-tained similar results. The sensitivity of PETwas 4% higher than that of computed tomogra-phy for hepatic metastases and 38% higher forextrahepatic metastases (specificity of PETgreater than or equal to that of CT).

The studies by Flamen et al. and Valk et al.show that PET is superior to computed tomo-graphy in detecting hepatic and extrahepaticmetastases in cases of suspected local recur-rence [Ghersi et al., 2000; Tunis et al., 2000].

For detecting hepatic metastases in cases of lo-cal recurrence of colorectal cancer, the MSACconcludes that "the concordance between CTand PET results is high, though PET offers theadvantage of detecting a small number of can-cers that would have been missed by CT….Little is known about how outcomes are af-fected by changing management in people whohave abnormalities that are detectable only onPET. Although there is evidence that chemo-therapy for early asymptomatic disease pro-longs survival…, it is unclear if this advantageextends to disease detectable only on PET andwhether there are improvements in quality oflife" [Ghersi et al. 2000].

For the detection of extrahepatic metastases,the MSAC concludes that "the relatively largeimprovement in sensitivity of PET in the de-



tection of extrahepatic metastases has a majorpotential impact on avoiding surgery and itshuman and financial costs. The effect ofavoiding surgery on survival and quality of lifehas not been directly assessed" [Ghersi et al.2000].

Three new studies of suitable quality (see Ap-pendix 9) were identified up to February 2001[Imdahl et al., 2000, 71 patients; Zhuang et al.,2000, 80 patients; and Staib et al., 2000, 100patients].

For the detection of metastases, Imdahl et al.report a sensitivity of 100% for PET (vs. 87%for computed tomography and 100% for MRI,although MRI was performed in only 22 pa-tients) and a specificity of 98% (vs. 91% forcomputed tomography and 100% for MRI).Furthermore, the treatment could have beenmodified in 16/71 patients, thanks to PET. Theresults of the study by Zhuang et al. are equallyrobust, with PET having detected extrahepaticlesions in 6/80 patients versus computed tomo-graphy, which did not detect any. Staib et al.'sresults confirm these figures, with a sensitivityof 100% for PET compared to 97% for con-ventional imaging (specificity: 99% for PETvs. 96% for computed tomography). Accord-ingly, surgery in these patients can therefore beavoided with PET.

A new study, by Strasberg et al. [2001], in-volved 43 patients with hepatic metastasesidentified by computed tomography who wereconsidered suitable candidates for surgery. PETidentified extrahepatic metastases in 10 ofthese patients, thus avoiding unnecessary sur-gery.

This update lends, by way of new data, supportto the MSAC's and HCFA's conclusions. Themost important clinical contribution of PET inthis type of cancer consists in identifying pa-tients with extrahepatic metastases, in whomsurgery is contraindicated, which would reduce

the morbidity and hospital costs associatedwith such surgery.

The clinical utility of PET in detecting metas-tases from recurrent colorectal cancer is recog-nized.

� Recurrence: Detection, differentiation ofpostoperative scar and determination of thelocation of recurrent tumors in the contextof a rising CEA level

� The HCFA recognizes the use of PETfor detecting recurrent tumors (in con-junction with computed tomography),but the MSAC, MHTAC, AHTAA, VA-TAP and INAHTA only recognize itspotential in this regard.

� Debate: The HCFA and MSAC recog-nize the use of PET for differentiatingrecurrence from postoperative scar,while the BCBSA-TEC states that thereis no evidence in this regard. TheAHTAA recognizes PET’s potential forthis application.

� Up until 2000, the HCFA covered PETscans only for determining the locationof recurrent colorectal tumors when in-dicated by rising CEA levels (restrictionlifted in 2000). The BCBSA-TEC andAHTAA recognize this use.

In its March 2000 report, the MSAC recom-mends that PET be covered on an interim basisfor the evaluation of residual abnormalities onconventional imaging in patients who aresymptomatic following definitive treatment forcolorectal cancer. This decision was based onthe two studies mentioned earlier (Flamen etal., 1999; Valk et al., 1999], which show thatPET is superior to computed tomography indetecting recurrent colorectal cancer [Ghersi etal., 2000].



Following the HCFA's decision, Medicare has,since 1998, been covering PET scans for de-termining the location of recurrent tumors inthe context of a rising CEA level. In its 2000decision, the HCFA lifted the restriction of anelevated CEA. This use is now covered in thecontext of abnormal clinical findings duringposttreatment follow-up. In addition, theHCFA extended its PET coverage recommen-dation for colorectal cancer to include differ-entiating between local recurrence and postop-erative scar. Although the six studies reviewedby the organization were characterized by cer-tain methodological biases, the HCFA main-tains that PET scanning could have the abilityto influence posttest probabilities such that pa-tients and their physicians can choose an ap-propriate biopsy strategy, which would in-crease the chances for curative resection [Tuniset al., 2000].

No new studies on the utility of PET in differ-entiating between recurrence and postoperativescar or in evaluating residual lesions after de-finitive treatment were identified.

As for determining the location of recurrenttumors in the context of abnormal clinicalfindings after curative-intent treatment, Imdahlet al. [2000] report that, in their study, PETprovided additional information in 86% of thecases compared to conventional imaging,which had an impact on the surgical decision in61% of the cases. Similarly, the results of thestudy by Staib et al. [2000] show a sensitivityof 98% for PET (vs. 91% for computed tomo-graphy and 76% for the CEA assay) and aspecificity of 90% (vs. 72% and 90% for com-puted tomography and the CEA assay, respec-tively).

This update shows that PET seems to improvethe detection of recurrence (in the context ofabnormal clinical findings). No new studies ondifferentiating between residual tumor andpostoperative scar (or radionecrosis) were

identified. The MSAC's and HCFA's conclu-sions are repeated here: The clinical utility ofPET in detecting recurrence, differentiatingbetween recurrence and postoperative scar, anddetermining the location of recurrent tumors inthe context of abnormal clinical findings isrecognized.

4.1.2.4 Conclusions

In colorectal cancer, the clinical utility of PET

is recognized for the following uses:� detecting preoperatively hepatic and extra-

hepatic metastases in patients in whom alocalized recurrence is detected;

� determining the location of recurrent tu-mors in the presence of clinical symptomsor abnormal paraclinical findings (conven-tional imaging, CEA, etc.); and

� differentiating between recurrence andpostoperative scar in the context of diag-nostic imaging that shows abnormalities;

has potential for:� monitoring response to therapy;

and is not recognized for:� diagnosing the primary lesion.

4.1.3 Melanoma


The incidence of melanoma has increasedsteadily over the past few years. Thelen andBares [2000] cite a melanoma incidence ratefor whites in Europe of 8 per 100,000, with therate increasing up to 20 per 100,000 near theequator.

It is estimated that there will be 3,800 newcases of melanoma and 820 melanoma deathsin Canada in 2001, with 570 new cases and 140deaths in Quebec [NCIC, Canadian CancerStatistics, 2001].



One of the characteristics of melanoma is itsrapid metastasis. For staging regional lymphnodes, high-resolution ultrasound is the mostwidely used method. Chest x-rays and abdomi-nal ultrasound are used as well. When the re-sults are inconclusive or reveal an abnormality,CT scans of the chest, abdomen and, if appli-cable, the head are performed [Thelen andBares, 2000].

There is currently no curative treatment formelanoma with widespread metastases.


In melanoma, the following uses were exam-ined: diagnosis, initial staging, evaluatinglymph nodes, detecting extranodal metastasesin the context of a preoperative workup or of aposttreatment or postoperative follow-up, andevaluating recurrence during a postoperativefollow-up.

PET is less effective than conventional diag-nostic methods for detecting the primary lesionand diagnosing local recurrence and thereforecannot, at this time, be used as a diagnosticstandard, according to Thelen and Bares[2000]. It can however, be used to visualizesuspected malignant lesions in structures thatappear normal on clinical examination (e.g.,normal-size lymph nodes), thus permittingearly localization of a tumor and thus improv-ing the patient's prognosis. Furthermore, thedetection of widespread metastases can sparethe patient "extensive, mutilating surgery"[Eigtved et al., 2000].


The list below is based on the positions (con-clusions and decisions) of organizations thatused explicit assessment criteria. In practice, ofthe more recent reports, those of the MSACand HCFA are cited first. Expert opinions areincluded to shed further light on the subject, ifneed be. A complete list of the positions is pro-

vided in Appendix 7. The studies selected andevaluated for the updates using the criteria setout in Chapter 3 and in Appendix 6 are pre-sented in Appendix 8.

� Diagnosing the primary lesion

� The only organization that mentions thisuse of PET is the INAHTA, which rec-ognizes its potential in this regard.

Up to February 2001, no new studies on the di-agnosis of the primary melanoma tumor wereidentified.

� Initial staging

� /� The HCFA, MSAC and BCBSA rec-ommend coverage of PET for the stag-ing of melanoma. The INAHTA recog-nizes its potential in this regard.

The Australian report identified 11 studies onthe detection of melanoma metastases thatqualified according to the selection criteria.Only four of these studies were actually com-parative (one additional study compared thesensitivity rates but not the specificity rates -See Appendix 9). In the first of these studies[Blessing et al., 1995, 20 patients], computedtomography was not found to be more accuratethan ultrasound. In the second study [Valk etal., 1996, 35 patients], the sensitivity of PETand that of computed tomography were compa-rable (96% and 94%), but PET’s specificitywas superior, being 55% versus 12%, which islow. In the third study [Holder et al., 1998, 76patients), the higher sensitivity of PET (94%)versus that of computed tomography (83%)was achieved at the cost of low specificity(55% for PET and < 84% for computed tomo-graphy). The fourth study, the most robust andrigorous one [Rinne et al., 1998] involved 52patients. The sensitivity of PET was 100%, thatof conventional imaging 85%. This sensitivitywas achieved while at the same time main-



taining 96% specificity compared to 68% forconventional imaging.

Based on these studies, mainly that of Rinne etal. [1998], the Australian report concludes that"PET appears to have improved diagnostic ac-curacy over conventional imaging in the detec-tion of metastatic lesions, but methodologicallimitations of the studies should be borne inmind."

Ten studies published between 1999 and Feb-ruary 2001 were identified. Only three of themwere comparative, and they were of low meth-odological quality. The first one [Paquet et al.,2000], which involved 24 patients, reportedequivalent accuracy for PET and computed to-mography (> 80%), which is not surprising,given the small, statistically insufficient patientsample (study rated D).

The second study, rated C, involved 38 patients[Eigtved et al., 2000] and reports 97% sensi-tivity for PET and 62% for conventional im-aging, with 62% specificity (vs. 22% for con-ventional imaging). The third study, rated D,involved 68 patients with advanced melanoma[Dietlein et al., 1999]. It merely concludes that"PET detected fewer pulmonary and hepaticmetastases..., but more lymph node and bonemetastases than conventional radiology or CT."

Since the new studies are of inadequate meth-odological quality, they add nothing to theMSAC's and HCFA's conclusions, which arethat the clinical utility of PET for initial stagingof melanoma is recognized.

� Evaluating lymph nodes

� The HCFA, MSAC and BCBSA-TECconsider that there is no evidence tosupport the use of PET for this applica-tion because of its low sensitivity formicrometastases. The AHTAA ac-knowledges that the evidence is lacking,but that there is nonetheless potential for

this application.

Based on the results of a study by Wagner et al.[1999], the MSAC concludes that PET is notaccurate for diagnosing micrometastases.Wagner et al. reported quite low results for thediagnosis of lymph node metastases comparedto lymph node biopsy. They state that if PET isused to screen patients for the purpose of re-secting seemingly isolated metastases, micro-metastatic disease in other sites cannot be ruledout. The MSAC states that "[t]his does not ne-gate PET's potential role in identifying patientsfor whom surgery would achieve local controland delay the onset of major symptoms orcomplications from metastatic disease. How-ever, the value of PET in changing manage-ment to avoid surgery in those patients withearly metastatic disease is unclear and wouldrequire long-term controlled cohort stud-ies...before any conclusions could be drawn."[Ghersi et al., 2000].

The HCFA concludes that, since there is a lackof evidence to the contrary, certain stronglynegative studies (e.g., that of Wagner et al.,1999, mentioned above) form the basis for itsrefusal to cover PET for lymph node evaluation[Tunis et al., 2000].

A recently published meta-analysis [Mijnhout,2001] of 11 studies, several of which werementioned above (initial staging), draws thefollowing conclusion: PET seems to be highlyaccurate in detecting metastases, but it seems tobe less accurate for regional than systemicstaging and for stages I and II disease thanstage III melanoma.

Two fairly recent studies [Tyler et al., 2000;Wagner et al., 1999] (that of Wagner et al. wasused by the MSAC and HCFA to draw theirconclusions), illustrate, in part, the large dis-parity between the performance results re-ported in various studies. Although Tyler et al.



examined PET in stage III melanoma (in lymphnode metastases), their sensitivity of 87% wasobtained with a specificity of only 44%. Wag-ner et al.'s study [1999], which is of high-quality, found that PET's sensitivity for occultlymph node metastases was only 17%, with96% specificity. Sentinel lymph node biopsy,on the other hand, had a sensitivity of 95%,with 100% specificity.

It seems to be accepted that PET is more sensi-tive in detecting disease with metastatic dis-semination but that it is insensitive in detectingmicrometastases from early-stage lesions. Inshort, the evidence concerning the improve-ment in diagnostic accuracy for lymph nodemetastases with PET is very limited. Wagner etal.'s study shows that PET cannot replace sen-tinel lymph node biopsy for detecting regionallymph node metastases during initial staging orduring posttreatment monitoring.

Since no new studies that could support orcontradict the MSAC’s or HCFA's conclusionswere identified in the update, the clinical utilityof PET in evaluating regional lymph nodes isnot recognized.

� Extranodal metastases (preoperative staging orpostoperative follow-up)

� The HCFA, MSAC and BCBSA-TECrecommend coverage of PET for thedetection of extranodal metastases dur-ing initial staging or in the context of apostoperative follow-up. The AHTAAmentions PET's potential for detectingmetastases in cases of advanced lesions(Clark III and IV).

In its conclusions, the MSAC states that, in thestudies where metastatic disease was not de-tected by conventional imaging modalities,PET was reported to have nearly 100% sensi-tivity [Steinert et al., 1995; Rinne et al., 1998;and Holder et al., 1998]. However, the authors

also mention the possibility of selection bias insome of these studies, since only PET-positivelesions were confirmed histologically. Fur-thermore, they once again cite Wagner et al.'sstudy [1999], which reports a very low sensi-tivity for PET and in which histological con-firmation was done in entire series of cases.Nevertheless, the MSAC recommends cover-age of PET on an interim basis for the preop-erative evaluation of patients being consideredfor surgical resection of apparently limitedmetastatic disease from melanoma [Ghersi etal., 2000].

For its part, the HCFA bases its recommenda-tion that PET be covered for the detection ofextranodal metastases in the context of a post-operative follow-up on two studies cited by theMSAC [Rinne et al., 1998; Holder et al., 1998]and on that of Valk et al. [1996]. Rinne et al.'sstudy [1998] provides promising data to sup-port the use of PET when added to conven-tional imaging in this situation, despite the lackof evidence of health benefits. The HCFAdraws the following conclusion: "However, itis likely that the expected clinical impact willbe limited. In patients where there is a concor-dant result..., there will be no significantchange in management. The true impact willlikely be realized when PET detects lesionsmissed by conventional imaging since patientswill receive necessary treatment in a timelyfashion without the delay from underdiagnosis"[Tunis et al., 2000].

None of the ten new studies identified up toFebruary 2001 is of sufficient quality to con-firm or invalidate the MSAC's or HCFA's con-clusion, which is as follows: The clinical utilityof PET in evaluating extranodal metastases isrecognized.

� Preoperative evaluation of recur-rence/postoperative follow-up

� The HCFA recommends coverage of



PET for the evaluation of recurrencewhen used as an alternative to a galliumscan. The AHTAA also recognizes thisuse. The MSAC makes no mention of it.

Since there are no new studies contradicting its1999 conclusion, the HCFA renewed, in De-cember 2000, its decision concerning PET cov-erage for the evaluation of melanoma prior tosurgery when used as an alternative to a gal-lium scan.

No new studies on the clinical utility of PET inthe evaluation of recurrent melanoma wereidentified up to February 2001.

4.1.3.4 Conclusions

In melanoma, the clinical utility of PET

is recognized for the following uses:� detecting extranodal metastases during

initial staging or in the context of apostoperative follow-up;

� evaluating a potentially treatable recur-rence; and

is not recognized for:� diagnosing the primary lesion; or� detecting lymph node metastases.

4.1.4 Head and neck cancer


Head and neck tumors account for about 3 to5% of all malignant tumors and occur mainlyin males (male-to-female ratio = 6:1). In gen-eral, the worldwide incidence is about 20 in100,000 (males), and the mortality rate, in1994, was 6.3 in 100,000 males and 1.1 in100,000 females [Bender and Straehler-Pohl,2000].

It is estimated that there will be 3,100 new

cases of mouth cancer and 1,250 new cases ofcancer of the larynx in Canada in 2001, with1,570 resulting deaths. In Québec, in 2001, it isestimated that there will be 1,290 and 560 newcases of mouth and laryngeal cancer, respec-tively, with about 450 resulting deaths [NCIC,Canadian Cancer Statistics, 2001].

The diagnosis of head and neck tumors is basedmainly on a physical examination, togetherwith ultrasound, computed tomography, mag-netic resonance imaging and biopsy. In general,palpation and visual examination seem to havebetter specificity than morphological imaging,as they provide information on tumor structureand consistency. The use of imaging tech-niques permits a better evaluation of the size,structure, relationship to the adjacent issuesand extent of the disease. There are limitationswhen the lesions are small and without anymorphologic change (micrometastases, orsmall metastases in normal-size lymph nodes,etc.); when lymph nodes have increased in sizein the absence of any typical signs of malig-nancy; and when the anatomical regions havebeen deformed by surgery or radiation therapy,thus complicating the task of differentiatingbetween postoperative scar and tumor [Benderand Straehler-Pohl, 2000].

Newly diagnosed malignant lesions are consid-ered "unknown primaries" in 2% of cases.These tumors often manifest as adenopathies inlymph nodes throughout the body (37%), and31% of them occur in the head and neck region[Jungehulsing et al., 2000; Scheidhauer et al.,2000]. Squamous-cell cervical metastases fromunknown primary sites probably originate froman unknown head and neck primary, since thetendency with head and neck tumors is thatthey are originally regional, not distant, metas-tases, which makes them potentially curable[Lassen et al., 1999; Dr. Jacques Laplante, per-sonal communication, 2001].




In head and neck cancer, the following useswere examined: identifying an unknown pri-mary tumor, staging regional lymph nodes,monitoring response to therapy, detecting re-currence or residual tumor, and differentiatingpostoperative scar.

It has been suggested that the use of functionalimaging techniques improves tissue evaluationand probably provides clues as to a tumor'shistology. This is very important because typi-cal functional changes can occur at an earlystage, often well before morphological changescan be detected [Bender and Straehler-Pohl,2000].

4.1.4.3 Previous positions and update


� Identifying an unknown primary tumor in thepresence of cervical lymph node metastases

��

The HCFA and BCBSA-TEC recom-mend coverage for PET for this usewhen conventional imaging is unable toidentify the primary tumor. TheAHTAA considers that the data are in-sufficient to support this use, while theVA-TAP recognizes PET’s potential forthis application.

The MSAC does not mention unknown pri-mary tumors in its March 2000 report. It statesthat there are several other potentially impor-tant uses of PET but that they were not studied

in its report and stresses that it should not beassumed that PET has no role to play in theuses that are not mentioned in its assessment.

Of the eight studies identified on the subject,the HCFA selected four for review, based ontheir ability to demonstrate the incrementalbenefit of this use of PET. The four studiesshowed a pooled true-positive rate of 30% inpatients who had negative findings on exami-nation and conventional imaging [Tunis et al.,2000].

PET seems to have positive utility in identify-ing an unknown primary tumor when there arecervical node metastases. When the primarytumor is identified by PET and when this isconfirmed by biopsy, directed tumor manage-ment is initiated, thus avoiding the morbidityassociated with unnecessary radiation therapyor surgery. Long-term survival has thus far notbeen studied. Although management is im-proved, it is not certain that this leads to betterhealth outcomes.

The HCFA concludes that it is reasonable tosupport coverage of PET for identifying un-known primary tumors, since this could bebeneficial in nearly one third of patients inwhom diagnosis might otherwise have failed[Tunis et al., 2000].

Our literature searches identified four newstudies [Jungehulsing et al., 2000; Bo-huslavizki et al., 2000; Perie et al., 2000; andLassen et al., 1999] on the identification of un-known primary tumors, although none of thesestudies is of sufficient quality (see Appendix9). The HCFA's conclusion is therefore re-peated here: The clinical utility of PET inidentifying unknown primary tumors is recog-nized.

� Staging cervical lymph node metastases

� The HCFA and BCBSA recommend



� coverage of PET for the initial stagingof cervical lymph nodes involved inmetastatic disease. In the AHTAA'sopinion, there is no evidence to supportthis use. The VA-TAP recognizes PET’spotential for staging cervical lymphnode metastases.

Based on the BCBSA-TEC's conclusions andon a small, methodologically robust study (n =19) [Wong et al., 1996] that is supported bydata from other, less rigorous studies, theHCFA concludes that PET should be coveredfor staging cervical lymph nodes when there ismetastatic disease. The results of Wong et al.'sstudy showed that computed tomography alonecorrectly staged the disease in 69% of the pa-tients, while computed tomography followedby PET correctly staged it in 92% of the pa-tients [Tunis et al., 2000].

No new studies on staging head and neck can-cer were identified. The HCFA's conclusion isstated here: The clinical utility of PET is rec-ognized for staging cervical lymph nodes.


� In the AHTAA’s opinion, there is noevidence to support this use, while theVA-TAP recognizes PET’s potential forsuch use, although it states that furtherstudies are required. The HCFA doesnot cover PET for monitoring responseto therapy.

No new studies were identified up to February2001. PET has potential clinical utility in thisapplication.

� Detecting recurrence or a residual tumor anddifferentiating postoperative scar

� /� The HCFA recommends coverage ofPET for this use. The MHTAC and VA-TAP recognize PET's potential for thisapplication. The AHTAA recognizes its

potential in cases where there are nega-tive findings on computed tomographyor magnetic resonance imaging.

The HCFA reviewed 11 comparative studies inorder to make its coverage decision regardingPET for detecting recurrence and residual tu-mors and for differentiating them from postop-erative scar. Six of the studies showed PET tohave superior sensitivity and specificity thancomputed tomography, magnetic resonanceimaging and the physical examination [Lowe etal., 2000; Wong et al., 1997; Anzai et al., 1996;Farber et al., 1999; Rege et al., 1994; and Kaoet al., 1998]. Four of them reported neutral ormixed rates [Hanasono et al., 1999; Manolidiset al., 1998; Nowak et al., 1999; and Greven etal., 1997], and the last one [Paulus et al., 1998]reported a less favourable diagnostic perform-ance with PET in relation to computed tomo-graphy. The HCFA bases itself on the study byWong et al. [1996] to support its coverage de-cision, despite the small sample size in thisstudy (which was of relatively strong design),stating that a more informed clinical decisionprobably leads to improved health outcomes[Tunis et al., 2000].

Four new studies were identified up to Febru-ary 2001 [Di Martino et al., 2000; Lowe et al.,2000; Lonneux, 2000; and Farber et al., 1999].They seem to suggest that PET is superior toconventional imaging (computed tomography,MRI, ultrasound), but these studies were of lowmethodological quality (grades C or D; seeAppendix 9).

Since this update did not identify any studiessupporting or contradicting the HCFA's con-clusion concerning head and neck cancer, thatconclusion is stated here: The clinical utility ofPET in detecting recurrent disease and residualtumors in the presence of abnormalities onconventional imaging is recognized.



4.1.4.4 Conclusions

In head and neck cancer, the clinical utility ofPET

is recognized for:

� identifying an unknown primary tumorin the presence of cervical node metas-tases;

� staging cervical lymph nodes whenthere are negative findings on conven-tional imaging; and

� detecting recurrent disease or residualtumor and differentiating postoperativescar;

and is not recognized for:� monitoring response to therapy.

4.1.5 Lymphoma


It is estimated that there will be 6,200 newcases of non-Hodgkin's lymphoma and 2,700deaths due to this type of cancer in Canada in2001. In addition, it is estimated that there willbe 810 new cases of Hodgkin's lymphoma and120 deaths due to this disease. In Québec, theestimates are 1,590 new cases of non-Hodgkin's lymphoma and 580 deaths [NCIC,Canadian Cancer Statistics, 2001].

The introduction of computed tomography andmagnetic resonance imaging led to vastchanges in the diagnostic imaging of lympho-mas. Although there have been some meth-odological improvements, these changes haveapparently now slowed down. Because of thelimitations of diagnostic methods like ultra-sound, staging malignant lymphoma presentlydepends on an increasing number of diagnosticmethods specific to the disease of interest (or-gan system-specific). These tests are associatedwith significant costs in terms of time and lo-

gistics, to say nothing of the considerable pa-tient discomfort associated with the usual pro-cedures for staging lymphoma (bone marrowaspiration and biopsy, CT scans of the abdo-men, pelvis and chest, gallium imaging and, insome cases, bone imaging) [CMI: Beanlands etal., 1999; Moog, 2000].


In lymphoma, the following uses were exam-ined: initial staging, posttreatment follow-upand monitoring response to therapy.

The numerous modalities used for the usualstaging of lymphoma can be replaced by a sin-gle PET scan. Furthermore, PET often revealsa more advanced stage of the disease and leadsto more aggressive management. It has beensuggested that PET is as useful for evaluatingresponse to therapy as it is for staging. AnAmerican study [Young et al., 1997] found thatupon using PET prior to treatment and againafter two cycles of chemotherapy for evaluat-ing the intermediate response to treatment, themortality rate due to lymphoma was reducedby one half in relation to the national averagerate [CMI: Beanlands et al., 1999].

For example, PET is a one-day procedure thatis more advantageous than gallium imaging be-cause it has higher resolution, permits betterdosimetry, induces less intestinal activity andshows quantitative potential. Furthermore,FDG-PET can be used to evaluate residual le-sions posttreatment with a high level of accu-racy [Laplante, 2000].


The list below is based on the positions (con-clusions and decisions) of organizations thatused explicit assessment criteria. In practice, ofthe more recent reports, those of the MSACand HCFA are cited first. Expert opinions areincluded to shed further light on the subject, ifneed be. A complete list of the positions is pro-



vided in Appendix 7. The studies selected andevaluated for the updates using the criteria setout in Chapter 3 and in Appendix 6 are pre-sented in Appendix 8.

� Initial staging and posttreatment follow-up

� /� The HCFA and BCBSA-TEC both rec-ommend coverage of PET for theseuses. The MHTAC and AHTAA recog-nize PET's potential for these applica-tions.

The MSAC does not mention Hodgkin's ornon-Hodgkin's lymphoma in its March 2000report. It states that there are several potentiallyimportant uses of PET that were not examinedin its report and stresses that it should not beassumed that PET has no role to play in theuses that are not mentioned in its assessment.

Since 1999, the HCFA has recommended cov-erage of PET for initial and posttreatmentstaging when used as a substitute for a galliumscan.

The BCBSA-TEC identified 14 studies pub-lished between April 1997 and March 2000 onthe staging of Hodgkin's and non-Hodgkin'slymphoma during initial staging or posttreat-ment follow-up [Bangerter et al., 1999; Jeru-salem et al., 1999; Moog et al., 1999; Zinzaniet al., 1999; Stumpe et al., 1998; Hoh et al.,1997; Bares et al., 1993; Carr et al., 1998; deWit et al., 1997; Rodriguez et al., 1997;Bangerter et al., 1998, Rodriguez et al., 1995;Leskinen-Kallio et al., 1991; and Okada et al.,1994]. Of these studies, three compared PETwith conventional imaging, and two includeddata concerning sites with and without diseasefor PET and computed tomography. These twostudies showed that PET provides greater diag-nostic accuracy than computed tomography.The BCBSA-TEC concludes that when PET isadded to conventional imaging, it provides use-ful information for selecting effective treatment

appropriate to the correct stage of disease.

The update to February 2001 identified onlyone study [Jerusalem et al., 2000], which, al-though of low quality, suggests that PET isclinically useful.

The fact that the available data are evolving ata rapid pace, that is, within a span of a fewmonths, can lead to different conclusions con-cerning certain uses. Referring to the results oftwo studies identified after February 2001 thatare of sufficient quality for our analysis[Buchmann et al., 2001; Spaepen et al., 2001](see Appendix 9), one might note evidence, inthe form of new data, that would support rec-ognition of two uses of PET in lymphoma: 1)when restaging could affect the choice oftreatment; and 2) for evaluating residual dis-ease after treatment.

Buchmann et al.'s study [2001] compared PETwith computed tomography for staging lym-phoma. Discrepant results were verified by bi-opsy, magnetic resonance imaging or a clinicalfollow-up of 4 to 24 months. PET detectedlymph node manifestations of lymphoma with99.2% sensitivity (100% specificity) comparedto computed tomography, which had a sensi-tivity of 83.2% (specificity of 99.8%).

Spaepen et al. [2001] studied the clinical utilityof PET in identifying lymphomas with lymphnode, extranodal, supradiaphragmatic and in-fradiaphragmatic manifestations. PET was su-perior to computed tomography in all cases,and its sensitivity in detecting extranodal mani-festations was 100% versus 80.8% for com-puted tomography (specificity 99.4% for both).These authors also examined the diagnostic ac-curacy of PET in posttreatment follow-up andshowed that PET predicted persistent disease in14/26 cases in which the results were positive.

A study of lesser quality (grade C) [Huelten-schmidt et al., 2001] confirms that PET is more



specific for restaging (95% sensitivity and 89%specificity for PET vs. 95% and 39%, respec-tively, for conventional imaging) and for de-tecting recurrence (91% sensitivity and 71%specificity for PET vs. 91% sensitivity for con-ventional imaging, with the latter’s specificitynot mentioned).

Although it is outside the limits that we set forselecting studies (1999 to February 2001), thisupdate supports the HCFA's and BCBSA-TEC’s conclusions with new data. The clinicalutility of PET in initial staging and posttreat-ment follow-up is therefore recognized.


� The MHTAC is the only organizationthat mentions this use for PET. TheARQ recognizes PET's potential for thisapplication, and the CMI recognizes thisuse. The HCFA does not cover PET formonitoring response to therapy

No new studies on monitoring response totherapy in lymphoma were identified. At pre-sent, this is a potential use of PET.

4.1.5.4 Conclusions

In Hodgkin's and non-Hodgkin's lymphoma,the clinical utility of PET

is recognized for:� initial staging when restaging could affect

the choice of treatment; and� evaluating residual disease after treatment;

and has potential for:� evaluating response to therapy.

4.1.6 Breast cancer


It is estimated that 19,500 Canadian womenwill be diagnosed with breast cancer in 2001,

with 5,000 new cases in Québec. The numberof deaths due to breast cancer in all the prov-inces in 2001 is estimated at 5,500, with 1,450of these deaths occurring in Québec [NCIC,Canadian Cancer Statistics, 2001].

When a clinical examination in conjunctionwith mammography and aspiration cytology(triple approach) does not permit a definitivediagnosis, invasive procedures need to be con-sidered. It is important to note that, in patientswith an abnormal mammogram who have un-dergone surgery, the masses that are operatedon are malignant in only 2 to 4 cases in 10. Inaddition, about 10% of breast cancers are notidentified by mammography, even when theyare palpable. This is why the diagnosis shouldbe made on the basis of the morphological re-sults obtained by aspiration, needle biopsy oreven open biopsy [Avril et al., 2000a].


In breast cancer, the following uses were ex-amined: detecting the primary tumor, stagingprimary and recurrent tumors, and monitoringresponse to therapy.

PET cannot be used as a routine test for de-tecting breast cancer because of its low sensi-tivity or because its sensitivity is comparable tothat of mammography and scintimammogra-phy, which are less expensive procedures.However, it is suggested that PET could be ofsignificant value in clinical cases that are con-sidered difficult, especially in women withlarge breasts or fibrocystic disease, those whohave had a first biopsy, surgery or radiationtherapy, and those with breast implants. PET isalso useful in evaluating an axillary mass sus-pected of being breast cancer and can evenavoid axillary dissection. It is superior to car-tography in detecting osteolytic bone metasta-ses, although it can miss certain osteoblastic le-sions. It can also prove useful in detecting re-current or metastatic tumors and very useful in



monitoring response to therapy, since it can beperformed earlier than the other methods forevaluating treatment [Avril et al., 2000a; Lap-lante, 2000; Létourneau, 2000].



� Detecting the primary tumor

� Although the AHTAA and BCBSA-TEC consider that there is no evidenceto support PET for this use, the VA-TAP recognizes PET's potential for thisapplication. None of the other organiza-tions mention this use, but theAMSMNQ, the CMI and the FMSQ'sPTC have all stressed its importance inclinical cases that are considered diffi-cult.

The MSAC does not mention breast cancer inits March 2000 report. It states that there areseveral other potentially important uses of PETthat were not examined in its report andstresses that it should not be assumed that PEThas no role to play in the uses that are notmentioned in its assessment.

The HCFA did not make a PET coverage deci-sion for breast cancer but instead referred thematter for consideration to the MCAC Diag-nostic Imaging Panel.

The BCBSA-TEC states that the medical lit-erature is incomplete when it comes to sup-

porting the use of PET in breast cancer.

Up to February 2001, no new studies, with suf-ficiently convincing data, of the use of PET inbreast cancer were identified.

PET has potential clinical utility in detectingprimary breast cancer tumors in cases consid-ered difficult.

� Staging primary and recurrent tumors

� Three assessment agencies (AHTAA,VA-TAP and INAHTA) recognizePET's potential for this application, es-pecially for recurrent tumors.

Up to February 2001, no new studies of PETfor the detection of breast cancer metastaseswere identified. This is therefore a potential useof PET.


� The MHTAC and VA-TAP recognizePET's potential for this use.

Up to February 2001, no new studies, with suf-ficiently convincing data, of this use of PET inbreast cancer were identified. This is thereforea potential use of PET.

4.1.6.4 Conclusions

In breast cancer, PET's clinical utility is notclearly recognized, but

it has potential for:� staging primary and recurrent tumors;� detecting axillary and internal mammary

lymph node metastases;� detecting the primary tumor in the context

of an equivocal complete evaluation; and� monitoring response to therapy.



4.1.7 Prostate cancer


It is estimated that there will be 17,800 newcases and 4,300 deaths in Canada in 2001, withan estimated 3,800 new cases and 890 deaths inQuébec. (NCIC, Canadian Cancer Statistics,2001]. It should be noted that autopsy resultshave shown that prostate cancer is present inabout 30% of men aged 50 and in about 60 to70% of men aged 70, but that this type of can-cer presents clinical manifestations in only onethird of cases [Avril et al., 2000b].

Although different schools of thought havequestioned the different approaches, the diag-nosis of prostate cancer is based primarily on adigital rectal examination, transrectal ultra-sound (TRUS), a prostate-specific antigen(PSA) assay and a prostate biopsy. A PSA as-say does not enable one to distinguish betweenorgan-confined and more advanced disease.TRUS permits visualization of the internal ar-chitecture and entire contour of the prostateand is an important tool in staging the disease[Avril et al., 2000b].

Computed tomography and magnetic resonanceimaging (MRI) are also used to evaluate theprostate. However, computed tomography doesnot allow one to distinguish between prostatecancer and benign hyperplasia or normalprostatic tissue. Studies have shown that MRIcannot provide adequate information for visu-alizing the anterior margin of the prostate.Technological developments in magnetic reso-nance imaging have improved its capacity tovisualize the prostate, but 35 to 52% of prostatecancers go undetected by magnetic resonanceimaging [Avril et al., 2000b].


In prostate cancer, the following use was ex-amined: detecting recurrence or residual tu-

mors.

It has been shown that FDG-PET has too low adetection rate to identify prostate cancer, andthere is no evidence that staging regional anddistant metastases and detecting recurrent can-cer can be achieved with sufficient accuracy byPET. Presently, there are no hard data in favourof PET scanning in patients with prostate can-cer [Avril et al., 2000b].

The members of the FMSQ's PTC consider thatFDG has a limited ability to detect primarycarcinoma of prostate and to differentiate be-tween a malignant tumor and benign prostatichyperplasia [Laplante, 2000]. However, the useof PET with other radiotracers, such as [11C]-acetate, seems promising for the detection ofprostate cancer [Avril et al., 2000b]. Furtherstudies are necessary.



� Detecting recurrence or residual tumor

� Only the MHTAC mentions PET's po-tential in prostate cancer.

Only one recent study of the use of PET inprostate cancer was identified [Seltzer et al.,1999], and its methodological quality is ratherlow (grade C; see Appendix 9). This study in-volved 45 patients with an elevated PSA afterlocal curative-intent therapy (prostatectomy,



radiation therapy and cryosurgery). The resultsof the whole-body PET scan were comparedwith those of the CT scan of the pelvis and ab-domen in all the patients and with those of amonoclonal antibody scan in 22 of them.

The study found similar detection rates for PETand computed tomography but also showed theinferiority of the monoclonal antibody scan inpatients with a markedly high PSA. The contri-bution of these three imaging techniques seemslimited for detecting distant metastases in pa-tients with a slightly elevated PSA, but this re-sult might have been due to a low incidence ofmetastases in such cases.

The authors conclude that other studies areneeded to determine the benefits of using oneor more imaging modalities when choosing theappropriate treatment for patients with PSArelapse.

Since our searches identified only one incon-clusive study [Seltzer et al., 1999] and since itsmethodological quality does not meet the se-lection criteria, no conclusion can be drawn infavour of the use of PET in prostate cancer.This is therefore a potential application.

4.1.7.5 Conclusions

In prostate cancer, FDG-PET has, althoughthere are presently no data to demonstrate itsefficacy, potential clinical utility in:

� detecting recurrence or residual tumors.

However, its use, as proposed in recent articles(under evaluation), is sparking keen interest.

4.1.8 Other cancers

Mention is made in the literature of PET'sclinical utility in other specific situations forthe following cancers, which are not discussed

in detail in this report: gynecological cancers(ovarian, uterine, cervical); certain genitouri-nary cancers (testicles, metastases from hy-pernephromas); mesotheliomas; soft-tissue sar-comas; thyroid cancer; and esophageal andpancreatic cancer.

4.1.9 Review of the positions published be-fore or in 2000 (oncology)

Current and potential users of PET in Québecfeel that this technology would provide greaterdiagnostic accuracy than the current conven-tional imaging by x-ray, ultrasound or mag-netic resonance. They foresee a significanttechnological delay, which could result in in-equities between the care provided to patientsin Québec and that provided in other industri-alized countries if PET is not deployed in Qué-bec [AMSMNQ, 2000; ARQ, 2000].

However, research is necessary to reinforceseveral aspects of the use of this technology,namely, finding new uses; lowering equipmentcosts; improving PET camera resolution; im-proving software performance; increasing thenumber of radioisotope distribution centres;conducting new studies of PET's validity; andbroadening the use of modified PET scanners[MHTAC, 2000].

Although the evidence concerning PET's clini-cal efficacy and cost-effectiveness in oncologyis insufficient to recommend unrestricted cov-erage of PET by various health-care programs(e.g., Medicare, BCBSA, VA-TAP), the evi-dence regarding its safety and its potentialclinical efficacy and potential cost-effectiveness do permit the recommendation ofinterim coverage in certain clinical situations.The MSAC states that, although it has beenshown that PET often leads to a change inmanagement, it is not always clear how thiswill impact the clinical outcome. It points out



the assumed relationship between a test's diag-nostic accuracy and its results on health out-comes is not restricted to PET and that there isseldom any evidence on the effects of diagnos-tic tests on health. It concludes that “[t]here isno direct evidence available at this time thatcan demonstrate that improvements in diag-nostic accuracy provided by PET, and any sub-sequent management changes, lead to im-provements in health outcomes for patients"[Ghersi et al., 2000].

Various conditions govern the different termsof coverage. For example, the coverage of PETby Medicare (U.S.) for diagnosis and staging inseveral areas of oncology is subject to specificconditions of acceptance: coverage is author-ized only if the PET scan results obviate theneed for an invasive diagnostic procedure orpermit one to accurately locate a tumor beforeperforming an invasive diagnostic procedure.Coverage is not authorized for any other diag-nostic use or for screening asymptomatic pa-tients [Tunis et al., 2000].

4.2 NEUROLOGY

In neurology, the following uses were exam-ined: dementia and Alzheimer's disease, re-fractory epilepsy and brain tumors (mainlyglioma).

Although neurology was one of the first areasin which PET was used, its utilization in clini-cal practice has not increased at the same paceas in oncology [Couillard, 2000; Ghersi et al.,2000].

The applications of PET in clinical neurologyare presently quite limited. With a few excep-tions, the information provided by PET cannotbe used directly to treat neurological or psychi-atric diseases [Beanlands et al., 1999].

4.2.1 Dementia and Alzheimer's disease


Close to 8% of Canadians aged 65 and overand close to 34% of those aged 85 and overhave Alzheimer's disease or related dementia[Alzheimer’s Society of Canada, 1999:www.alzheimer.ca].

Alzheimer's disease is the most common ofseveral neurodegenerative dementias that havesimilar clinical manifestations. This is whythere are three objectives to the diagnosticevaluation of elderly individuals with cognitiveproblems, namely, to determine: 1) if the pa-tient has dementia; 2) if the clinical picture andcourse confirm the presence of Alzheimer'sdisease; and 3) if there is another primary causeof dementia or a coexisting disease (especiallya cerebrovascular disease or Parkinson’s dis-ease or, less often, Pick's disease or primaryprogressive aphasia) that is contributing to thedevelopment of Alzheimer's disease or result-ing in an atypical manifestation [Bennett,2000a and 2000b].

The clinical diagnosis of the likelihood of Alz-heimer's disease is based on many differentsources of information that are considered to-gether, since there is no single biologicalmarker of this disease or other types of demen-tia (e.g., cerebrovascular and frontotemporaldementias) [Almkvist and Windblad, 1999].There is no reliable diagnostic test for Alz-heimer's disease, although certain blood andcerebrospinal fluid tests and neurological im-aging procedures are currently being re-searched [Bennett, 2000a and 2000b].


Various diseases can cause dementia, memoryloss and symptoms similar to those of Alz-heimer's disease, for which there is no treat-ment. PET can be used to confirm the degen-



erative process associated with this disease be-fore the onset of conclusive clinical symptomsand to make a differential diagnosis betweenAlzheimer's disease and diseases that are treat-able or reversible. Early diagnosis enables pa-tients and their families to plan the patient'senvironment and provide the necessary homemaintenance resources. Psychosocial tech-niques and pharmacologic treatments forslowing the progression of the disease are nowavailable and can improve these patients' qual-ity of life [Adams et al., 1999; AMSMNQ,2000].

Despite the lack of therapeutic modalities forthis disease, PET might possibly play a role inthe differential diagnosis of dementia and othercognitive diseases (e.g., Parkinson's disease).



� Diagnostic

�/� The HCFA recognizes PET's potentialfor this use. The VA-TAP, AHTAA andINAHTA recognize this potential aswell but point out that, since there is notreatment for this disease, it is unlikelyfor PET to have any diagnostic value inthis application.

The HCFA referred the matter of PET andAlzheimer's disease to the Medical CoverageAdvisory Committee (MCAC) Diagnostic Im-aging Panel in 2000. For the time being, the

use of PET for diagnosing this disease is notcovered [Tunis et al., 2000].

In the 1998 update of its 1996 report, the Vet-erans Affairs Technology Assessment Program(VA-TAP) concludes that "[t]he value of im-proved diagnostic information to AD patientsand their families should not be dismissed;however, this value should be quantified in thecontext of accessibility and accuracy of alter-native imaging technologies and of phenotypi-cally or genetically defined subsets of AD. Inthe absence of effective treatments for AD, anaccurate diagnostic test may be needed primar-ily in research for epidemiologic studies andevaluations of potential therapies" [Flynn andAdams, 1998].

Between the publication of the VA-TAP reportand February 2001, two new studies wereidentified. One of them [Reiman et al., 2001]involved only four patients. The other [Kawanoet al., 2001] examined the relationship betweenvarious IQ tests and the Mini-Mental State Ex-amination and regional cerebral FDG uptake.Neither of these studies is applicable, based onour selection criteria, and since there is no ef-fective treatment for this disease, the use ofPET to diagnose Alzheimer's disease or relateddementia is not recognized.

4.2.1.5 Conclusions

In Alzheimer's disease, the clinical utility ofPET is not recognized.

4.2.2 Refractory epilepsy


About 1% of the general population has epi-lepsy. It affects about 300,000 Canadians, andeach year 1 Canadian in 2,000 is diagnosedwith it, which translates into approximately14,000 new cases per year [Epilepsy Canada,1998: http://www.epilepsy.ca/eng/mainSet.



html].

Twenty to 30% of epileptics are refractory toantiepileptic therapy [Leung et al., 1999]. Themost frequent type of refractory epilepsy is thepartial complex seizure, which, in most cases,originates in the temporal lobe (temporal lobeepilepsy). For patients whose epilepsy is re-fractory to the usual treatments, surgery maylead to complete or partial seizure control. Thechoice of surgical treatment depends on the ex-act location of the epileptogenic focus [Leunget al., 1999].

It has been suggested that more than 200,000epileptics in the United States could benefitfrom surgical treatment, but the need to per-form numerous, complex preoperative tests(e.g., using conventional imaging [ictal andinterictal], various invasive methods and EEG)greatly limits these interventions [Devous etal., 1998].

The cognitive and psychosocial sequelae ofconstant epileptic seizures in a child should beconsidered differently from those occurring inadults because they can lead to developmentaland growth stagnation. For each individualchild, the potential risk/benefit ratio for surgeryshould be carefully weighed. Some studieshave shown that delaying surgery for child-hood-onset epilepsy until adulthood can lead tosignificant psychosocial, behavioural and edu-cational problems [Wyllie, 1998].


The advantage of interictal PET over other di-agnostic modalities is that it can be performedbetween seizures, unlike ictal digital SPECT,for example, which must be performed severaltimes. In addition, PET can reveal areas of cor-tical dysplasia that cannot be visualized withmagnetic resonance imaging, delineate the areaof epileptic dysfunction in conjunction withEEG and digital SPECT, substitute for preop-

erative functional localization tests, especiallyin pediatric patients, and be used for the preop-erative evaluation [Carmant, 2000]. Further-more, the combined use of PET and MRI per-mits the precise localization of epileptogenicfoci and obviates the need for invasive moni-toring by deep electrodes, a labour-intensiveand expensive procedure whose results are dif-ficult to interpret [AMSMNQ, 2000].



� Determining the location of epileptogenic foci

� /� The HCFA, MSAC and BCBSA-TECrecommend coverage of PET for thisuse. The AHTAA also recognizes theutility of PET in such cases, while theINAHTA concludes that the quality ofthe efficacy evidence for interictal PETin epilepsy is lacking.

The authors of the MSAC report base theirconclusion concerning the clinical utility ofPET for localizing epileptogenic foci in epi-lepsy on five studies [Swartz et al., 1992, 37patients; Ferrie et al., 1996, 32 patients; Del-beke et al., 1996, 38 patients; Chee et al., 1993,40 patients; and Markand et al., 1997, 67 pa-tients]. The authors state that the sensitivitywas relatively high but mention certain prob-lems that should be taken into consideration,such as reference standards based on a combi-nation of tests; the fact that the surgical out-come was used to measure the accuracy of PET



(only patients with positive PET results under-went surgery, which results in 100% sensitiv-ity); and patient selection (most studies se-lected patients who had undergone surgery).The authors did not provide any information onthe patients who had undergone a presurgicalworkup but who were not operated on, whichwould have included patients who may havehad a negative result if they had had a PETscan, etc. [Ghersi et al., 2000].

The Australian report nonetheless states that"[it] is reasonable to conclude that a subset ofpatients who would be helped by surgery willbenefit as a result of a positive PET scan. It isunclear how much PET is helping all patientswith refractory epilepsy, because we lack in-formation regarding false-negative PET re-sults...and patients testing positive by PET butnot undergoing surgery." The authors never-theless recommend interim coverage of PETfor evaluating patients with refractory epilepsywho are being considered for surgery in thecontext of a comprehensive epilepsy program,where localization based on a standard assess-ment, including EEG, MRI and semiology, isinconclusive.

The HCFA bases itself on a study that theAustralians used as well [Delbeke et al., 1996].The study showed that PET had a high positivepredictive value for postsurgical improvement(94%). The PET and EEG findings were con-cordant in 86% of the cases. The authors of thestudy concludes that, in terms of health out-comes, the impact of PET's diagnostic accuracywould be the ability to quantify postsurgicaloutcomes more accurately [Tunis et al., 2000].

To underscore this positive impact on morbid-ity and quality of life, the HCFA also mentionsthe study by Engel et al. [1990], who reportthat some patients may have avoided an inva-sive EEG (and other, noninvasive tests), thanksto localization of the epileptogenic focus byPET.

Three studies between 1999 and February 2001were identified [Dupont et al., 2000; Muzik etal., 2000; and Hwang et al., 2001]. Of thesethree studies, only that by Hwang et al. [2001]compared FDG-PET with conventional imag-ing.

Hwang et al.’s study [2001], which involved117 patients (grade B), compared PET withinterictal SPECT, ictal SPECT and MRI. Itshowed that PET can localize epileptogenicfoci with 77.7% accuracy versus 59.8% forMRI and 70.3% for ictal SPECT (concordancerate between the three modalities: 38%). Theaccuracy rate for the localization of neocorticalepileptogenic foci was 86.7% for PET, with64% and 80.6% for MRI and ictal SPECT, re-spectively. PET localized extratemporal epi-leptic foci with 70.7% accuracy, which wasbetter than MRI and SPECT (56.7% and63.6%, respectively). The authors found thatPET was the most sensitive of the three meth-ods in detecting each substrate of epilepsy andthat MRI was as sensitive as PET in detectingtumors causing certain epilepsies and the leastsensitive in detecting neuronal migration disor-der.

They also discuss the impact of imaging on thehealth of individuals with refractory epilepsy,reporting good surgical outcomes in 81.4% ofthe patients with imaging abnormalities (com-pared to 59.5% for those without imaging ab-normalities). The authors conclude that PETand ictal SPECT are generally more sensitivethan MRI (despite the low concordance ratebetween the two tests and variable sensitivitythat depended on the substrate), that the detec-tion of abnormalities by MRI is associated withfavourable outcomes and that PET and ictalSPECT can be used in a complementary fash-ion, especially when the MRI is negative.

Hwang et al.'s [2001] study supports the con-clusions of the MSAC's and HCFA's reports



and provides additional data for recognizingthe clinical impact that PET can have on healthoutcomes in individuals with refractory epi-lepsy while at the same time obviating the needfor numerous invasive, diagnostic procedures.

4.2.2.5 Conclusions

In refractory epilepsy, the clinical utility ofPET

is recognized for:� localizing epileptogenic foci in patients

with refractory epilepsy who are being con-sidered for surgery and where inconclusivelocalizing information is provided by astandard assessment, including seizuresemiology, EEG and MRI.

4.2.3 Brain tumors (mainly glioma)


It is estimated that there will be 2,400 newcases of brain cancer and 1,550 deaths due tothis disease in Canada in 2001. Québec is theCanadian province where there are the mostbrain tumors, with an estimated age-adjustedincidence for 2001 of 10 in 100,000 males and7 in 100,000 females compared to an averageof about 7 in 100,000 males and 4 in 100,000females in the other provinces [NCIC, Cana-dian Cancer Statistics, 2001].

Although various types of tumors can occur inthe brain, nearly 50% of these neoplasms de-rive from glial cells. The other 50% includeother types of tumors, such as metastases andmeningiomas, which are the most common.Gliomas are the only brain tumors that havebeen evaluated in detail by PET [Kuwert andDelbeke, 2000].

Magnetic resonance imaging and computedtomography are the two most commonly used

imaging techniques for the detection and dif-ferential diagnosis of brain lesions, but in somecases, they are unable to distinguish betweenneoplastic and nonneoplastic processes. It canalso be difficult, with these diagnostic methods,to differentiate between a low-grade gliomaand chronic inflammation or between a high-grade brain tumor and certain benign proc-esses, such as toxoplasmosis or a hemorrhagicinfarction [Kuwert and Delbeke, 2000].

The staging of brain tumors is based on tumorcontrast, which can be accomplished by com-puted tomography or magnetic resonance im-aging [Kuwert and Delbeke, 2000].


In glioma, the following uses were examined:evaluating the malignant transformation of alow-grade glioma, preoperative workup, de-termining the histology of the tumor, choice oftreatment, prognosis, detecting metastases,posttreatment follow-up (differentiating be-tween radionecrosis and recurrence).

FDG-PET is not suitable for differentiating alow-grade glioma from a benign process, sincelow-grade gliomas do not take up FDG. FDG-PET can, however, be used for correctly differ-entiating a high-grade glioma from nonneo-plastic processes, since such gliomas exhibithigh FDG uptake. FDG-PET can also differen-tiate, for example, between a hemorrhagic in-farction and a malignant astrocytoma [Kuwertand Delbeke, 2000].

FDG-PET is often used as a complement tomagnetic resonance imaging, and, in certaincases, the correlation between PET images andamplified MRI images is crucial for differenti-ating between FDG uptake by high-grade tu-mors and its uptake by normal cortex. False-positive results can occur in patients with gradeI gliomas and certain low-grade oligodendro-gliomas, which exhibit high FDG uptake



[Couillard, 2000; Kuwert and Delbeke, 2000].

The level of FDG accumulation in gliomas isindicative of the degree of malignancy. PETcan therefore be used both for staging braintumors and for guiding the biopsy toward themost undifferentiated region (maximum meta-bolic activity). PET-guided biopsy increasesdiagnostic accuracy to nearly 100% [Kuwertand Delbeke, 2000].

During follow-up, computed tomography andmagnetic resonance imaging cannot always dif-ferentiate between radionecrosis and recurrenttumors. PET is an important tool for follow-uppurposes, since high-grade tumors exhibithigher FDG uptake than radionecrosis. How-ever, the clinical utility of PET in detecting re-currence is not always reliable, for the moreaggressive the radiation therapy, the lower thediagnostic efficacy. Since low-grade gliomasdo not accumulate FDG, functional imagingwith radiolabelled amino acids is one of theonly modalities with digital SPECT thatcan detect a recurrent low-grade glioma[Couillard, 2000; Kuwert and Delbeke, 2000].

At present, functional imaging with PET can-not substitute for computed tomography ormagnetic resonance imaging but does provideadditional information that cannot be obtainedby morphological imaging alone [Kuwert andDelbeke, 2000].


The list below is based on the positions (con-clusions and decisions) of organizations thatused explicit assessment criteria. In practice, ofthe more recent reports, those of the MSACand HCFA are cited first. Expert opinions areincluded to shed further light on the subject, ifneed be. A complete list of the positions is pro-vided in Appendix 7. The studies selected andevaluated for the updates using the criteria setout in Chapter 3 and in Appendix 6 are pre-

sented in Appendix 8.

The MSAC mentions several potential uses ofPET in glioma. However, it limited its assess-ment to differentiating between radionecrosisand recurrent glioma, based on a retrospectiveaudit of clinical records by Deshmukh et al.[1996, 75 patients], who found that 87% of thePET scans between September 1990 and June1992 in cases of glioma were for the purpose ofdifferentiating between radionecrosis and tu-mor recurrence [Ghersi et al., 2000].

� Evaluating the malignant transformation of alow-grade glioma

� The MSAC, MHTAC and INAHTArecognize PET's potential for thisapplication.

Four new studies [De Witte et al., 2001; Earyet al., 1999; Derlon et al. 2000; and Roelcke etal., 1999] were identified, but all are methodol-ogically flawed. De Witte et al. [2001] con-clude that FDG-PET is not superior to patho-logic grading for predicting survival. Eary et al.[1999], Derlon et al. [2000] and Roelcke et al.[1999] compared FDG-PET with C-11 me-thionine (MET) PET. Their findings were con-cordant. Results with FDG-PET were inferiorto those with MET PET, which sharply deline-ated the tumors in most of the cases. Despitethe poor quality of these studies, these resultsindicate that C-11 methionine PET has a po-tential role in identifying tumors progressing tomalignancy.

Since this update did not uncover any robustevidence, this use of PET is a potential appli-cation.

� Preoperative workup

� The MSAC and INAHTA state that PEThas potential for this use.



No new studies were identified. This use ofPET is a potential application.

� Grading tumors

�/� The MSAC and INAHTA mention thisuse and recognize PET's potential for it.In the MHTAC's opinion, there are nohard data in favour of PET for this ap-plication.

No new studies were identified. This use ofPET is a potential application.

� Choice of treatment

� The MSAC, MHTAC and INAHTAmention PET's potential in the man-agement of patients with brain tu-mors.

A single study (grade C) was identified. Nuuti-nen et al. [2000] studied the impact of C-11methionine PET on the planning of the man-agement of patients with low-grade astrocy-tomas. Their results suggest that this modalitywas helpful in choosing the treatment in only27% of the patients, whereas PET used incombination with MRI was helpful in 46% ofthe patients. This update does not permit rec-ognition of this use of PET, which remains apotential application.

� Prognosis

� Only the MSAC mentions this use andrecognizes PET's potential in this re-gard.

Up to February 2001, two studies on deter-mining the prognosis of glioma by PET wereidentified. One of them [De Witte et al., 2001],although of poor quality, did not demonstratePET's superiority to the pathologic examinationin grading brain tumors. The other [Nuutinen etal., 2000], also a grade C study, suggests a

strong association between MET-PET and theprediction of survival in patients with a stan-dardized uptake value greater than 3.5. Sincethese results are not convincing, this remains apotential application of PET.

� Detecting metastases

� The MHTAC and INAHTA mention PET'spotential for this application.

No new studies were identified. This is a po-tential use of PET.

� Posttreatment follow-up: differentiating be-tween radionecrosis and recurrence

� /� The AHTAA and MSAC support theuse of PET for this use. All the potentialusers consider that PET has clinicalutility in this application. The MHTACand INAHTA recognize its potential forthis use.

Although the MSAC based its conclusionsconcerning the use of PET for evaluating glio-mas on 16 studies, the authors state that thereare too few data to draw any firm conclusionsas to the superiority of PET to SPECT with re-gard to differentiating radionecrosis from re-current tumor. In addition, the studies of the di-agnostic accuracy of PET vary widely from amethodological standpoint. None of the studiesexamined in the Australian report for the pur-pose of making a coverage recommendationconcerning the differentiation of radionecrosisfrom recurrent glioma in patients who have re-ceived radiation therapy and who have residualstructural abnormalities on diagnostic imagingclearly demonstrated PET's superiority. Fur-thermore, the Australian report cites a study byBader et al. [1999, 30 patients], which, al-though limited by a small patient sample,seems to suggest that FDG-PET's efficacy indistinguishing between radionecrosis and re-currence varies considerably according to thegrade of the glioma.



Bader et al. [1999] compared the efficacy ofFDG-PET and that of I-123 SPECT in detect-ing and grading recurrence in patients previ-ously treated for glioma. Based on the histo-pathological data, SPECT and PET findingsconfirmed recurrence in all the cases of high-grade (IV) gliomas. However, for grade II andIII gliomas, PET was inferior to SPECT, withtrue-positive results in 71% of the cases ofgrade III and 50% of the cases of grade II tu-mors compared to 86% (grade III) and 75%(grade II) for SPECT. The authors concludethat grading brain tumors by PET is hinderedby problems confirming the malignant progres-sion of a grade II glioma to a grade III or IVglioma. Nonetheless, in all, FDG-PET con-firmed an upgrading in 75% of the cases.

The MSAC also cites a retrospective audit ofclinical records by Deshmukh et al. [1996],which showed that PET had a positive impacton the clinical management of patients. Theauthors of this report state that little is knownabout the effects of PET on the health of pa-tients with recurrent glioma, but that it wouldbe reasonable to expect improvements in theirhealth in terms of morbidity, mortality andquality of life. Deshmukh et al. [1996] foundthat the FDG-PET findings led to the consid-eration of a new therapy in 31% of the cases,including the decision to initiate chemotherapyin 21% of the cases and surgery in 10% of thecases, with a biopsy in only 2/89 cases and sur-gical resection in 7/89 cases. The PET findingsalso contributed to a decision to withhold ag-gressive therapy in 59% of the cases: chemo-therapy in 15% of the cases, surgery in 7% ofthe cases, radiosurgery in only 2/89 cases andunspecified "aggressive action" in 36% of thecases. These decisions were made in light ofPET results only in 28% of the cases and ofPET results supported by other information in72% of the cases. In general, the authors con-clude that PET played a valuable clinical rolein 86 of the 89 cases.

Despite the methodological weaknesses, whichreduce the convincingness of the availabledata, the MSAC recommends that PET befunded on an interim basis for the differentia-tion of radionecrosis from tumor recurrence inpatients treated with radiation therapy whohave residual structural abnormalities on diag-nostic imaging [Ghersi et al., 2000].

Our searches did not reveal any new studies ondifferentiating between radionecrosis and re-currence. The MSAC's conclusion is statedhere: This use is recognized until such time asnew data confirm or invalidate this conclusion.

4.2.3.5 Conclusions

In brain tumors (mainly gliomas), the clinicalutility of PET is

recognized for:� evaluating residual lesions after treatment

of a recurrent glioma and differentiatingbetween radionecrosis and recurrence inpatients treated with radiation therapy whohave abnormalities on diagnostic imaging;

and has potential for:� the initial staging of patients suspected of

having a primary brain tumor, in order toguide biopsy to the highest area of activity;

� evaluating the progression of a low-gradeglioma to malignancy;

� the preoperative workup;� grading tumors;� the choice of treatment;� determining the prognosis; and� detecting metastases.

4.2.4 Other clinical uses in neurology andpsychiatry

Other avenues of application in neurology andpsychiatry are mentioned in the literature, butthey are not examined in detail in this study.



They include Parkinson's disease, neurologicalpain, multiple sclerosis, language problems andschizophrenia.

4.2.5 Review of positions published in or be-fore 2000 (neurology)

The need for PET in the field of neurology ispresently quite limited because the specific in-formation that it provides cannot, at this time,be used for therapeutic purposes, with a fewexceptions. However, knowledge in this field isevolving, which could lead to a rapid increasein the demand for PET. This is why the strat-egy concerning the use of PET in neurologyshould be one of close observation, includingthe integration of limited applications at exist-ing PET centres. The CMI recommends thatthe need for PET in neurology be monitoredclosely and that services be made available atcertain tertiary care centres for validated neu-rological applications [CMI: Beanlands, 1999].

In light of the available data, the INAHTA rec-ognizes PET's contribution to our knowledgeof the biochemical and physiological mecha-nisms of several cerebrovascular and neuro-psychiatric diseases but does not conclude thatthe information provided by PET can or cannotimprove patient management or final healthoutcomes. The available evidence is based onsmall, methodologically flawed studies.

The clinical indications for PET in neuropsy-chiatry have been under investigation since theearly 1980s, and after two decades, questionsabout its clinical utility persist and are limitingits diffusion in clinical practice. Furthermore,PET's diagnostic value in a number of neuro-logical applications has been called into ques-tion because there is no treatment for improv-ing the prognosis of certain diseases, e.g., Alz-heimer's dementia [INAHTA: Adams et al.,1999].

4.3 CARDIOLOGY

The research methodology used was explainedabove (Chapter 3 and Appendix 6). Briefly, itconsists of a synthesis of the available assess-ment reports and an update of the recent lit-erature published on the subject

4.3.1 General data

4.3.1.1 The pathophysiology of myocardialimpairment

Viable myocardium is that which still has theability to contract. Viable myocardium can benormal, stunned (postischemic contractile dys-function) or hibernating (persistent but poten-tially reversible deterioration of myocardialfunction). Briefly, myocardial stunning is a re-versible state of regional contractile dysfunc-tion that occurs after an ischemic episode, evenin the absence of myocardial necrosis. It can bean acute ischemic episode or repeated ischemicepisodes, symptomatic or nonsymptomatic, in apatient with one or more coronary stenoses.Resting myocardial blood flow may be rela-tively normal. On the other hand, myocardialhibernation is a state of potentially reversiblechronic ventricular dysfunction associated withpermanent, inadequate blood supply. Thesetwo terms describe different pathophysiologicalprocesses. However, in clinical practice, thedividing line between the two states is not al-ways clear.

� Why evaluate myocardial viability?Several studies have shown that revasculariza-tion techniques (bypass and angioplasty) im-prove global and regional ventricular dysfunc-tion in coronary patients [Rees et al., 1971;Chatterjee et al., 1973; Brundage et al., 1984;Braunwald et al., 1986; Rahimtoola, 1989; Vanden Berg et al., 1990; Elefteriades et al., 1993;and Vanoverschelde et al., 1997], which im-proves quality of life, reduces the occurrence of



cardiac events and can prolong life expectancy.Thus, it is estimated that the left ventricularejection fraction (LVEF) can improve signifi-cantly after revascularization in 25 to 40% ofcoronary patients with global left ventriculardysfunction [Bonow et al., 1996].

In patients with coronary ischemia and more orless preserved left ventricular function, myo-cardial revascularization, if indicated, involvesa small surgical risk [Kennedy et al., 1981].However, in patients with severe left ventricu-lar dysfunction (LVEF ≤ 35%), revasculariza-tion provides little or no benefit if the myocar-dial tissue is not viable. The mortality rate as-sociated with bypass surgery in such patients is5 to 37% [Zubiate et al., 1977; Hellman et al.,1980; Hochberg et al., 1983; Marwick et al.,1995]. It is for this category of patients thatstudying myocardial viability is so important[Bax et al., 1998], for it is this population thatbenefits the most from revascularization in thepresence of viable myocardium [Yusuf et al.,1994]. Patients with low myocardial viability,as evaluated preoperatively, have a poorerprognosis than those who have better viability[Pagley et al., 1997; Pasquet, 1999]. Patientswith nonviable myocardium do not benefitfrom revascularization and must be treatedpharmacologically or undergo transplantation.However, in patients in whom transplantationis considered and in whom a sufficient quantityof viable myocardium has been detected, thestrategy should be modified in favour of myo-cardial revascularization. By identifying pa-

tients with viable myocardium, the therapeuticoptions (medical, revascularization or trans-plantation) will be optimized and the prognosisimproved accordingly.

In Québec, 6,000 to 7,000 coronary bypassesand 9,000 to 10,000 angioplasties are per-formed each year, increasingly in patients withsevere left ventricular dysfunction. These by-passes and angioplasties cost $100 million ayear. To ensure optimal resource utilization,patients who are to undergo revascularizationprocedures should be properly selected. Thedetection of myocardial viability is a deter-mining factor in the process of identifyingthese patients, especially those with severe leftventricular dysfunction.

4.3.2 The role of PET in cardiologyPET can be used to characterize and quantifythe different functions of cardiac tissue. Its usein research has provided new information aboutcardiac physiology and pathophysiology[Camici, 2000].

The isotopes most widely used in cardiologyare listed in the table below, together with theirhalf-lives, the parameters they measure andtheir source of production. Because of the veryshort half-life of most of the radiotracers usedin cardiology, the use of PET in this field re-quires a cyclotron on the same site as the scan-ner.

Table 2: Radiopharmaceutical tracers used most often for PET in cardiology

[Pirich and Schwaiger, 2000]ISOTOPE RADIOPHARMACEUTICAL HALF-LIFE PARAMETER MEASURED SOURCERb-82 Rubidium 76 sec Blood flow GeneratorO-15 Oxygen 2 min Blood flow CyclotronN-13 Ammonia 10 min Blood flow CyclotronC-11 Acetate 20 min Oxidative metabolism/ Cyclotron



perfusionF-18 Deoxyglucose 109 min Metabolism Cyclotron

4.3.2.1 Myocardial perfusion studies

Oxygen-15-labelled water and N-13 ammoniaare the most widely used radiopharmaceuticaltracers for quantifying regional myocardial per-fusion by PET. One can also use rubidium-82or Cu-62-PTSM (copper-62-pyruvaldehydeIIbis(N4-methyl)thiosemicarbazone). UnlikeFDG, these tracers have a very short half-life(from a few seconds to a few minutes). As a re-sult, a cyclotron is required on the premiseswhere they are used. Presently, coronary angi-ography is the gold standard for studying myo-cardial perfusion. Myocardial perfusion canalso be studied with single-photon emissioncomputed tomography (thallium-201 or tech-netium-99m-sestamibi). The advantage of PETover SPECT is that it enables one to quantifydifferent parameters and accurately correct forthe photon attenuation caused by soft tissues,which sometimes make the scans difficult tointerpret (obesity, mammary or diaphragmaticattenuation).

PET can also be used to study coronary vaso-dilation reserve (study of coronary microvas-cular function). Quantifying this reserve is use-ful in evaluating the functional significance ofcoronary stenoses.

Perfusion studies have other potential uses, butthey have yet to be validated: follow-up on theeffects of angioplasty or cholesterol-loweringagents, the postbypass evaluation of blood flowin transplants, the detection of restenoses, andposttransplant evaluation.

4.3.2.2 Study of the metabolism of free fattyacids, glucose and oxygen

Myocardial glucose uptake can be assessedwith PET and the glucose analog 18-fluorodeoxyglucose (FDG). Glucose plays akey role in metabolism in ischemic myocar-

dium. In normal myocardium, fatty acids arethe primary energy substrate for the myocytes'energy metabolism, but in ischemic myocar-dium, glucose becomes their primary energysource. Glucose uptake can be measured withFDG, which is transported across the mem-brane in the same manner. In ischemic and hy-poxic myocardium, FDG competes with glu-cose. It is phosphorylated to FDG-6-phosphateby an enzyme, hexokinase, and cannot subse-quently be dephosphorylated or metabolized. Ittherefore remains trapped in the myocardium.FDG uptake is increased or preserved in sev-eral altered states of the myocardium associ-ated with contractile dysfunction and/or re-duced perfusion, which indicates persistence ofmetabolic activity and therefore residual vi-ability with a potential for recovery after revas-cularization.

Increased or preserved FDG uptake has beenobserved during ischemia, postischemic myo-cardial stunning and myocardial hibernation.

4.3.2.3 Detection of myocardial viability

There are several methods for studying myo-cardial viability. They include nuclear imagingtechniques, such as late redistribution thallium-201 imaging after rest injection, resting tech-netium-99m perfusion imaging, 18F-FDG-SPECT and 18F-FDG-PET. Dobutamine stressechocardiography is widely used as well, and ithas also been proposed that nuclear magneticresonance imaging could be used.

Viable myocardium can exhibit normal or di-minished coronary blood flow [Haas et al.,1997]. If resting metabolism is preserved, glu-cose utilization may be normal or increased.The presence of viable myocardium can bedetected by evaluating both coronary blood



flow and metabolism.

PET scans (perfusion/metabolism) can yieldthree possible results:

1 . Normal coronary blood flow and normalFDG uptake.

2 . Reduced blood perfusion and preserved orincreased FDG uptake (perfusion/metabolismmismatch).

3. A proportional reduction in blood flow andFDG uptake (perfusion/metabolism match).

Areas of the myocardium where there is con-cordance between reduced blood flow (meas-ured by means of N-13 ammonia, O-15-labelled water or rubidium-82) and decreasedglucose utilization (measured by FDG uptake)are nonviable infarcted areas. However, if thereis decreased myocardial blood flow with pre-served FDG uptake, the myocardium is consid-ered viable. Thus, scenario 2 depicts a potentialfor reversible myocardial dysfunction (viablemyocardium), while scenario 3 depicts irre-versible myocardial dysfunction and thereforeimplies infarcted areas.

4.3.2.4 Study of autonomic cardiac function

PET technology can be used to conduct uniquestudies of the links between the heart and nerv-ous system [Schwaiger et al., 1990a and 1990b;Bengel et al., 1999; Stevens et al., 1999,Tamaki, 1997]. For example, one can studycardiac innervation with specific neuronal trac-ers, pre- and postsynaptic circuits, and thenumber, density, location and quality of the re-ceptors in the autonomic nervous system[Wieland et al., 1990; Goldstein et al., 1990;Allman et al., 1993; Calkins et al., 1993a and1993b; Lefroy et al., 1993; Schafers et al.,1998a and 1998b; Merlet et al., 1993; Choud-hury et al., 1996; Law et al., 2000]. The natureof the interactions between the sympathetic andparasympathetic nervous systems and the rela-

tionship between innervation and the regulationof coronary blood flow can be studied as well[Di Carli et al., 1997; Stevens et al., 1998a].These imaging and neurocardiac characteriza-tion methods, which are presently unique toPET, are expanding our knowledge of thepathophysiology of heart failure [Goldstein etal., 1990; Merlet et al., 1993]; cardiac trans-plantation [Di Carli et al., 1997, Stevens et al.,1998a]; dilated [Merlet et al., 1993] and re-strictive and hypertrophic [Lefroy et al., 1993;Schafers et al., 1998a; Choudhury et al., 1996]cardiomyopathies; myocardial infarction [All-man et al., 1993]; diabetic cardiomyopathy[Stevens et al., 1998a and 1998b]; severe car-diac rhythm disturbances [Calkins et al., 1993a;Schafers et al., 1998b]; and sudden death[Calkins et al., 1993b]. The tracers used are11C-hydroxyephedrine (11C-HED), a norepi-nephrine analog [Di Carli et al., 1997]);[11C](S)-CGP 12177, a nonselective beta-antagonist employed in postsynaptic studies;[11C]GB67, which is used to study alpha-1 re-ceptors [Law et al., 2000]; and [11C]m-hydroxyephedrine, which is used to study pre-synaptic sympathetic innervation. Presently, atleast 15 or so radiotracers are being used in thisarea of research [Tamaki, 1997].

Another, rather new but extremely promisingfield is the exploration, using PET, of gene ex-pression in cardiac physiology and pathophysi-ology [Gambhir et al., 1999].

4.3.2.5 Follow-up of transplanted patients

In addition to PET being used as a tool for se-lecting cardiac transplantation candidates, sev-eral studies have employed PET in the follow-up of transplanted patients for the purpose ofevaluating regional sympathetic reinnervationof heart transplants [Bengel et al., 1999; Uber-fuhr et al., 2000a; Uberfuhr et al., 2000b]. PETis also used to study posttransplant myocardialperfusion [Kushwaha et al., 1998] and the pro-



gression of atheromatous disease in grafts [Al-len-Auerbach et al., 1999].

4.3.2.6 Monitoring the effect of therapeuticinterventions

PET can be used to monitor coronary vasomo-tor response to various therapeutic interven-tions, whether pharmacologic or risk factormodifications [Czernin et al., 1995]. Studieshave revealed, by means of PET, that risk fac-tor modifications accompanied by healthierliving results in a decrease in the severity ofstress-induced myocardial perfusion abnor-malities [Gould et al., 1995]. In addition, theliterature contains several studies of coronaryflow reserve changes during cholesterol-lowering therapy [Gould et al., 1994; Guethlinet al., 1999; Yokoyama et al., 1999; Baller etal., 1999]. PET has also been used to studymyocardial perfusion under the effect of vera-pamil in patients with hypertrophic myocardio-pathy [Choudhury et al., 1999] and to quantifyendothelium-dependent coronary vasodilationduring the administration of L-arginine or hor-mone replacement therapy [Campisi et al.,1999a; Campisi et al., 1999b].

4.3.2.7 Other uses

Other applications of PET have been proposedfor studying 11C-acetate metabolism in conges-tive heart failure and cardiac arrhythmias (18F-dopamine, 11C-HED).

4.3.3 Previous positions and update

A more detailed summary of the following po-sitions is provided in Appendix 7.

The MSAC (March 2000) specifically studiedthe role of PET in evaluating coronary arterydisease in patients with left ventricular dys-function who, on digital SPECT, are found tohave nonviable myocardium or where viabilityis uncertain. The MSAC concludes that there is

presently not enough evidence to draw anyfirm conclusions regarding the clinical efficacyand cost-effectiveness of PET in this applica-tion. In most cases, PET is added to other diag-nostic modalities. Further assessments of thistechnology are necessary. The committee rec-ommends that FDG-PET be funded on an in-terim basis, on the condition that clinical and/oreconomic data from MSAC-approved, pro-spectively designed studies are provided to acentral body so as to enable more long-termdecisions to be made about the role of FDG-PET in clinical practice.

For the HCFA, as regards the impact of FDG-PET findings on patient health, in cases wherethe SPECT scan is negative and the PET scanis positive, there are insufficient literature datato state that a change in patient managementwould result in improved health outcomes. Incases where the SPECT scan is positive butquestions remain as to the appropriateness ofrevascularization, PET could be a promisingtest, based on the literature examined. Medi-care (U.S.) was already covering rubidium-82PET for evaluating myocardial perfusion at restor with pharmacological stress in the context ofmanaging patients with known or strongly sus-pected coronary artery disease, when PET wasused in place of SPECT or following a SPECTscan that was inconclusive. In December 2000,Medicare also began covering FDG-PET forevaluating myocardial viability following apositive SPECT scan, but when there is doubtas to myocardial viability if revascularization isconsidered. Coverage for this test in other ap-plications was referred for study to an advisorycommittee consisting of nuclear imaging ex-perts.

The INAHTA [1999] concludes that the meta-bolic information provided by PET may im-prove patient selection for revascularizationand increase the chances of successful surgery.PET may offer cost savings by eliminating un-necessary angiography or revascularization in



certain patients. As for evaluating myocardialperfusion, PET appears to have superior per-formance, but the improvement in performancein relation to digital thallium SPECT and theextent of its contribution to managing patientsare not clear. PET is more expensive than othernoninvasive strategies and has not yet replacedcoronary angiography for assessing coronaryartery disease. For patients at intermediate risk(25 to 50% probability of having either a 50%or greater left main coronary artery occlusionor a greater than 70% occlusion of an othercoronary artery), PET is not a cost-effectivealternative to immediate coronary angiographyor other noninvasive tests, such as stress echo-cardiography or SPECT. There are not enoughdata to determine the cost-effectiveness of PETin diagnosing coronary artery disease. For de-termining myocardial viability and/or predict-ing the risk of cardiac events, most assessmentshave found that PET has comparable sensitivityand superior specificity to other modalities,although the studies were few in number andoften lacked methodological robustness. As re-gards improving the likelihood of successfulrevascularization and cost savings, the data areinsufficient to confirm that PET is favourablycost-effective. As for monitoring the effective-ness of treatment in coronary patients whohave hypertension or cardiomyopathy, the evi-dence is insufficient. This application is stillbeing researched.

The Alberta Heritage Foundation for MedicalResearch [Cowley et al., 1999] reports thatPET and dobutamine stress echocardiographyhave the same level of diagnostic efficacy, butthat the evidence is limited. There is somemethodologically weak evidence pointing tothe predictive value of PET as regards theclinical outcome of patients who undergo thistest. In Alberta, the use of these methods toevaluate myocardial viability should be ac-companied by prospective studies with a long-term patient follow-up.

The reports submitted by the FMSQ,AMSMNQ and ARQ [September 2000] notethat the recognized and potential uses of PETare numerous. Accordingly, PET is the besttechnique for identifying ischemic but viablemyocardium in patients with compromised leftventricular function and for evaluating myo-cardial perfusion, thus permitting better patientselection for revascularization or a heart trans-plant. PET is considered to be an imaging mo-dality with an established role. The Québecpopulation's accessibility to it should thereforebe ensured, especially in cardiology. Over thenext few years, frequent use will also be madeof PET in cardiology research. The reports alsomention the very short half-life of the ra-diotracers used in cardiology and call attentionto the need to install cyclotrons at cardiologycentres so as not to compromise the use of PETin this field.

For the CMI [1999], too, FDG-PET imaginghas proven utility in evaluating myocardial vi-ability for the purpose of selecting patientswith reduced cardiac function for revasculari-zation and candidates for a heart transplant.Similarly, according to this organization,studying myocardial perfusion with PET hasproven utility in the diagnosis and prognosis ofcoronary artery disease; the diagnosis of coro-nary artery disease in the subset of patientsprone to false-positive results with conven-tional nuclear imaging; the evaluation of theprogression or regression of the disease in thecontext of therapeutic interventions; and thedetection of a decrease in coronary reserve inpatients with ischemic disease not associatedwith coronary atherosclerosis. The CMI rec-ommends that when deciding on the location ofPET centres, one take into account patientswith heart disease as one of the two most im-portant priorities, the other being oncology.

The tables below summarize the different posi-tions regarding the uses that were examined.



� Coronary perfusion

Diagnosis, prognosis and management of coro-nary artery disease

� /� Use recognized by the HCFA andBCBSA and considered a potential ap-plication by the INAHTA and MSAC,which include cost-effectiveness in theiranalyses (lack of high-level evidence).

Evaluating response to therapy

� Recognized use, according to potentialusers, and is in the process of being as-sessed for the INAHTA.

Detecting restenosis

NM Use recognized by the ARQ. Not men-tioned by assessment agencies, whichhave mainly examined myocardial vi-ability.

Detecting diminished coronary reserve in patientswith ischemia associated with noncoronary dis-ease

NMUse recognized by potential users butnot mentioned by the other organiza-tions.

� Myocardial viability

Initial evaluation and selection for revasculari-zation

� /� The HCFA recommends coverage forthis use of PET if the SPECT scan ispositive and there is some question as tothe appropriateness of revascularization.The MSAC recommends coverage if theconventional tests for evaluating viabil-ity are negative and the patient is beingconsidered for revascularization. For theINAHTA, PET is an effective evalua-tion method, but there is insufficientdata to determine its cost-effectiveness.The AHFMR considers this technology

effective but that its use should be ac-companied by prospective studies.

Evaluating patients for a heart transplant

NM Use recognized by potential users butnot mentioned by assessment organiza-tions.

The diversity of opinions and conclusions inthe different assessment reports stems from alack of data for drawing any firm conclusionsand clearly underscores the need for furtherclinical research in this area.

� UpdateWe focused our assessment on the benefit ofintroducing PET into clinical practice forevaluating myocardial viability in patients withheart failure who are being considered formyocardial revascularization. The second usefor which there is abundant literature is evalu-ating myocardial perfusion.

PET is more effective (sensitivity and specific-ity) in diagnosing coronary artery disease thansingle-photon emission computed tomographyor stress electrocardiography [Adams et al.,1999]. Presently, the limitations of this use ofPET are its high cost and low availability.Thus, although PET can provide additionalquantitative information in relation to coronaryangiography, the latter is the reference test forthis use.

There are insufficient data to draw any firmconclusions about the cost-effectiveness ofPET as a tool for diagnosing coronary arterydisease. For patients with an intermediate car-diovascular risk (25 to 50% probability ofhaving a 50% or greater left main coronary ar-tery occlusion or a greater than 70% occlusionof any other coronary artery), Garber's meta-analysis [1999] shows that PET is not a cost-effective alternative to immediate coronary an-



giography or other noninvasive tests, such asstress echocardiography or SPECT, for study-ing myocardial perfusion. However, this meta-analysis has certain limitations. In particular,only three PET studies where considered, withno direct comparison between the different di-agnostic modalities. However, other economicstudies, such as Patterson's [1995], have shownthat PET is the most cost-effective test in pa-tients with a pretest probability of coronary ar-tery disease of less than 70%.

The update was done after reviewing the lit-erature, in order to describe the current knowl-edge on the basis of published data on the sen-sitivity and specificity of PET in detectingmyocardial viability.

Most of studies that have evaluated PET havesignificant methodological flaws. Furthermore,there are few highly reliable data on the impactof using the technology on patient manage-ment.

Tables 33 and 34 (Appendix 8) show the twotypes of available studies. Table 34 presentsstudies that have evaluated PET using an in-termediate assessment criterion, it most oftenbeing the recovery of segmental myocardialmotion after revascularization, and that do notpermit a true assessment of the positive clinicalimpact of PET. Numerous factors can explainthe differences in the results of these studies[Cowley et al., 1999, Part 2]. In some of them,the calculations are based on all the dysfunc-tional segments, in others, only the successfullyrevascularized segments. In addition, the leftventricular ejection fraction varies from studyto study (coronary arteries affected to varyingdegrees). In other studies, different protocolswere used for PET scanning and for the seg-mental division of the ventricle.

Table 33 shows studies that used a clinical as-sessment criterion. Most of them were retro-spective and of little power and had methodo-

logical biases. However, their results may sug-gest that PET has a positive impact on clinicaloutcomes. Nonetheless, because of the absenceof comparison groups in most of the studies,one cannot be completely affirmative or, in anyevent, draw any conclusions.

Corroborating the results of Dreyfus et al.'sstudy [1994] of 46 patients, Akinboboye et al.[1999] showed, in 33 patients with ischemicmyocardiopathy in whom a heart transplantwas indicated (LVEF < 35%), that PET de-tected myocardial viability in 50% of the pa-tients for whom thallium-201 scintigraphy wasnegative. Thus, these patients underwent suc-cessful coronary bypasses with the same clini-cal outcome at 12 months as all the patients re-vascularized during the same year at the facil-ity in question. This therefore made it possibleto reserve heart transplants for the patients whoactually required them, thus optimizing the al-location of this scarce resource. The study byDuong et al. [1995] showed, in 112 patientswith ischemic cardiomyopathy (LVEF < 35%)who were being considered for a heart trans-plant, that 5-year survival in the patients withPET-evaluated myocardial viability who werevascularized was approximately the same asthat in the transplanted patients (80% vs. 71%).

The recent study by Siebelink et al. [2001] in-volved 103 patients and is the only random-ized, controlled study. Its objective was tocompare prospectively two strategies, onePET-guided, the other SPECT-guided, formanaging coronary patients who were beingconsidered for myocardial revascularization.All the patients underwent FDG-PET and tech-netium-99m-sestamibi SPECT and were sub-sequently randomized to two groups in order totake the test into account in the managementmodality (bypass, angioplasty or pharma-cologic treatment). After a follow-up of 28 ± 1months, there was no statistically significantdifference in terms of the occurrence of themain assessment criterion (cardiac event-free



survival) between the two groups (11 events inthe PET group vs. 13 in the SPECT group).The authors therefore conclude that either mo-dality can be used equally for making revascu-larization decisions in patients suspected ofhaving viable myocardium. It will be noted thatonly one third of the patients in this study hadan LVEF of less than 30%.

A randomized, prospective, multicentre Cana-dian study, PARR Phase 2 [Beanlands, 2000-2001], is presently underway. It has a recruit-ment goal of 412 coronary patients with severeleft ventricular dysfunction (LVEF < 35%).The study has a double objective. The mainobjective is to compare the clinical outcomes(composite criterion: cardiac deaths, cardiac ar-rests, infarctions, transplantation, admission tohospital for unstable angina, or heart failure)and the cost of PET-guided management withthe outcomes obtained without PET (standardapproach) at one and two years. The secondaryobjective is to compare left ventricular functionand quality of life at one and two years of fol-low-up on the basis of the two managementstrategies. The results of this study should pro-vide a better assessment of the clinical andeconomic impact of using PET in patients withsevere left ventricular dysfunction, patientswho benefit the most from revascularizationtechniques and in whom determining myocar-dial viability is essential for optimal manage-ment.

Presently, there are insufficient data to drawany conclusions, in light of the clinical assess-ment criteria, concerning efficacy and cost-effectiveness. Although it is of a high level ofevidence, Siebelink et al.'s study [2001] is alow-power study of all manner of coronary pa-tients. It would seem that it is not these patientswho benefit the most from PET scanning, butrather more severely affected patients or pa-tients with equivocal SPECT results. ThePARR 2 study [Beanlands, 2000-2001] could

provide answers to some of these questions andthus permit informed decision making.

4.3.5 The role of PET among the currentlyused nuclear medicine tests

In patients with minimal to moderate left ven-tricular dysfunction, the positive predictivevalue of stress echocardiography, digitalSPECT and PET for identifying hibernatingmyocardium is comparable (69 to 83%), with anegative predictive value of between 81 and90% [Winjs et al., 1998]. Dobutamine stressechocardiography is the least expensive andmost widely available test. However, in pa-tients with severe left ventricular dysfunction,the false-negative rate is greater than with nu-clear imaging techniques.

PET is the method with the highest predictivevalue in patients with left ventricular dysfunc-tion, according to Arrighi et al. [1997] andFath-Ordoubadi et al. [1998].

The Alberta Heritage Foundation's assessmentreport [Cowley et al., 1999] on the differentmethods of evaluating myocardial viability in-dicates sensitivity and specificity ranges for thedifferent diagnostic modalities discussed in theliterature:

� PET: Sensitivity, 71 to 95%; specific-ity, 66 to 86%.

� Dobutamine stress echocardiography:− Low to moderate dose of dobutamine:

Sensitivity, 60 to 97%; specificity, 30 to94%.

− High dose of dobutamine: Sensitivity,60 to 87%; specificity, 83 to 90%.

� SPECT: Sensitivity, 72 to 100%; speci-ficity, 38 to 98%.



The latest review, by Bax et al. (published inFebruary 2001) [Bax et al., 2001], which in-volved a comparison of various noninvasivemodalities for detecting hibernating myocar-dium (FDG-PET, thallium SPECT, dobutaminestress echocardiography), showed that FDG-PET had the highest sensitivity (93%, p <0.005) and the highest negative predictivevalue (86%) in relation to the other techniquesand that echocardiography had the lowest sen-sitivity (81%). However, stress echocardiogra-phy had the highest positive predictive value(77%, p < 0.05) compared to 71% for PET, andthe best specificity (80% vs. 58% for PET).

These results should be interpreted with cau-tion, since different study methodologies arefound among the studies that were selected, interms of both the heterogeneity of the partici-pants, the assessment criteria, the protocols andthe analytical methods. Furthermore, most ofthe studies involved patients with an LVEF >35%.

One of the technical problems encounteredwith SPECT scanning is nonhomogeneousphoton attenuation in the chest. Thus, attenua-tion artifacts reduce the specificity of imageanalysis, especially in obese patients andwomen with large breasts. The recent intro-duction of new multihead SPECT camerasmight resolve these problems in the future.

In addition, new technologies or test method-ologies are being developed and evaluated[Bengel et al., 1998]. Hybrid PET/SPECTcameras permit imaging with PET tracers, suchas FDG, using new generations of scanners[Beller et al., 2000]. Also, magnetic resonanceimaging has yielded promising preliminary re-sults in this use [Kim et al., 2000].

4.3.6 Conclusion

From what is stated in the literature (arguments

in favour of clinical benefit based on severalstudies with a low level of evidence and a sin-gle study of good methodological quality but oflow power that did not involve patients who, itwas thought beforehand, would benefit themost from PET), one cannot draw any firmconclusions as to the systematic introduction ofPET in day-to-day clinical practice.

Nonetheless, PET permits an unprecedentedevaluation, both quantitative and qualitative, ofcardiac pathophysiological processes. It is apromising technology that is leading to numer-ous areas of research that will provide a betterknowledge of cardiac processes, which mayhave important clinical applications in patientmanagement.

The use of PET might make it possible to op-timize the management and prognosis of moreseverely affected coronary patients throughbetter selection of those in whom a heart trans-plant is truly indicated (certainty that there isno viable myocardial tissue) and through theuse of revascularization only in those caseswhere myocardial viability has been demon-strated. Certain publications and our simulationconcerning the clinical use of PET to assessmyocardial viability suggest a favourable cost-effectiveness ratio.

In the field of cardiology, the clinical utility ofPET is

recognized for the following uses:� studying myocardial perfusion for the

purpose of diagnosing and managingcoronary artery disease; and

� studying myocardial viability;

and has potential for� monitoring heart transplant patients (de-

tecting posttransplant arteriopathy andmeasuring coronary reserve); and



� monitoring the effect of treatments andthe response to therapy in coronary arterydisease.

4.4 ECONOMIC CONSIDERATIONS

4.4.1 Australian assessment (MSAC)

The observations of the MSAC (Australia) willbe used here as the latest guidelines for as-sessing the cost-effectiveness of PET [MSAC:Ghersi et al., 2000]. A few other referenceswill be included and commented on.

The MSAC notes that models have mainlyconcerned non-small-cell lung cancer andsolitary pulmonary nodules and that there islittle information on the evolution of patientsfollowing PET scanning. Given that there areno direct measurements of PET’s impact onquality of life and survival, the economicevaluations are based on intermediate evalua-tion measures of patients' clinical evolution.Only Gambhir et al. [1996] and Scott et al.[1998] have attempted to evaluate survivalbased on the reduction in the mortality associ-ated with surgical procedures.

Nevertheless, the data tend to suggest that PETcould generate savings [Gambhir et al., 1996;Valk et al., 1996; von Schulthess et al., 1998]

or at least be cost-effective [Scott et al., 1998],if the upper limit of cost-effectiveness is set at$50,000 per life-year saved.

However, the MSAC notes that all the studiesof the diagnosis and management of non-small-cell lung cancer used a simulation methodologyand were therefore based on assumptions. Itconcludes that it would be advisable to vali-date, with clinical trials, the assumptions inthose models, which are nonetheless plausible.Such trials are presently underway and willconfirm or invalidate the results of the models.

From a narrower perspective, given the diver-sity of the data and of the methodologies usedin these studies, it is difficult to make directcomparisons and to extrapolate the results tothe situation in Québec. Furthermore, none ofthese studies explicitly factored in the cost ofimplementing PET in its methods of calcula-tion. Lastly, it is possible that the models useddo not represent a clinical and epidemiologicalapproach specific to the situation in Québec.

4.4.2 1999-2000 update - Economic aspects

4.4.2.1 Literature review - Lung cancer

The economic studies of PET published afterthe MSAC report are summarized in Table 3.

Table 3: Literature review on the cost-effectiveness of PET in lung cancer

Year Author,country

Citation Use Design Results

2000a Dietlein et al.,Germany

Eur J NuclMed 2000;27:1441-56.

Managementof patientswith solitarypulmonarynodules

• Gambhir et al.’smodel [1998] withmodifications

• Decision tree• Four strategies

compared: wait andwatch, exploratorysurgery, transtho-racic needle biopsy

Incremental cost-effectiveness ratio inrelation to the wait-and-watch strategy:• 3,218 euros*• 4,210 euros per

life-year saved forsurgery

• 6,210 euros per



and PET life-year saved forbiopsy

2000b Dietlein et al.,Germany

Eur J NuclMed 2000;27:1591-7.

Managementof patientswith non-small-cellcancer

• Decision tree• Five strategies

compared

Incremental cost-effectiveness ratio inrelation to the conven-tional strategy.• 143 euros per life-

year saved for PET

* 1 euro = 0.7403 Canadian dollars

� Studies concerning solitary pulmonary nodules(SPNs)

The objectives of Dietlein et al.'s cost-effectiveness study [2000a] were to determinethe impact of PET utilization on life expec-tancy and to estimate the incremental cost-effectiveness ratio for implementing PET.

A decision tree was constructed to assess thecosts and efficacy, in terms of the number oflife-years saved, of the four competing strate-gies for managing patients with SPNs: 1) waitand watch; 2) exploratory surgery; 3) transtho-racic needle biopsy; and 4) PET. The parame-ters used in the model came from the literature.Only the following direct costs, which relate tothe public insurer's perspective, were consid-ered by the authors: hospitalization, PET, com-puted tomography, biopsy, surgery, mediasti-noscopy and palliative treatment. Indirect costswere not considered. An annual discount rateof 5% was used, although the time horizon wasnot explicitly indicated.

The study population was a hypothetical cohortof 62-year-old operable men with SPNs of upto 3 cm, with no calcification, spiculae, metas-tases or recent history of extrapulmonary tumordiagnosed by computed tomography or chest x-rays.

A univariate sensitivity analysis was performedfor the following variables in order to test the

robustness of the conclusions: prevalence ofsolitary pulmonary nodules, specificity andsensitivity of PET and biopsy, surgical mortal-ity rate, and costs of PET and palliative treat-ment.

In the baseline scenario, and in comparisonwith the wait-and-watch strategy, the incre-mental cost-effectiveness ratios of PET, biopsyand surgery are, respectively, 3,218, 6,120 and4,210 euros per life-year gained. The resultsare different if the surgery strategy is substi-tuted for the wait-and-watch strategy as thecomparator. Thus, PET has a negative incre-mental ratio of 6,912 euros per life-year gainedcompared to 4,210 euros per life-year gainedfor the wait-and-watch strategy and 3,343 eu-ros per life-year gained for the transthoracicneedle biopsy strategy.

According to the authors, the results suggestthat PET is cost-effective both for patients whoare at risk and those not at risk for dying uponsurgery. They conclude that, in applyingAmerican prevalence data, the budget impactof implementing and disseminating PET in apublic program would be less than one euro perinsured person.

� Studies concerning non-small-cell lung cancer

The objectives of Dietlein et al.’s economicstudy [2000b] were to identify the groups of



patients with non-small-cell cancer most likelyto benefit from PET, to determine the potentialsavings that would result from introducing PETand to determine the role of PET in relation tomediastinoscopy in the arsenal for evaluatingthis disease.

A decision tree was constructed to assess thecosts and efficacy, in terms of life-yearsgained, of five competing strategies: 1) com-puted tomography; 2) PET for patients withnormal-size lymph nodes; 3) PET for all pa-tients; 4) PET for all patients without mediasti-noscopy if the computed tomography and PETscans are positive; and 5) PET for all patientswithout mediastinoscopy if the PET scan ispositive. The parameters used in the modelcame from the literature. Only the following di-rect costs, which relate to the public insurer'sperspective, were considered: hospitalization,PET, computed tomography, biopsy, surgery,mediastinoscopy and palliative treatment. Indi-rect costs were not considered. An annual dis-count rate of 5% was used.

The study population was a hypothetical cohortof 62-year-old operable men in whom the di-agnosis of a tumor was made by computed to-mography and bronchoscopy and confirmedhistologically.

A univariate sensitivity analysis was performedfor the following variables in order to test therobustness of the conclusions: the prevalenceof mediastinal metastases, the specificity andsensitivity of PET, the sensitivity of mediasti-noscopy, and the costs of PET and palliativetreatment.

In the baseline scenario, and in comparison tothe computed tomography strategy, the incre-mental cost-effectiveness ratios for the fol-lowing strategies, 1) PET for patients withnormal-size lymph nodes; 2) PET for all pa-tients; 3) PET for all patients without mediasti-noscopy if the CT and PET scans are positive;

and 4) PET for all patients without mediastino-scopy if the PET scan is positive, are 143,11,100, 24 and 19,830 euros, respectively, perlife-year gained. The results are different if thePET strategy is substituted for the computedtomography strategy in patients with normal-size lymph nodes as the comparator. Thus, theincremental ratios for the following strategies,1) PET for all patients; 2) PET for all patientswithout mediastinoscopy if the CT and PETscans are positive; and 3) PET for all patientswithout mediastinoscopy if the PET findingsare positive, are, respectively, 36,667, 15,325and 15,716 euros per life-year gained.

The results suggest that the scenario in whichPET is used for all patients with normal-sizelymph nodes would be more cost-effective thanthe competing options. The budget impact isacceptable in this scenario.

4.4.2.2 Studies of other uses in oncology andneurology

We did not find any studies evaluating the effi-ciency of PET in other oncological and neuro-logical applications.

4.4.3 Study specific to Québec: non-small-cell lung cancer (NSCLC)

It is difficult to extrapolate the results of thetwo studies by Dietlein et al. [2000a; 2000b] tothe situation in Québec. These authors at-tempted to identify the most efficient strategyfor managing solitary pulmonary nodules(SPNs) and detecting metastases, using clinicalalgorithms reflecting the common practices intheir health-care system.

In Québec, it is unlikely that this technologywill be used by clinicians as a diagnostic toolfor solitary pulmonary nodules because accesswill remain limited (even if this technology be-comes available in Québec's health-care sys-



tem).

Furthermore, given that the prevalence of lungcancer in patients with SPNs varies from 80 to90% when they are referred to a tertiary centre,the use of PET will obviously be limited pri-marily to determining the degree of metastasisin order to assess the appropriateness of un-dertaking surgical treatment.

It was in considering these specific aspects thatan analysis using Québec parameters was per-formed in order to be able to assess the eco-nomic impact and potential efficiency of PETin its implementation context.

4.4.3.1 Materials and method

An analytical model was developed to predictthe cost and effects of using PET to detect me-diastinal and distant metastases. To be able tocompare the results with those of previousstudies, the reference case technique recom-mended by the Washington panel was used inthis analysis [Russell et al., 1996].

The costs and life expectancy were determinedfor each strategy examined. The explicit prob-abilities used in this model are presented in Ta-ble 4. Bayes' theorem was used to extract mostof the probabilities. A health-care system per-spective was adopted in this study. Only the di-rect costs were considered.

4.4.3.2 Construction of the decision tree

The prediction model (Figure 4, Appendix 11)was developed from several assumptions: thestudy population consists of a hypothetical co-hort of 100 65-year-old males with histologi-cally confirmed non-small-cell lung cancer inwhom the preoperative metastatic evaluation,based on conventional detection techniques, isnegative for mediastinal and distant metastases.The model thus excludes any patient with apositive preoperative evaluation. In addition,because of the number and types of scanners

available on the market and of the lack of dataon their efficacy, we assumed, for the sake ofsimplicity, that a dedicated, full-ring, multide-tector PET scanner would be the type deployedin Québec.

The costs associated with the treatment re-quired for the disease, such as chemotherapyand radiation therapy, were not included in ouranalysis. We assumed that a patient in whomsurgical intervention is going to be avoidedthanks to PET incurs the same chemotherapyand radiation therapy costs as a patient who isoperated on and who had, prior to surgery,metastases that were not detected by conven-tional diagnostic methods. The model thus as-sesses the costs associated with the diagnostictests, the surgery and the reimbursement for themedical procedures performed by health pro-fessionals. Consequently, the model mainly ex-amines the difference in surgical costs betweenthe two options.

4.4.3.3 Strategies

In the CT option, patients with positive resultsfor mediastinal metastases undergo mediastino-scopy for the purpose of determining whetheror not they are candidates for surgery. In thosecases, where the CT findings are negative, thepatient undergoes surgery.

As for the CT + PET option, all the patients arefirst evaluated by computed tomography. Theuse of PET is limited to detecting distant me-tastases when the CT scan is positive for medi-astinal metastases and to detecting mediastinalmetastases when the CT scan is negative. In thelatter case, PET is also used for detecting dis-tant metastases. To confirm the PET findings,biopsy and mediastinoscopy are used to detectthe presence of distant metastases and medi-astinal metastases, respectively.



4.4.3.4 Literature review: Data acquisition

The prevalence of mediastinal and contralateralmetastases in patients with NSCLC has beenestimated at 31% (range: 28 to 38%) in the lit-erature [Gross et al., 1988; McLoud et al.,1992; Dillemans et al., 1994; White et al.,1994; Primack et al., 1994]. However, this fig-ure is not the prevalence of NSCLC but ratherthat of N2 or N3 lymph node metastases in pa-tients with a histological diagnosis of NSCLC

who are eligible for surgery. As for the prob-ability of distant metastases detected by PETand not detected by conventional methods, itvaries from 5 to 11% [Valk et al., 1995; Buryet al., 1998; Gupta et al., 1999; Pieterman etal., 2000]. The baseline values and intervals tobe used in the sensitivity analyses and the othervariables used in this analysis are presented inTable 4.

Table 4: List of the variables used in the lung cancer model

Variable Baselinevalue

Lowerlimit

Upperlimit

Reference(s)

Fee* for biopsy (physician) 75 Medical Specialists’ Manual, RAMQ

Fee for computed tomography (physician) 60 Medical Specialists’ Manual, RAMQ, code8262

Fee for mediastinoscopy (physician) 280 Medical Specialists’ Manual, RAMQ, code3036

Fee for PET (physician) 250 Medical Specialists’ Manual, RAMQ, code8700

Fee for surgery (physician) 672 Medical Specialists’ Manual, RAMQ, code3196

Cost* of hospital stay for mediastinoscopy 5,054 4,977 7,932 APR-DRG 076 (subgroup, code 46.81)

Cost of hospital stay for surgery 8,424 7,544 11,781 APR-DRG 075 (no subgroup, code 46.81)

Cost of hospital stay for biopsy 6,130 APR-DRG 076 (subgroup, code 89.00 and90.90)

Cost of hospital stay for mediastinoscopy, biopsy andsurgery

9,163 7,609 12,147 APR-DRG 075 (with subgroups 46.81, 89.00and 90.90)

Cost of PET scan, including capital cost (hospital) 1,313 Appendix 10

Sensitivity of computed tomography 0.75 0.6 0.9 Pieterman et al., 2000Specificity of computed tomography 0.66 0.55 0.77 Pieterman et al., 2000

Life expectancy** with palliative treatment 1 0.1 2 Scott et al., 1998; Gambhir et al., 1996

Life expectancy with surgical treatment 7 1 15 Cummings et al., 1986; Beck et al., 1982

Mortality rate associated with computed tomography 0.000025 0 1 Hartman et al., 1982

Surgical mortality rate 0.03 0 0.2 Williams et al., 1981; Evans, 1973; Ginsberg etal., 1983

Sensitivity of PET in detecting distant metastases 0.82 0.64 1 Pieterman et al., 2000

Specificity of PET in detecting distant metastases 0.93 0.88 0.98 Pieterman et al., 2000

Sensitivity of PET in detecting mediastinal metastases 0.91 0.81 1 Pieterman et al., 2000

Specificity of PET in detecting mediastinal metastases 0.86 0.78 0.94 Pieterman et al., 2000

Probability of metastases detected by PET 0.07 0.05 0.11 Valk et al., 1995; Bury et al., 1998; Gupta et al.,1999, Pieterman et al., 2000

Prevalence of mediastinal metastases 0.31 0.28 0.38 Gross et al., 1998; McLoud et al., 1992; Dille-mans et al., 1994; White et al., 1994; Primack etal., 1994

* in Canadian dollars** in years

Appendix 8 of this report (selected studies) wasconsulted for CT and PET sensitivity andspecificity data. As regards the detection ofmediastinal metastases, the sensitivity of PET

is estimated at between 72 and 96%, with 90 to100% specificity. As for its sensitivity andspecificity in detecting distant metastases,Pieterman et al. [2000] estimate them at 82%



(range: 64 to 100%) and 93% (range: 88 to98%), respectively. For analytical purposes,Pieterman et al.'s study was used as the mainsource of data for PET and CT sensitivity andspecificity values, since it is the only prospec-tive study that estimates the performance ofPET and computed tomography in detectingmediastinal and distant metastases. Lastly, weassumed that the accuracy of biopsy as a diag-nostic tool is 100%.

The mortality rate associated with the surgicalresection of lung cancer reported in the litera-ture varies from 2.4% [Williams, 1981] and20% [Evans, 1973]. Ginsberg et al. [1983] re-port an overall mortality rate of 3.7% and athoracotomy-specific mortality rate of 2.9%.We chose 3% (range: 0 to 20 %) as the baselinevalue in our analysis.

The risk of mortality associated with computedtomography is mainly due to the contrastagents used. A risk of 1 in 40,000 (0.0025%) isreported in the literature [Hartman et al., 1982].The risk associated with FDG-PET is assumedto be negligible, since, to date, no complica-tions or adverse effects have been reported inthe literature.

Life expectancy was calculated according tothe DEALE method, which was developed byBeck et al. [1982], where mean life expectancy= 1/(ASR + DSR). The ASR is the age-, sex-and race-adjusted annual mortality rate in thegeneral population. DSR is the disease-relatedsurvival rate. In Québec, the life expectancy ofa 65-year-old man is 15.64 years [Institut de lastatistique du Québec, 2001]. The ASR is thus0.067 (1/15.64). The DSR for a 2.3-cm lungtumor is 0.075 [Cummings et al., 1986]. Thecombined mortality rate for a typical patient is0.142. His life expectancy (the reciprocal func-tion of this sum) is 7 years. This figure wasused in the baseline analysis, with a range of 1to 15 years for the sensitivity analysis.

The life expectancy for inoperable cancers inpatients with an advanced stage of the disease,based on radiographic findings, is estimated at0.47 years [Gambhir et al., 1996; Scott et al.,1998]. Although there are certainly a few sur-vival data for patients with mediastinal metas-tases that are not detected radiographically,these patients probably exhibit better survivalthan those in whom the chest x-ray is une-quivocal. To take this into account, we as-sumed a life expectancy of 1 year (range: 0.1 to2 years) as the baseline value for inoperablepatients. As for the life expectancy of patientswhose distant metastases are not detected byPET and for patients whose mediastinal me-tastases are not detected by computed tomo-graphy, it was estimated at 1 year.

The cost of a PET scan was estimated from thedata presented in Appendix 10 of this report.We calculated that a PET scan costs $581, ex-cluding the capital cost. The cost of a PET scanis $1,313, which includes the capital cost, or$732 more per scan. This figure was estimatedon the basis of the following assumptions. Weassumed a cost of $3.2 million for a scannerand of $4.3 million for a cyclotron. As for con-struction costs, we assumed that they were$500,000 for the cyclotron and $175,000 forthe scanner. The life span of the scanner andthat of the cyclotron were set at 10 and 25years, respectively. Lastly, we used a 5% inter-est rate to take into account the loss of returnon the capital, had it been invested.

The cost of the hospital stay for surgery alone,mediastinoscopy alone, biopsy alone and for allfour interventions was determined by consult-ing the Ministère de la Santé et des Servicessociaux's 1998-1999 APR-DRG database. Thecost is $8,424, $5,054, $6,130 and $9,163, re-spectively. These average figures include thehospital stay and all the interventions and man-agement that the hospital is to provide for thepatient during the stay. Only the physicians'professional fees are excluded. For example,



the costs associated with a CT scan are in-cluded in one of the APR-DRGs, excluding thephysicians' fees.

The figures for the physicians' fees are fromthe Medical Specialists' Manual (September1999 edition), published by the Régie de l'as-surance-maladie du Québec, and are providedin Table 4 for each intervention. Lastly, thecost of a biopsy depends on the type. It canvary considerably. We therefore assumed anaverage cost of $75.

4.4.3.5 Analysis

The cost and life expectancy estimates for eachoption were obtained by adding the products ofthe probabilities and cost and life expectancyvalues. The mean cost-effectiveness ratios wereused to evaluate the CT scan strategy in rela-tion to the CT + PET strategy. The lowest ratiois that of the most efficient strategy. The in-cremental cost-effectiveness ratio (cost of theCT option - cost of the PET option)/(CT lifeexpectancy - PET life expectancy) was used forthe marginal analysis of PET efficiency.

Given the low accuracy of certain variables, aunivariate sensitivity analysis was performed.

This type of analysis consists in varying thevalues of a given variable in a predefined rangewhile maintaining the values of the other vari-ables constant, and assumes independence be-tween the variables.

To take into account the interdependence be-tween the variables, a Monte Carlo dynamicsensitivity analysis was performed in order toobtain a 95% confidence interval for the costs,efficacy and cost-effectiveness ratio. This typeof simulation executes the model numeroustimes (1,000 in our analysis), modifying thevalues of all the variables simultaneously, us-ing a predefined probability distribution. Thisapproach provides a distribution of samplesand therefore descriptive values (mean, me-dian, maximum, minimum and probability dis-tribution) for the result. The variables, theirranges and their predefined distributions areshown in Table 48 (Appendix 11).

Given that the time horizon for the analysiswas one year, no discounting of the costs or ef-fects was necessary.

Table 5: Results for each strategy per patient (lung cancer)

StrategyCost

IncrementalCost

Efficacy(life-years)

Incremental efficacy

Mean cost-effectiveness

ratio

Incremental cost-effectiveness ratio

CT $8,455 4,551 $1,858 per life-year

CT + PET $9,723 $1,268 4,823 0.27 $2,017 per life-year

$4,689 per addi-tional life-year

gained

4.4.3.6 Results

Table 5 shows the results per patient and thecosts, efficacy, mean cost-effectiveness ratiosand the incremental cost-effectiveness ratio.

In general, the mean cost-effectiveness ratiossuggest that PET is almost as efficient an inter-vention as computed tomography, the meancost-effectiveness ratio for PET being slightly



higher than that for computed tomography. Themean cost of the CT strategy is $8,455 per pa-tient compared to $9,723 for the PET strategy,for a cost differential of $1,268. The PET strat-egy extends life expectancy by slightly morethan three months (0.27 years) compared tosurvival with the CT strategy.

As for the incremental cost-effectiveness ratio,i.e., the cost associated with one additional life-year gained, it is $4,689. Considering the num-ber of new cases and the prevalence of medi-astinal metastases, 1,837 new patients a yearcould potentially benefit from this technology.The budget impact, from the perspective of theincremental cost, i.e., the additional cost to thehealth-care system, would be $8,613,693. If weperform the same analysis but exclude thecapital cost from the cost of PET scanning, theincremental cost-effectiveness ratio would be$1,983, with a budget impact of $3,642,771.

Table 49 (Appendix 11) shows the results ofthe univariate sensitivity analysis for all thevariables included in the model. It can be seenfrom the results that the variation in the base-line values of the variables does not have animpact on the decision, the mean cost-effectiveness ratio for the PET strategy alwaysbeing higher than that for the competing op-tion.

Figure 6 (Appendix 11) is a graph of the MonteCarlo simulation for the incremental cost-effectiveness ratio. It can be seen that most ofthe simulations are in the right quadrant, thequadrant where PET improved life expectancywhile at the same time requiring an investmentfor each additional health gain. Table 50 (Ap-pendix 11) shows the mean, median and quar-tiles for the incremental cost, efficacy and cost-effectiveness ratio values, as they appear in thedynamic sensitivity analysis.

Table 51 (Appendix 11) shows the frequenciesof occurrence of incremental cost-effectiveness

ratios, by intervals, of the 1,000 Monte Carlosimulations. It can be seen that in 95% of thecases, the incremental ratio is less than $50,000per life-year gained.

4.4.3.7 Discussion and conclusions

Our study suggests that PET is a cost-effectiveintervention in the specific context of its clini-cal utilization and deployment in Québec.Thanks to the reference case technique, wewere able to compare our results with thosepresented in the literature. We note that theyare similar to the results compiled by theMSAC, that is, that PET can lead to savings orrequire only a very small and very acceptableinvestment for each life-year gained, providedthe purchase cost, which is high, has been am-ortized.

Our univariate and Monte Carlo sensitivityanalyses confirmed the robustness of the re-sults. Table 49 (Appendix 11) shows thatvarying the baseline values does not signifi-cantly affect the results, the incremental cost-effectiveness ratio ranging from $3,000 to$5,000 per life-year gained.

Table 51 (Appendix 11) shows the distributionof the incremental cost-effectiveness ratios. Itcan be seen that, in about 50% of the simula-tions, the investment required to gain one life-year is less than $5,000. In 73% of the cases, itis less than $10,000, which is not, in itself, avery large investment. If the standard cost-effectiveness ratio is set at less than $50,000per life-year gained, 95% of the simulationcases would be below this figure. Thus, eco-nomically, PET would not necessarily generateany savings in the health-care system, but itmight avoid unnecessary surgeries and there-fore, to a certain extent, shorten the waitinglists and contribute to the overall efficiency ofthe health-care system.

Using the available data, we attempted to as-



sess the economic impact and efficiency ofPET on clinical outcomes and the health-caresystem. Since there is no direct measure of pa-tient survival, we used an intermediate end-point, such as the reduction in the mortality dueto surgical procedures. However, using such anindicator does not enable us to fully assess theutility of PET. In the short term, it is unlikelythat PET will improve survival in patients withlung cancer. The efficacy of PET resides in-stead in its ability to improve the patients’quality of life by sparing them an unnecessary,debilitating intervention and providing themwith quicker access to treatment, on the onehand, and to enable them to express their pref-erences in terms of the clinical approach to beused, on the other. Quality-of-life and patientpreference measurements obtained by prospec-tive studies will permit a more adequate as-sessment of the efficacy of this technology.

We included the costs required to implementPET in Québec. For analytical purposes, weconsidered a very cautious scenario and as-sumed that this technology would be used onlyto detect local and distant metastases in pa-tients with NSCLC. Given that this technologywould be used for clinical purposes other thandetecting metastases in patients with NSCLC,the capital cost should not be done on the basisof this single application. Thus, the capital costto be included in the cost of PET scanningwould probably be less than $732 per use,which would significantly reduce the incre-mental cost-effectiveness.

Because the analysis is limited to a single clini-cal use (the detection of metastases in cases ofnon-small-cell lung cancer) and to a very lim-ited population, other studies are needed toconfirm PET's efficiency or lack thereof inother applications. Our prediction model esti-mates the reduction in the number of surgeriesat 12%, with patient survival prolonged bythree months. These figures are similar to theresults of a randomized, prospective study [van

Tinteren et al., 2000] presented at the 2000ASCO conference.

In conclusion, with the current state of knowl-edge and from an economic perspective spe-cific to the situation in Québec, our modelshows that the use of PET to detect local anddistant metastases in NSCLC is potentially anintervention that would require an acceptableinvestment for each life-year gained.

4.4.4 Study of myocardial viability in theQuébec context: An exploratory analy-sis of the potential impact

In the very near future in Québec, it is likelythat PET will be used in cardiology mainly fordetecting viable myocardium that can respondfavourably to revascularization.

At this time, only a few efficacy studies areavailable for estimating the potential impact ofPET as a diagnostic tool in this application.Furthermore, no economic evaluation of thisclinical application is, to our knowledge, avail-able. This lack of data makes it difficult to as-sess the anticipated efficacy of PET in this dis-ease.

The literature abounds with differing opinionsabout the usefulness of models in this specificproblem. Some authors do not see this ap-proach as having any validity, while othersmaintain that early modelling is an importantmeans of aiding the decision-making process.They recognize the need to use different ana-lytical approaches to make up for the lack ofexperimental data. They also recognize thatutilizing the results of an early model to deter-mine research priorities does have its useful-ness.

This is where our analysis comes in. Our ob-jective was to model the use of PET to detect



myocardial viability and to assess its potentialimpact. The results of this modelling will en-able us to make a global assessment of PET'sutility at this stage of its development. To thisend, a Monte Carlo-type mathematical modelwas used for the analysis. This approach yieldsresults in the form of intervals with a prede-fined level of confidence. The objective of thisanalysis is not to provide accurate data, butrather to estimate, in an exploratory manner,the anticipated economic impact of PET in thisspecific clinical application, i.e., evaluatingmyocardial viability.

4.4.4.1 Materials and method

Given the context in which PET will be used inQuébec and the availability of data in the lit-erature, we adopted, for the purposes of thisanalysis, a data extraction approach by sur-veying expert clinicians. Expert clinicians atHôpital Laval in Québec City were consultedfor the purpose of developing decision trees,determining the explicit probabilities for thevariables and determining the population forwhich the technology would be used first. Alist of the experts who were consulted is pro-vided in Appendix 12.

The decision tree that was constructed is shownin Appendix 11 (Figure 8). For patients withsevere left ventricular dysfunction who havenot previously undergone coronary angiogra-phy, which is the case for about half of the pa-tients in tertiary cardiology, when the clinicianneeds to make a diagnostic and therapeutic as-sessment, the use of PET on a first-recoursebasis may be an attractive option.

In current clinical practice, where the use ofPET is not possible, the clinician, in about onehalf of the cases, will first perform coronaryangiography. If he or she detects coronary ar-tery disease, he or she will, in 75% of these pa-tients, then have to determine their myocardialviability. The use of conventional modalities,

i.e., SPECT or stress echocardiography (dobu-tamine) will be inconclusive or "negative" in50 to 75% of these cases, which will lead theclinician to make a relatively arbitrary decisionregarding the possibility of revascularization(or to opt for transplantation or pharmacologictreatment).

For each strategy examined, the costs and theproportion of individuals surviving at fiveyears were estimated. The explicit probabilitiesused in the prediction model are based on thevariables presented in Table 6. A health-caresystem perspective was adopted in this study.Only the direct costs were considered. Thecosts associated with the use of medical serv-ices and the reimbursement of professional feeswere estimated for each strategy.

The cost of a PET scan was estimated using themethod described in Section 4.4.3.4. The costof a thallium scan was estimated from financialreport AS-471 and expert opinions. It is $350.This figure includes the cost of the procedureand that of the radiotracer. The cost of revas-cularization was obtained by consulting the1998-1999 APR-DRG database. The costs ofmedical treatment and transplantation wereobtained from expert opinions. They are$20,000 and $60,000, respectively. This latterfigure does not include the cost of obtaining anorgan or of the coordination involved in a hearttransplant, given that the time horizon in ouranalysis precedes transplantation. The figuresused for the physicians' fees are from theMedical Specialists' Manual (September 1999),publish by the RAMQ. For the purposes of ouranalysis, a variation of ± 20% was applied toeach cost component in order to obtain an in-terval. These costs are shown in Table 6.

For measuring efficacy, we used the patients'mean probability of survival at five years afterrevascularization, medical treatment and/ortransplantation.



Lastly, the study population consists of a hy-pothetical cohort of male patients with a left

ventricular ejection fraction of less than 30%.

Table 6: List of the variables used in the myocardial viability model

Description Value Source

Cost* of a PET scan (hospital) 1,050 - 1,575 Appendix 10Fee* for PET (physician) 225 - 275 Medical Specialists’

Manual (RAMQ)Fee for a thallium scan (physician) 94.4 Medical Specialists’

Manual (RAMQ)Cost of revascularization 8,262 - 10,099 APR-DRGCost of a thallium scan (hospital) 315 - 385 Financial report AS-

471Cost of medical treatment 16,000 - 24,000 Expert opinionCost of transplantation 48,000 - 72,000 Expert opinion5-year postrevascularization survival probability 0.8 Expert opinion5-year post-medical treatment survival probability 0.5 Expert opinion5-year posttransplantation survival probability 0.75 Expert opinionProbability of medical treatment 0.6 - 0.95 Expert opinionProbability of an unequivocal thallium scan 0.3 - 0.4 Expert opinionProbability of viable myocardium when the thallium scan isequivocal in the thallium-alone option

0.15 - 0.3 Expert opinion

Probability of viable myocardium when the thallium scan isequivocal in the thallium + PET option

0.5 Beanlands et al. 1997,Dreyfus et al. 1994

* in Canadian dollars

4.4.4.2 Analysis

The mean and incremental cost and efficacyintervals were used to compare the PET optionwith the no-PET option. These intervals weregenerated by a Monte Carlo analysis. This typeof simulation executes the model numeroustimes (1,000 times in our analysis, with a 95%confidence level), modifying simultaneouslythe data and randomly choosing values from apredefined probability distribution. The vari-

ables and their predefined distribution areshown in Table 52 (Appendix 11).

4.4.4.3 Results

Table 7 shows the results and the 95% confi-dence intervals for the costs, the 5-year sur-vival probability, the incremental cost and theincremental efficacy.

Table 7: Results of the economic analysis for myocardial viability

Strategy Cost ($) Efficacy Incremental cost ($)

Incrementalefficacy

Thallium + clinicaldecision

10,547 to 29,993 0.63 to 0.71

Thallium + PET 10,119 to 24,753 0.69 to 0.73 -7,182 to 687 0.02 to 0.07



The thallium + PET strategy seems to be themost cost-effective option. It appears to be lessexpensive, and its efficacy is superior to that ofthe thallium-alone strategy. It seems that PET,based on the basic assumptions, would yieldsavings for the health-care system (the incre-mental cost is negative).

Figure 7 (Appendix 11) is a graph of the in-cremental cost-effectiveness ratio simulationresults. Each point represents a simulation. Itcan be seen that almost 100% of the points arein the left quadrant, which means that the thal-lium + PET strategy would yield savings and atthe same time be more effective, with a 95%confidence level.

As for efficiency, i.e., its cost-effectiveness,our model, which is based on several assump-tions (efficacy, target population and costs),suggests that PET is a very cost-effective inter-vention.

4.4.4.4 Discussion and conclusions

Our objective was to estimate the efficiency ofPET in detecting myocardial viability. Ourpreliminary analysis, based on a Monte Carlo-type dynamic simulation, suggests that PET isa very cost-effective intervention in the contextof tertiary cardiology centres when used for aspecific purpose in a population of patientswith an ejection fraction of less than 30%.

However, this finding needs to be qualified.Since there are no available data in the litera-ture, our sources of data were mainly expertopinions. Despite the inherent weakness of thisapproach, its use in certain types of economicstudies should not be avoided. Data from ran-domized clinical studies and formal literaturereviews certainly provide more-valid data, butwhen there is a paucity of data from suchsources, one must resort to other methods, es-pecially in a context of diffusing new tech-

nologies, such as PET in cardiology.

Given that the time horizon for the analysis incardiology was several years, the costs andconsequences should have been discounted.However, for the sake of simplicity, this wasnot done. It is unlikely that discounting wouldhave reversed the result, that is, that PET isvery cost-effective. To compensate for the un-certainty concerning the accuracy of the data,we used intervals rather than estimates and aMonte Carlo-type dynamic simulation. Usingthis strategy enabled us not to overestimate orunderestimate the efficacy and cost data. As forthe variables that were kept constant, the sensi-tivity analysis performed for these variables re-veals that varying the baseline values does notaffect the results.

Although this is early modelling, if we acceptits limitations, it suggests that PET might bemore efficient, because the analysis is limitedto assumptions and to a very small population,and that further studies are necessary, espe-cially in the case of using PET as a diagnostictool for coronary artery disease and myocardialviability. Perfusion studies are performed todetect the presence of coronary artery disease,since in the vast majority of noncoronary car-diomyopathies, there should be little or no al-teration of the regional myocardial flow orcoronary reserve. Thus, PET could potentiallylead to savings in coronary angiography utili-zation. In addition, in the case of coronary ar-tery disease, a combined myocardial perfusionand myocardial viability PET study might en-able one to determine if there is any benefit inrevascularizing and, if the absence of viabilityis not demonstrated, to avoid performing coro-nary angiography and any other subsequent di-agnostic test.

In conclusion, given the current state of knowl-edge and with an economic perspective specificto the situation in Québec, we note that the use



of PET for detecting myocardial viabilityseems to be an efficient intervention. However,it would be important to document evidence ofthe incremental efficacy of PET in relation tothe diagnostic tools that are currently available.

4.4.5 Studies concerning other uses in oncol-ogy and neurology

No study on the other uses of PET was carriedout. The efficiency of PET in these clinical ap-plications, specifically for Québec, is unknown.

4.5 GRADING CLINICAL USES

Referring to the conclusions stated above inthis section for each of the uses and rearrangingthem, not by discipline and by use, but ratheraccording to the recognition of the clinical usein each discipline, we obtain the followinggrouping:

� Clinical uses said to be recognized if the harddata are acceptable (in terms of clinical effi-cacy);

� Potential, if the hard data are acceptable but in-complete; and

� Not recognized, if the data are insufficient toreach a verdict, if they demonstrate nonperform-ance or if hard data were not found during theliterature surveys performed for the purposes ofthis report.

It will be noted that this classification reflectsthe current hard data rather than a limitedopinion about the use of this technology.

4.5.1 Recognized clinical uses

4.5.1.1 Oncology

� Lung cancer� Characterizing solitary pulmonary nodules.� Initial staging when a diagnosis of non-small-

cell lung cancer is made, including:� detecting mediastinal lymph node metasta-

ses; and� detecting distant metastases.

� Colorectal cancer� The preoperative detection of hepatic and ex-

trahepatic metastases in the context of detect-ing localized recurrence.

� Determining the location of recurrence in thepresence of clinical symptoms or abnormalparaclinical findings (conventional imaging,carcinoembryonic antigen, etc.).

� Differentiating between residual tumor andpostoperative scar when diagnostic imagingreveals abnormalities.

� Melanoma� Detecting extranodal metastases during the

initial workup or the postoperative follow-up.� Evaluating a potentially treatable recurrence.

� Head and neck cancer� Identifying an unknown primary tumor in the

presence of cervical lymph node metastases.� Staging cervical lymph nodes when conven-

tional imaging is negative.� Detecting recurrence or residual tumors and

differentiating between a tumor and postop-erative scar.

� Lymphoma� Staging when restaging could influence the

choice of treatment.� Evaluating residual disease after treatment.

4.5.1.2 Neurology and psychiatry

� Refractory epilepsy� Localizing epileptogenic foci in patients with

refractory epilepsy who are candidates for sur-gery and in whom the information on the loca-tion of the foci (usual work-up, including sei-zure semiology, EEG and magnetic resonanceimaging) is inconclusive.



� Brain tumors (mainly glioma)� Evaluating residual lesions after treating recur-

rent glioma and differentiating between radi-onecrosis and recurrence in patients treatedwith radiation therapy with abnormalities ondiagnostic imaging.

4.5.1.3 Cardiology

� Myocardial perfusion studies� Diagnosis and management of

coronary artery disease.

� Myocardial viability studies

4.5.2 Potential clinical uses

4.5.2.1 Oncology

� Lung cancer� Monitoring response to therapy.� Detecting recurrence or residual tumors.

� Colorectal cancer� Monitoring response to therapy.

� Hodgkin's and non-Hodgkin's lymphoma� Evaluating response to therapy during treat-

ment.

� Breast cancer� Staging of primary and recurrent tumors.� Detecting axillary and internal mammary

lymph node metastases.� Detecting the primary tumor in the context of

an equivocal complete evaluation.� Monitoring response to therapy.

Mention is made in the literature of avenues ofclinical utility in particular circumstances forthe following cancers, which were not exam-ined in this report: gynecologic cancers (ovar-ian, uterine, cervical), certain genitourinarycancers (testicular, metastases from hyperneph-romas), mesotheliomas, soft-tissue sarcomas,esophageal cancer, and pancreatic cancer.

4.5.2.1 Neurology

� Brain tumors (mainly glioma)� Initial staging of patients suspected of having a

primary brain tumor, in order to guide the bi-opsy to the highest area of activity.

� Evaluating the progression of a low-gradeglioma to malignancy.

� Preoperative workup.� Histologic grading of tumors.� The choice of treatment.� Determining the prognosis.� Detecting metastases.

4.5.2.2 Cardiology

� Monitoring heart transplant patients (detectingposttransplant arteriopathy and measuringcoronary reserve).

� Monitoring the effect of treatments and re-sponse to therapy in coronary artery disease.

4.5.3 Unrecognized clinical uses

4.5.3.1 Oncology

� Colorectal cancer� Diagnosis of the primary lesion.

� Melanoma� Detecting lymph node metastases.

� Head and neck cancer� Monitoring response to therapy.

� Prostate cancer

4.5.3.2 Neurology and psychiatry

� Alzheimer's disease� The clinical utility of PET is not recognized,

since there is presently no treatment, but itcould play a role in the differential diagnosisof dementia and other cognitive diseases, suchas Parkinson's disease.

It will be noted that this classification reflectsthe conclusions of recent assessment reportsavailable in 2000 and available hard data gath-ered up to February 2001. As was seen, for ex-ample, with the clinical use of PET in lym-



phoma, it sometimes only takes a few weeksfor new publications to corroborate the recog-nition of these uses or to move certain usesfrom the "potential" category to the "recog-nized" category.

It will also be noted that, although this classifi-cation will change quickly over the next fewmonths because of clinical data, the efficiencydata on PET still need to be strengthen, as indi-cated in the studies (discussed above) con-cerning the economic aspects of this technol-ogy.


Discussion67

5. DISCUSSION

5.1. ASSESSMENT OBJECTIVES AND STRATEGY

The objectives of AÉTMIS’s assessment are:

1. To gather hard data on the clinical useof PET in different fields, in particular,oncology, neurology and cardiology;and

2. To make recommendations concerningthe possible deployment of PET inQuébec.

When these objectives were set out, PET's effi-cacy in a certain number of clinical applica-tions was taken for granted. Once constructed,the list of these applications, together with eco-nomic estimates, would serve as a basis forrecommendations concerning the deploymentof this technology in Québec.

Ideally, the demonstration of the clinical effi-cacy of a given diagnostic test should be basedon the following conventional criteria: 1) itsperformance in terms of sensitivity, specificityand the other usual parameters (positive andnegative predictive values, etc.); 2) the unique-ness of the information obtained in relation toother tests; 3) the improvement in patient man-agement in terms of beneficial outcomes; and4) acceptable cost-effectiveness, given the per-spective adopted.

In light of the data gathered thus far, the rangeof responses seems vast. For current and po-tential users, the clinical utility of PET hasbeen demonstrated for a very large number ofapplications, and this technology should be de-ployed immediately in Canada and Québec.

For reimbursement organizations (e.g., theDHAC-DTB in Australia and the HCFA in theUnited States) faced with a technology that isalready deployed in their countries, after sev-

eral years if not decades of research, the con-clusions concerning the recognition of PET’sclinical efficacy are still guarded.

An evaluation of the existing data suggests thatthe demonstration of the efficacy of PET whenused for clinical purposes is only partly sup-ported. The DHAC-DTB recognizes severalclinical applications, but they are being cov-ered on a temporary basis until the existingdata are completed with performance valida-tions and economic studies. Furthermore, ex-plicit conditions concerning the qualificationsof clinical staff and the implementation ofquality assurance protocols must be met in or-der to obtain reimbursement. Additionally,while the HCFA recognizes several reimburs-able applications, coverage of these uses issubject to various conditions.

This difference between the current and poten-tial users and the coverage policymakers showsthe extent to which the modalities used by thedifferent players concerned for evaluatingPET's efficacy differ. For current and potentialusers, the demonstration of PET's efficacy is, ingeneral, based on expert opinions based, inturn, on publications selected according to un-stated criteria.

For organizations and task forces whose man-date has been to study coverage requests forPET scans (e.g., the VA-TAP, MSAC, DHAC-DTB, BCBSA and HCFA), the criteria for as-sessing the studies submitted in support of therequests are explicit, verifiable and supple-mented with consultations with expert groups.These still-recent, various attempts by reim-bursement organizations and assessment agen-cies to apply an appropriate evaluative processto PET reflect the difficulties inherent in as-


Discussion68

sessing diagnostic tests or examinations [Fry-back et al., 1991]. In this regard, mention willbe made of the difficulties experienced by theDepartment of Veterans Affairs in 1996 and in1998, despite efforts to establish appropriateassessment criteria [Adams et al., 1997].

AÉTMIS opted for a methodology equivalent tothat in previous reports, modifying the assess-ment criteria as follows: in addition to selectioncriteria identical to those of the VA-TAP andMSAC, the methodological assessment criteriawere the same as those used by these two orga-nizations and were supplemented with the ex-plicit requirement that a study compare PETwith another technology in order to make it ahigher-grade study (see Table 19 in Appendix6).

It will be recalled that the strategy chosen byAÉTMIS consisted in accepting the conclusionsof reports based on a structured assessmentmethodology, then in carrying out an update onthe publications that postdate these reports,from 1999 to February 2001. In these circum-stances, the quality of the studies cited in theprevious reports was not reassessed. Occasion-ally, articles prior to 1999 were consulted orevaluated if they had not been included in theprevious reports. This approach was identicalfor each of the areas of application examined,i.e., oncology, neurology and cardiology.

While the assessment of the methodologicalquality of studies is the same for the differentareas of application, the measurement of theultimate effects of PET scanning may differaccording to the type of patients. Two exam-ples in oncology and one in cardiology will il-lustrate these differences. Because of the poorsurvival prognosis in non-small-cell lung can-cer, the clinical efficacy of PET may reside inthe avoidance of unnecessary surgical inter-ventions, which would lead to an increase insurvival by reducing the complications of these

interventions. The judicious use of PET inlymphoma could translate into increased sur-vival, although there is still insufficient data inthis regard. In cardiology, a change of clinicaldecision can have an immediate and measur-able impact on patient health. Here, too, harddata are still not available, although somestudies in progress might provide informationin the near future.

The economic models constructed by AÉTMISwere aimed at compensating for this lack ofdata.

5.2 ECONOMIC CONSIDERATIONS

The observations of the MSAC (Australia)were used, to a large extent, as the latestguidelines for assessing the cost-effectivenessof PET. Although, according to the MSAC, thehard data are still not sufficient to draw anyfirm conclusions as to the efficacy and cost-effectiveness of PET, recent data and our mod-els tend to suggest that PET may be cost-effective for the two cases in question, i.e., thestaging of non-small-cell lung cancer and de-tecting myocardial viability. The models evensuggest that, for these applications, PET couldgenerate savings or at least be cost-effective ifthe upper limit of cost-effectiveness accept-ability is set at $50,000 per life-year saved.

However, all of the studies use a simulationmethodology. They are thus based only on as-sumptions, not on validated data. In addition,the MSAC concludes that it would be advisableto validate, with clinical trials, the assumptionsin these models, which are nonetheless plausi-ble. Such trials are presently underway andmay enable one to confirm or invalidate the re-sults of these models.

From a narrower perspective, given the diver-sity of the data and methodologies in these


Discussion69

studies, it is difficult to make direct compari-sons and to extrapolate the results to the situa-tion in Québec, for the models used may notrepresent a clinical, epidemiological and eco-nomic approach specific to the situation in thisprovince.

Assuming limited diffusion of PET at the out-set and in light of the epidemiological and eco-nomic data specific to the situation in Québec,two models were constructed, one for non-small-cell lung cancer (Section 4.4.3), the otherfor myocardial viability (Section 4.4.4). Theresults of these studies suggest that PET is acost-effective intervention for staging non-small-cell lung cancer. As for detecting myo-cardial viability, based on the available pre-liminary data and given the very nature of thepotential impact of PET on patient manage-ment, PET is potentially very cost-effective.

As regards uses other than those mentionedabove, PET's efficiency remains an unknown inthe Québec context, as in most other health-care systems.

To illustrate the rapid pace at which economicdata are changing, mention will be made of anarticle published after our survey cut-off date.It concerns a model applied to the use of PETin colorectal cancer (Park et al., 2001) and in-dicates that adding PET to computed tomogra-phy is cost-effective for staging.

5.3 LIMITATIONS OF THE ASSESSMENT

As in a good number of industrialized coun-tries, although the situation has still not beengeneralized on a worldwide scale (see Appen-dix 3 for the geographic distribution of PET),Québec's health-care system is witnessing a faitaccompli: a particular research technology hasgradually developed over the past 25 years orso. During the last decade, manufacturers have

focused on its use for clinical purposes, placingincreasingly sophisticated equipment on themarket. For about the past five years, health-care systems have been facing stronglymounting pressure to cover PET scans, yetthere is no real evidence of this technology'sefficiency.

This situation is a close repeat of the assess-ment of computed tomography. Over 20 yearsago, the first formal assessment of this technol-ogy encountered the methodological difficul-ties evaluating diagnostic tests and the absenceof cost-effectiveness data. It will be noted,however, that this aspect was barely beginningto be taken into consideration by health-caresystem policymakers [Banta and McNeil,1978].

This situation is also similar to that of the as-sessment of magnetic resonance imaging(MRI) some ten years later, where the efficacyand efficiency evidence was also scarce whenthis technology was first used for clinical pur-poses. The parallel between MRI and PET wasnoted by the VA-TAP’s evaluators [Adams,1997]. They cite the first exhaustive study ofthe clinical efficacy of MRI [Cooper et al.,1988] and the research leading to the observa-tion that the more explicit the assessment crite-ria, the less favourable the conclusions regard-ing the clinical use of MRI [Kent and Larsen,1988].

To compensate for the lack of hard data andnot delay the development of this new yetpromising technology, it was, at the time, sug-gested [Chalmers, 1988] that patients be en-rolled in approved studies as a means for athird-party payer to subsidize the first studies.This would reduce the risk of paying for teststhe demonstration of whose efficacy was stillonly in the preliminary stage. This suggestionhas been echoed in the case of PET.

It would still be an appropriate suggestion,


Discussion70

since, with new promising technologies, thereis once again the question as to whether a still-preliminary demonstration of the clinical effi-cacy of a given technology, as is the case withPET, is enough to justify coverage by a publicbody without it first requiring efficiency dataon the new technology.

The coverage conditions imposed in Australiaand even in the United States seem, to a largeextent, to be aimed at checking the first-recourse use of PET and in making up for thelack of data on a large number of patients bycompiling such data as reimbursements aremade. The idea, it seems, is a bit as if adminis-trative data could, at the end of the day, be usedin place of data acquired from comparativestudies.

In other words, by imposing restrictions on theuse of PET, these health-care systems seem tobe approving the awarding of grants on a one-by-one basis, since there are no complete dos-siers on the clinical efficacy of this new tech-nology to rely on. Furthermore, Australia'sDHAC-DTB (a ministerial body), which en-dorsed the recommendations made by theMSAC (assessment organization) expects datato be gathered that will document the effi-ciency of the different uses of PET, in additionto formal research that is currently underway inthis area. The Australian and American ap-proaches could help guide the clinical use ofPET until the anticipated additional hard databecome available.

However, the means by which these additional

data are acquired will pose numerous chal-lenges both to the users, evaluators and poli-cymakers, for technological improvements areoccurring at such a rapid pace that studies setup for the purpose of generating the hard datathat are lacking at the present time could beoutdated before they are completed.

A similar comment could just as well be madeabout the technologies that "compete" withPET, such as computed tomography and mag-netic resonance imaging or other nuclear medi-cine modalities that continue to advance. In-deed, it is expected that technological advanceswill lead to hybrid technologies combining anincreased capacity of computed tomographyfor anatomical localization with an increasinglyrefined measurement of the metabolic activityin the observed lesions. The objectives of thisreport did not include this aspect of the evolu-tion of related technologies.

From a broader perspective, it should also bementioned that government agencies do notperform formal, rigorous clinical and economicassessments of most of the commonly used di-agnostic tests. This still-recent assessmentprocess has become more rigorous over thepast few years, and this change has created, inpotential users, the impression that a "newstandard" for assessing diagnostic tests hasbeen established for PET.

This advance in the assessment and acceptancecriteria for a new technology illustrates, withthe example of PET, the now-generalized needto tightly manage limited resources.


Suggestions for deploying PET in Québec71

6. SUGGESTIONS FOR DEPLOYING PET IN QUÉBEC

6.1 CLINICAL NEEDS

In all the areas of PET utilization, the list ofapplications whose efficacy is recognized isgetting longer by the day. As a result, the clini-cal uses enumerated in Chapter 4 do not con-stitute a closed list. The enumeration of recog-nized and potential uses raises the issue of thenumber of patients who could benefit from thistechnology.

However, for now, the number is very hypo-thetical, mainly because of the diversity of theapplications. A compilation of the incidencerates of the diseases for each recognized appli-cation would help approximate the extent ofthe needs. Incidence can, of course, serve as astarting point, and the general epidemiologicaldata found in Chapter 4 would enable us to dothis. However, the number of patients whocould benefit would have to be estimated, usingmanagement algorithms appropriate for eachtype of disease (complementarity with existingtests and examinations, clinical decisions, etc.).Such information from algorithms was notcompiled within the framework of this report.

Furthermore, a number of adjustments wouldhave to be made to prevalence data, since, inoncology, for example, any patient in remissionis considered a prevalent case, which wouldprobably distort the hypothetical PET utiliza-tion rates upwards, unless corrections are madeto take into account the prognosis, the length oftime before recurrence, survival times, etc.Thus, the incidence or prevalence data willhave to be weighted on the basis of the recog-nized roles of PET for the different specialties.

Models of these estimates were not performedwithin the framework of this assessment. How-ever, based on the opinions expressed by themembers of the advisory committee, the num-

ber of patients in Québec who could benefitfrom PET would put the annual number ofscans at about 15,000 or more.

The applications of PET with a potential forclinical use are an immediately fertile groundfor validating the efficacy and efficiency ofthese uses, and the validation activities wouldbe highly relevant to the new-technology as-sessment mandate that university hospital cen-tres and university institutes have under An Actrespecting health services and social services(R.S.Q., c. S-4.2, ss. 88 and 89, amendments ofJune 15, 2000), provided they are equippedwith PET.

In brief, however approximately they havebeen determined, there seems to be sufficientclinical need to justify the appropriate deploy-ment of PET at facilities that serve oncology,cardiology and neurology patients. For some ofthese facilities, it would be advisable to link theclinical activities with the assessment and re-search activities in order to promote the acqui-sition of the data needed for optimal PET utili-zation in Québec.

6.2 PHYSICAL AND HUMAN RESOURCES

REQUIRED FOR PET

The cost of providing space and purchasing asufficient number of cyclotrons and scannersfor Québec can be briefly estimated. Based onthe figures presented above, about 15,000 scansper year, at the rate of about 1,500 per scannerper year, calculated on the basis of one 8-hourshift per day, would monopolize at least about10 scanners. This approximate calculation doesnot, however, take into account the operationalaspects specific to each area of application. For


Avenues of deployment of PET in Québec72

example, in oncology, a PET scan may be per-formed daily on a larger number of patientsthan perfusion scans in cardiology with ammo-nia. We could thus consider adding, for thepurposes of our calculations, a few more scan-ners to take this situation into account.

While the number of scans per scanner dependson the type of application, the short half-life (afew minutes or less) of certain isotopes makestransporting them over long distances unfeasi-ble and requires a cyclotron near the patient,especially for perfusion studies in cardiology.In these conditions, the number of cyclotronscould be increased according to the number ofsites chosen, for perfusion can be measured bymeans of rubidium-82 as rubidium chloride,oxygen-15 as water or nitrogen-13 as ammo-nia. No rubidium generator has been approvedby Health Canada. Oxygen-15 is used mainlyin research.

Thus, the production of 13N for clinical pur-poses will require a cyclotron near facilitiesthat perform heart scans. See Chapter 2 andAppendices 1 and 2 for further details on half-lives and transport.

Myocardial perfusion can be measured withSPECT using radioisotopes that have longerhalf-lives than that of 13N, e.g., thallium-201and technicium-99m, and myocardial viabilitycan be evaluated with 18FDG. In such cases,there is no need to have a cyclotron near the fa-cility.

Since a cyclotron can provide the necessary ra-dioisotopes for operating about three or fourscanners, it would take about three or four cy-clotrons to supply 10 or 12 scanners.

Since there are already two cyclotrons in Qué-bec, namely, in Sherbrooke and Montréal,adding a third and a fourth or even a fifth cy-clotron, which would be located at an appropri-

ate distance from the new PET centres, could,together with the two existing cyclotrons, meetthe need for radioisotopes for all of Québec.

Without repeating the cost figures used in themodels in Section 4.4 or all the ranges for thecosts of the different components (providingspace for scanners and cyclotrons, purchase ofequipment, supplies and service contracts in-herent in PET operations; see Appendix 10), itcan be said that the overall cost of additionalPET deployment, in addition to the existingcentres in Québec, could vary from several tensof millions of dollars to more than about $100million, depending on the sites chosen (leadedvaults, etc.) and the proposed number of scan-ners and cyclotrons.

The variables in terms of equipment cannot,however, be disassociated from the human re-sources variables, and the issue now is mainlyno longer just the number and location of thescanners and cyclotrons required to meet theneeds of the Québec population in the shortterm, but also how many dedicated scannerscould be operational in the short term, giventhe specialized physical and human resourcesneeded for the clinical use of PET.

It should be borne in mind that the human re-sources requirements for PET utilization are afunction of the technological features that arespecific to it. Thus, the use of positron-emittingtracers that have to be produced near or evenon the premises requires personnel capable ofoperating cyclotrons, extracting the radioiso-topes that are generated, incorporating theminto products for human use, ensuring theirquality and transport, performing scans, andlastly, correctly interpreting the results. In otherwords, operating a PET centre equipped withscanners and a cyclotron requires a teamtrained specifically in PET consisting of radio-physicists, radiochemists, radiopharmacists,nuclear medicine technicians and nurses, and


Avenues of deployment of PET in Québec73

nuclear medicine physicians.

Although we are unable to present accuratedata on the available manpower other than thatat the existing centres, specifically in terms ofspecialized human resources, we can say thatthere is currently a shortage of such resourcesand that this shortage would become apparentif scanners and cyclotrons were installed atnew, clinically oriented sites in the short term.The training of specialized personnel becomesas limiting a factor as providing space and pur-chasing equipment. Self-training in this high-tech area does not seem to be a very feasiblesolution.

One possible approach for facilitating the de-ployment of PET in Québec might be to givethe centres that are already equipped with PETan explicit mandate to train the specialized per-sonnel needed to operate future PET centres.This would reduce the current need to obtainsuch training outside Québec. The clinical

service output capacity at these centres shouldbe bolstered in order to carry out this mandate.

It would also be advisable to plan the best sce-narios for providing other university hospitalcentres and university institutes with the facili-ties and human resources needed to offer clini-cal PET services. Whatever the scenarios, theyshould be part of a deployment strategy basedon meeting the population's needs and centredon cooperation between the different sectorsand the implementation of quality standards.

In short, because of the operational and ad-ministrative aspects to be taken into considera-tion when deploying PET in Québec, and thisbeyond simply recognizing its clinical efficacy,a ministerial master plan should govern thisdeployment, taking into account the popula-tion's clinical needs and the specialized humanand physical resources required by this tech-nology.


Conclusions and recommendations75

7. CONCLUSIONS AND RECOMMENDATIONS

7.1 CONCLUSIONS

� Because of its ability to provide informa-tion on both the anatomical location of tis-sues and their functional dynamics, PETmakes an important contribution to medi-cal imaging.

� Recent, structured assessment reports andAÉTMIS’s update on works published afterthese reports lead to the conclusion that, atthis time, although there are still few hardcomparative study data on PET's clinicalperformance, they are sufficient to supportthe recognition of various clinical applica-tions in oncology, cardiology and neurol-ogy.

� The reports underscore the paucity of dataon the cost-effectiveness of this technol-ogy. However more-recent economicstudies and the models performed byAÉTMIS and adapted to the situation inQuébec suggest that PET could prove to beefficient in managing patients with non-small-cell lung cancer and in myocardialviability studies. These findings are stillhypothetical and require further examina-

tion with more-complete, validated data.As for uses other than the two mentionedhere, PET's efficiency in the Québec con-text is still not known, as is the case formost other health-care systems.

� More than one payer already covers certainclinical applications of PET, which in-crease in number each year. The condi-tions governing reimbursement vary bothaccording to the organization and the ap-plication. These restrictions reflect theseorganizations' eagerness to contain the useof PET on a first-recourse basis for unrec-ognized applications and to generate datathat can further document the context ofrecognized uses.

The clinical uses examined thus far are, to alarge extent, based on diagnostic performanceor therapeutic guidance criteria, not on themeasurement of effects on patient survival orquality of life. Despite these limitations, the ef-ficacy of PET in clinical applications seemssufficiently documented and promising to in-crease the Québec population's access to it.


Conclusions and recommendations76

7.2 RECOMMENDATIONS

Deployment

� Since the clinical efficacy of PET is recog-nized for many oncological, cardiologicaland neurological applications, it would beadvisable to promote and support, in Qué-bec's public health-care system, the deploy-ment of this technology for clinical pur-poses.

� PET scans should be performed first forclinical applications with recognized clinicalefficacy. These applications should be re-viewed periodically as new hard data re-flecting rapid changes in the relevant infor-mation become available.

Deployment Specifics

� A master plan for deploying PET should beprepared by the Ministère de la Santé et desServices sociaux.

� The plan should include quantifying thepopulation's PET scan needs, both with re-gard to optimizing the other existing tech-

nologies and to the requirements for PET forwhich human and physical resources mightsometimes prove to be a limiting factor. Theplan should therefore be prepared both incollaboration with the existing PET centres,whose expertise can be used to good ad-vantage, and with the different key playersin tertiary intervention settings.

� The plan should take into account the factthat PET cannot be deployed for clinicalpurposes without conducting research intopromising applications that are presentlyrecognized as potential uses but whose effi-cacy and cost-effectiveness have not yetbeen demonstrated.

� Since the plan should include an evaluationof the efficiency of PET while it is beingdeployed for clinical purposes, the planshould be carried out in close collaborationwith university hospital centres and univer-sity institutes.

* * * * *


Appendix 1: Technical data on PET77

Appendix 1

Technical Data on PET



APPENDIX 1: TECHNICAL DATA ON PET

A1.1 EQUIPMENT FOR PRODUCING POSITRON-EMITTING RADIOISOTOPES

The main principles of interest in positron-emission tomography can be divided into twocategories according to the method of produc-tion: isotopes produced by a cyclotron andthose produced by a generator.

Isotopes produced by a cyclotron

Oxygen-15 (15O)Fluorine-18 (18F)Nitrogen-13 (13N)Carbon-11 (11C)

All four of the main isotopes produced by amedical cyclotron are used for clinical applica-tions of PET or for its routine research applica-tions. Of these, fluorine-18 (for labelling FDG)suffices for all the current clinical uses in on-cology. The advantage of fluorine is that it canbe delivered to locations far from the produc-tion site. In oncology, the other isotopes areused for research. In cardiology, however, ni-trogen-13, oxygen-15 and carbon-11 have di-rect usefulness in clinical (and clinical re-search) applications.

A cyclotron is a particle accelerator that accel-erates ions (in general, the new cyclotrons usedin medicine accelerate a proton that has twoelectrons (an H-, i.e., a negatively charged hy-drogen atom) to transform an isotope (nonra-dioactive) into a radioactive isotope. For ex-ample, to produce 18F, one generally bombardsabout 1 to 2 mL of H2

18O (which costs about$250/mL) with a beam of protons for the reac-tion 18O(p,n)18F. The proton strikes the nucleusof the oxygen-18 atom, and a neutron is ex-pelled. The new atom, whose nucleus containsexcess protons, becomes radioactive. The pro-tons come from the beam of hydrogen atoms,

since their electrons are removed beforereaching the target.

Cyclotrons come in different sizes:

< 5 MeV Dedicated to the production of oxy-gen-15 only (research).

10-13 MeV Can produce the four common iso-topes with good yields.

This is the type of cyclotron installedmost often in hospitals.

17-19 MeV The output is generally higher (canserve several hospitals).

The option of a deuteron beam onthese models reduces the cost of pro-ducing 15O if research studies in neu-rology and psychiatry are a priority.

In general, these cyclotrons are moreversatile for producing isotopes usedless frequently in research, such asiodine-124, copper-64 and bromium-76.

≥ 30 MeV These cyclotrons are used for thecommercial production of isotopesfor conventional nuclear medicine(Tl-201, I-123, In-111, etc.), but theycan also produce the four isotopescommonly used in PET. These in-struments are necessary only insituations where they are justifiedeconomically by the commercialneed for conventional isotopes in nu-clear medicine.

The purchase price of a cyclotron usually de-pends on its accelerating power (10-19 MeV),the number of targets (each target produces oneisotope), the number of irradiation ports (1 or2), the option of deuteron acceleration, theshielding (partial, complete or none) and theoption of varying the beam energy. It also de-pends on the facilities required for labelling



compounds for medical use (such as FDG)with isotopes (such as fluorine-18). The proc-ess of selecting a cyclotron is a complex onethat requires an accurate needs assessment andspecial expertise in this field.

Isotopes produced by a generator

Gallium-68Rubidium-82Copper-62

A generator is not an instrument or a machine.It is a concept based on the disintegration ofcertain long-half-life radioactive isotopes intoradioactive isotopes with shorter half-lives. Aclassic example is the 99Mo/99mTc generator,which is used in all nuclear medicine depart-ments. Molybdenum-99, which has a half-lifeof three days, disintegrates into Tc-99m, whichhas a half-life of six hours. Since the chemicalproperties of Tc-99m are different from thoseof Mo-99, they can be separated. Thus, by de-livering, each week, a 99Mo/99mTc generatorsystem, one can provide, Tc-99m to depart-ments of nuclear medicine on a daily basis. The68Ge/68Ga generator is used in PET in sealedsources or cylinders for certain quality con-trols. 68Ge does not emit positrons, but its deg-radation product, 68Ga, does. With the half-lifeof 68Ga approaching nine months, one can pur-chase, once a year, a 68Ga production sourcethat can last one year. Thus, the lifespan of agenerator depends on the half-life of the parentatom. Isotopes produced by generators are usu-ally heavier atoms that are not easily incorpo-rated into compounds of biological interest andare therefore of limited interest.

Rubidium-82 can be used in place of nitrogen-13 for cardiac perfusion studies. Rb-82's shorthalf-life permits shorter studies than nitrogen-13, but the images are of considerably lesserquality. However, the diagnostic reliability isthe same. The cost of a commercial generator

is about $40,000, and it needs to be replacedevery month. Some centres build their owngenerators. In such cases, the cost can drop toabout $8,000 a month.

Gallium-68 and copper-62 presently have novalidated clinical uses, but, in the long term,new diagnostic agents may be developed fromthese isotopes.

Thus, a generator cannot, under any circum-stances, replace a medical cyclotron for pro-ducing isotopes of interest in PET. Currently,the only clinical use of a generator is the utili-zation of Rb-82 for perfusion and viabilitystudies in PET, as an addition to FDG, for fa-cilities that do cardiology and that do not haveaccess to short-half-life isotopes.

Positron emission tomography differs by itsimaging principle, which is based on the detec-tion of two gamma rays emitted simultane-ously. Tomography in conventional nuclearmedicine (single-photon emission computedtomography [SPECT]) is based on a differentprinciple: a single gamma ray is emitted, and alead collimator (the counterpart of an opticallens in a camera) serves to identify the ray'sorigin. While PET images the photons emittedby positron emitters (18F, 15O, 13N, 11C) ,SPECT images the photons emitted by gammaemitters (99mTc, 131I , 201Tl, 111In, 67Ga). Theradioactive atoms used in SPECT are heavierand generally do not permit a direct measure-ment of biochemical processes in the humanbody. One cannot label deoxyglucose with io-dine, for example, and preserve its biologicalproperties.

SPECT scanners cannot image positron emit-ters without unacceptable compromises in im-age quality (Section A1.1.1). Some SPECTscanners are equipped with hybrid circuits thatenable them to detect the two photons emittedsimultaneously (Section A1.1.2).



There are several types of instruments designedfor positron emission tomography. These in-struments (PET scanners or PET cameras) areto be distinguished from cyclotrons, whichproduce radioisotopes. A PET scanner consistsof detectors (made of scintillating crystals)coupled to photomultiplier tubes and of anelectronic signal amplification and processingdevice. The scintillating crystal is a key deter-mining factor of a PET scanner's performance,but other aspects, such as the geometric con-figuration and the underlying electronics, alsodetermine an instrument’s clinical perform-ance. The detection principle in PET is basedon the emission of one positron per radioactiveatom. The positron travels a very short distance(1/10 of one millimeter) before annihilating,emitting two gamma rays in opposite direc-tions. Since two rays are emitted at 180°, de-tecting this radiation does not normally requirethe addition of collimators (lead lenses), as inconventional nuclear medicine. The detectionsensitivity (percentage of radiation emitted bythe patient that is picked up by the PET scan-ner) is thus higher than in conventional nuclearmedicine. The spatial resolution (the instru-ment’s ability to visualize small lesions) is alsovery good (typically 4.5 to 5.5 mm).

Technical parameters apart from sensitivity andspatial resolution are important as well, such asthe peak noise-equivalent count rate, which in-dicates the instrument’s ability to process alarge amount of information at the same time.The crystals used in dedicated or hybrid PETscanners generally consist of BGO or NaI(Tl).As of this year, new scanners are equippedwith detectors containing better-performingcrystals (LSO and GSO), but their availabilityis still limited.

There are three main categories of instrumentsthat can be used for detection in PET:

A1.1.1 Shielded conventional nuclear medi-cine camera equipped with a high-energy collimator (511 keV)

Since a heavy lead collimator masks more than95% of the radiation emitted by the patient, thisapproach has been rejected for imaging in on-cology, since tumors smaller than 3 cm cannotbe detected. In cardiology, some publicationsreport success with this approach for evaluatingmyocardial viability.

A1.1.2 Collimator-less hybrid nuclear medi-cine camera for detecting coincidencephotons

These systems use nuclear medicine gammacameras, to which shielding, a thicker crystaland special electronics for detecting the pho-tons emitted in coincidence upon the positron'sannihilation have been added. This option in-creases the cost of a new gamma camera byabout $250,000. Some recent cameras cansometimes be retrofitted with this option. Thisrequires certain compromises in the quality ofconventional nuclear-medicine studies. Most ofthe existing equipment in Québec cannot beupgraded for this technology. This approachwas developed for centres with a very smallpotential number of patients requiring PETscans. In theory, the spatial resolution is ex-cellent but in practice is poor: these scanners'peak noise-equivalent count rate is so low thatthe injected dose has to be decreased or extralead filters have to be added to prevent the oth-ers from becoming saturated. The scans take along time, and lesions smaller than 2 cm areoften not detected. Since many of these scan-ners cannot adequately correct for the absorp-tion and scattering of the rays by the patient'stissues, the studies are nonquantitative. Cardiacstudies are possible only on scanners equippedwith coincidence attenuation correction (pres-ently, the minority). These scanners may be



useful for certain applications (evaluating a re-sidual mass or pulmonary nodules ≥ 2 cm)when the anticipated volume of activity cannotjustify the purchase of a dedicated PET scan-ner. The literature still provides too little sup-port to justify the wide-scale use of these scan-ners. This is why the Health Care FinancingAdministration (which administers Medicare inthe United States) has not approved coverageof new uses for scans performed with these in-struments.

A1.1.3 Dedicated PET instruments

These scanners are fully dedicated to PET im-aging. They are generally suitable for clinicaluse, but their features vary. The more expen-sive ones are more versatile for research appli-cations, while the less expensive ones are gen-erally somewhat more limited to routine clini-cal use. There are several types of dedicated in-struments:

A1.1.3.1 Partial-ring scanners

These instruments are equipped with partialdetector rings, which reduces the constructioncosts, but a full ring can be simulated by rotat-ing banks of detectors. The detection sensitiv-ity is lower than that of the other scanners thatare currently available but is the same as that ofthe scanners that have been used in most of theclinical trials published to date.

A1.1.3.2 NaI(Tl) full-ring scanners

These scanners use full rings but large, curvedor rectangular crystals. They have a high de-tection sensitivity, good spatial resolution and apeak noise-equivalent count rate suitable foroncology but are of more limited use for re-search applications in cardiology and neurol-ogy, which require a very good count rate per-formance. These scanners' performance in on-cology imaging is generally equivalent to that

of multidetector scanners, but they are less ver-satile. They are less expensive.

A1.1.3.3 Full-ring multidetector scanners

Instead of using large crystals, these scannersuse matrices of small, individual crystals. Ingeneral, they have a better peak noise-equivalent count rate, which is not essential inclinical oncology imaging, but which may benecessary in several research and cardiologyapplications. These instruments’ performancevaries according to the size of the crystals andthe processing electronics. Their cost generallydepends on the number of crystals.

A1.1.3.4 PET scanner coupled to an axial CTscanner

Several prototypes have been announced or areavailable. They are PET scanners coupled to aCT scanner. Computed tomography is used foranatomical localization and to make certainimage corrections. These systems are expen-sive, costing $500,000 to $1 million more thana dedicated multidetector PET scanner. Theability to directly merge images is an attractiveone for certain applications, such as planningradiation therapy for oncology patients, butevaluative research still needs to be conductedto determine the incremental benefit of inte-grated computed tomography in relation to aPET study performed separately.

A1.1.3.5 Scanner with GSO and LSO crystals

Two scanners are now commercially available(or will be in the next few months). Within thenext few years, it is expected that most PETscanners will be using these better-performingcrystals. These scanners should reduce PETscan length by one half and therefore consid-erably improve patient flow, without scanquality being affected. Their cost will probablybe close to that of a top-of-the-line, dedicated,



multiple-crystal scanner. Their performancecharacteristics have not yet been published.

A1.2 SAFETY

The recent Australian report [Medicare Serv-ices Advisory Committee (MSAC): Ghersi etal., 2000] summarizes the available informa-tion. PET is a noninvasive diagnostic proce-dure whose safety is generally recognized.Safety-related issues mainly concern the radio-pharmaceuticals rather than the procedure as awhole.

A retrospective study [Silberstein et al., 1998]was conducted at 22 PET centres in the UnitedStates in order to determine the prevalence ofadverse reactions to positron emitters, mainly18F-fluorodeoxyglucose (FDG), but also 11C-CO2,

11C-methionine, 13N-NH3 and 15O-H2O.The centres provided retrospective data thathad been gathered from when they had openedup to 1994 and monthly prospective data gath-ered between 1994 and 1997. The data concern33,295 retrospective doses and 47,876 pro-spective doses, for a total of 81,801 doses ofpositron-emitting radiopharmaceuticals. Noadverse reactions to these some 80,000 doseswere reported or observed, with a 95% confi-dence interval of 3.7 per 100,000 doses.

According to this large prospective study, thesafety of the radiopharmaceuticals used inpositron emission tomography is well docu-mented and seems very high. The United StatesPharmacopeia - Drug Information [USP DI,1998] also states that there are no known ad-verse reactions associated with the use of FDG.Since the radiotracers are usually used in mi-crogram amounts, the incidence of adverse re-actions to these very small quantities of la-belled substances is very small.

The available information on the distribution of

PET centres in Europe and North America andon the number of PET scans per year perpopulation is provided in Appendix 3.

There are no absolute contraindications to aPET scan. Claustrophobia is seldom a problem(< 0.5%). In very severe cases, one must forgothe examination. This problem is more pro-nounced with PET/CT systems, where the pa-tient is placed in a long tunnel, as in MRI. Veryobese patients (> 158.5 kg) cannot be scanned.

A1.3 PRE-PET SCAN PATIENT PREPARATION

Typically, a PET scan consists of transmissionimages and emission images. Transmission im-ages are used to correct emission image datafor the attenuation effects of soft tissues and toalign different emission images (from the sameor a different session). The images can be in-terpreted visually or semiquantitatively. Twosemiquantitative parameters commonly usedare the standardized uptake ratio (SUR) and thelesion-to-background ratio [ICSI, 2001].

A1.3.1 Oncology studies

Patients need to be in the fasting state duringthe six hours preceding the scan. However,they are encouraged to drink water in order tomaintain good hydration, and they may taketheir usual medications. Fasting is necessary inorder to keep the plasma glucose and insulinlevels as low as possible. A high glucose levelcompetes with FDG and thus reduces tumoruptake. A high insulin level promotes uptake ofthe tracer by the muscles and heart at the ex-pense of tumor uptake.

Precautions should be taken for diabetic pa-tients. It is advisable to reduce their blood glu-cose level with intravenous insulin before in-jecting FDG. Since the insulin level should be



low when FDG is injected, it should always beinjected at least one hour after the last dose ofintravenous insulin.

Some very tense patients require the admini-stration of a benzodiazepine before FDG isadministered, in order to prevent strong muscleuptake, which occurs near the neck, axillae andgreater supraclavicular fossae. At many cen-tres, patients scanned for neck cancer system-atically take benzodiazepines to prevent theseartifacts.

A1.3.2 Brain studies

The initial preparation is the same as for oncol-ogy studies, i.e., fasting and glycemic control.

A1.3.3 Cardiac studies

Rest perfusionNo special preparation is required.

Stress perfusionPatients must be in the fasting state.

Myocardial viabilityUnlike in oncology studies, FDG cardiac up-take is desirable.

Patients need to eat a light meal in the morningof the PET scan. Upon their arrival, nondia-betic patients receive sweetened juice (oralglucose load) containing 25 to 75 grams of glu-cose. One hour later, FDG is administered, to-gether with a small dose of insulin, if need be.

Diabetic patients should be stabilized by meansof a long procedure consisting in administeringa supraphysiologic concentration of insulin byinfusion and in stabilizing the blood glucoselevel with a second infusion with a dextrosesolution. This procedure, referred to as "eugly-

cemic-hyperinsulinemic clamp", takes 90 to120 minutes.

A1.4 ROUTINE CLINICAL PROCEDURES USED

DURING PET SCANNING

A1.4.1 Oncology studies

The radiotracer used in oncology studies is18FDG. It is administered to the patient via anintravenous catheter introduced into a vein inthe arm. Direct intravenous injection should beavoided, since it could result in an artifactualaccumulation of the substance in the axillarylymph nodes. The injection is done outside thePET scanning room. An hour later, the patientis placed in the scanner, lying face up, prefera-bly with the arms folded above the head or, ifthe patient is uncomfortable doing this, along-side the body. Obtaining images generallytakes between 45 and 60 minutes, dependingon the patient's size and the extent of the ex-amination.

Since FDG is excreted by the kidneys and ac-cumulates in the bladder, it is occasionally nec-essary to obtain additional images of the pelvis,after administering a diuretic. This dilutes andeliminates urinary activity and permits a betterevaluation of the perivesical region, whichwould otherwise be masked by a significantaccumulation of FDG in the bladder. This re-quires an extra imaging session, which takesabout 90 minutes. Ovarian, cervical, colon,rectal and bladder cancers and lymphoma mayfrequently necessitate this procedure.

A1.4.2 Brain studies

The radiotracer used in brain studies is 18FDG.Images are taken 45 minutes after it is injected.The scanning takes between 15 and 30 minutes.



A1.4.3 Cardiac studies

Perfusion (rest and stress)

13NH3, 82Rb or H215O can be used as the ra-

diotracer.

An intravenous catheter is introduced into avein in the arm, and the patient is positioned,prior to injection, in the scanner. Positioningthe patient is a delicate operation and can take10 to 15 minutes. Dynamic acquisition beginsat the same time as the injection and lasts for atotal of 20 to 30 minutes.

The procedure for stress perfusion is the sameas that for rest perfusion, but the agent is in-

jected after the administration of a myocardialstress agent, such as dobutamine, adenosine ordipyridamole.

A1.4.4 Glucose metabolism studies

18FDG is used as the radiotracer.

The radiotracer is injected 45 minutes beforeimage acquisition, when the patient is outsidethe scanning room. The patient is positioned inthe same manner as for a perfusion study. Im-age acquisition takes 20 to 30 minutes. A myo-cardial viability study always consists of twoparts: a rest perfusion study and a glucose me-tabolism study. Both are usually performed onthe same day.


Appendix 2: Radiation protection in the positron emission tomography (PET) laboratory87

Appendix 2

Radiation Protection in the Positron Emission Tomography(PET) Laboratory



APPENDIX 2: RADIATION PROTECTION IN THE POSITRON EMISSIONTOMOGRAPHY (PET) LABORATORY

A2.1 INTRODUCTION

The problem of radiation protection is very dif-ferent in a PET laboratory than in a conven-tional nuclear medicine department, mainly be-cause of the very short half-lives of the radioi-sotopes used in such a laboratory.

Four radioisotopes are used at the laboratory ofCHUS's Clinical Research Centre (CRC). Theisotopes in question are listed in the followingtable, together with their half-lives.

Table 8: Cyclotrons products

PRODUCT HALF-LIFE

11C 20 minutes[C-11]CO2, [C-11]CO, [C-11]CH3I[N-13]ammonia 10 minutes15O 2 minutes[O-15]H2O, [O-15]CO2, [O-15]CO, [O-15]O218F 110 minutes[F-18]F2, [F-18]F-, [F-18]-L-DOPA, [F-18]FDG

These isotopes are created by specific chemicalreactions that take place during the collision ofelectrons with targets consisting of various sta-ble substances, as indicated in Tables 9 and 10.

The production yield varies from 70 to 90%,depending on the parameters used and the pu-rity of the target that undergoes bombardment.

Table 9: Nuclear reactions for producing radioisotopes

RADIOISOTOPE NUCLEAR REACTION TARGET

Carbone-11 14N[p,α]11C 14N - NitrogenNitrogen-13 16O[p,α]13N H2

16O/Ethanol solutionOxygen-15 15N[p,n]15O 15N - Enriched nitrogenFluorine-18 [18F]F2

18O[p,n]18F 18O2 - Enriched gasFluorine-18 [18F]F- 18O[p,n]18F H2

18O - Enriched water



Table 10: Radioisotope production parameters

TARGET REACTION CURRENT BOMBARDMENT RESULTING

TIME ACTIVITY14N 14N[p,α]11C 40 µA 40 min 1,500 mCi16O 16O[p,α]13N 20 µA 20 min 800 mCi15N 15N[p,n]15O 30 µA 6 min 1,000 mCi18O 18O[p,n]18F 40 µA 40 - 60 min 1,000 mCi

A2.2 REQUIRED LICENSES

For a PET centre to be able to operate properly,it needs several licenses from the CanadianNuclear Safety Commission (CNSC). To beginwith, a license to use a cyclotron is required, ifthe centre has one. It is a class II nonmedicalparticle accelerator operating license, whichauthorizes the holder to produce, use, disposeof and distribute radioisotopes.

The use of these radioisotopes also requires autilization license, which authorizes the holderto possess, use, dispose of, store, distribute andimport radioisotopes whose total activity ex-ceeds 10 GBq. A list of the radioisotopes usedmust appear on the license.

Distributing radioisotopes to other centres re-quires support from all those involved at thedistribution centre and the user centres. Thetransport distance between the distributioncentre and a user centre should be factored intothe radioisotope production process. The ra-dioisotope's physical half-life is used to calcu-late the activity to be generated. It is generallyagreed that the transport time between the twocentres should not exceed two half-lives. Forexample, 18F and 18FDG, in liquid form, can betransported between two cities more than 100km apart. On the other hand, the isotopes 11C,13N and 15O cannot be transported from onecity to another because their half-lives are tooshort. One would have to produce a very large

quantity of isotope in order to obtain the verysmall amount needed. This is especially thecase with the latter two isotopes, which cannotbe transported from one floor to another withina given building.

A2.3 THE PERSONNEL REQUIRED

1. Cyclotron laboratory personnel.

� The person who operates the cyclo-tron that produces the radioisotope(18F-).

� A radiochemist for collecting theproduct of the cyclotron (18F-) and forsynthesizing the radiotracer (FDG).

� A radiopharmacist for purifying thesynthesized radiotracer (FDG) and forits quality control.

2. A nuclear medicine specialist, who contactsthe user centre and determines the totalamount to be shipped.

3 . A radiation protection supervisor, whooversees the procedures for producing andpacking the product to be shipped.

4. A carrier accredited by Health Canada andthe CNSC. The carrier must have a licensein due form for transporting radioisotopes.



5. A nuclear medicine specialist at the usersite, who supervises the receipt of the pack-age and conducts leak tests.

A2.4 PACKAGING FOR THE RADIOISOTOPE

The radioisotope must be transported in a typeA package, which can be a cardboard orwooden box or a metal drum. The packagemust have been manufactured in accordancewith engineering standards. The general fea-tures of the package are as follows:

� The minimum edge length is 100 mm.

� The package must be sealable.

� The outside surfaces must be smooth andfree of protrusions, unless they are fortransporting the package.

� The package must the able to withstand awide range of temperatures (70°C to-40°C).

� It must be resistant to breakage and crack-ing if the temperature is outside the safetyrange.

� It must be able to retain the radioactivecontents if the ambient pressure drops tobelow 25 kPa.

� The package must contain enough absorb-ent material to contain more than twice thevolume of radioactive liquid being trans-ported (≤≤≤≤ 50 mL) in the event of a leak inthe package’s inner container.

� The outer package must have more thantwice the volume of the radioactive liquidbeing transported (> 50 mL) in the event ofa leak in the package’s inner container.

A self-adhesive radioactive label indicating theisotope being transported, its activity and itstransport category must be placed on the pack-age. There are three types of self-adhesive la-bels:

� Radioactive I: The radiation level does notexceed 5 µSv/hr anywhere on the externalsurface of the package.

� Radioactive II: The radiation level does notexceed 500 µSv/h anywhere on the externalsurface of the package. Transport index≤ 1.

� Radioactive III: The radiation level doesnot exceed 2,000 µSv/h anywhere on theexternal surface of the package. Transportindex ≤ 10.

Figure 1: Categories of radioactivity



A2.5 RADIOISOTOPE DISTRIBUTION

DOCUMENTS

The radioisotope distribution centre must havea certain number of important documents. Theyinclude the following:

� The procedure for packing the radioiso-topes produced.

� A license to possess, import, use, store anddispose of the radioisotope received (18F as18F- or 18FDG) or radioisotopes whoseatomic number is between 1 and 89. The li-cense is issued by the CNSC.

� A license to transport high-activity radioi-sotopes, which is issued to the carrier bythe CNSC.

� The carrier's emergency procedures in theevent of an accident.

A2.6 PROCEDURE FOR PREPARING RA-

DIOACTIVE PACKAGES

The packing procedure comprises five impor-tant steps. They may be supplemented by addi-tional verification procedures, at the facility'sdiscretion. The five steps are as follows:

Step 1: Check the package for contamination

� Measure the radiation level with agamma counter.

� Perform a dry-smear test on theinternal and external surfaces of theinner package and on the externalsurface of the outer package.

� No contamination is allowed on theinner package.

� Allowable fixed contamination onthe outer package: ≤ 0.5 µSv/h at0.5 m (higher fixed contaminationmeans reduced activity).

� Allowable nonfixed contaminationon the outer package: ≤ 0.5 Bq/cm_on a surface of 100 cm_ (test per-formed with dry cotton).

Step 2: Check that it is a type A package.

� Check that the package is not dam-aged.

� Check that any openings are sealed.

� Make unusable any component ofthe package that is not intended forits transport.

Step 3: Check that the inner vial is not con-taminated.

� Perform a smear test on the outersurface of the vial. No contamina-tion is allowed.

Step 4: Pack the vial in the inner package.

� Place enough absorbent material inthe inner package to contain twicethe volume of the liquid radioiso-tope being shipped.

� Seal the inner package.

� Perform a leak test on the surfaceof the inner package.

� Seal the inner package.

� Seal the outer package.

Step 5: Determine the index radiation level.

� Using a contamination counter,measure the radiation level at 1 mfrom the outer surface of the pack-age. Record the value in mR/hr.

Complete the hazardous materials bill of lad-ing. A copy accompanies the package and isgiven to the requesting site. Another copy iskept by the carrier. The distrubution site keepsthe original copy.



Figure 2: Hazardous materials bill of lading (specimen)

A2.7 THE CARRIER

The carrier must be a company recognized bythe CNSC as having all the necessary qualifi-cations for carrying out this type of transport.In Canada, two such companies that regularlydo this type of transport: DUPONT and NY-COMED AMERSHAM. Companies likePUROLATOR and FEDERAL EXPRESS dogeneral transport. It is not advisable to use suchcompanies, since the management of emer-gency situations is not as well documented asfor specialized companies and since the levelof responsibility is not the same. Transportingradioactive materials requires special handlingof the packages and specialized knowledge ofradioisotopes.

A2.8 EMERGENCY PROCEDURE

A2.8.1 General emergency procedure

The first step of an emergency procedure in theevent of an accident is to call 911. First-linesupport is provided by the local fire depart-ment. It will determine the extent of the dam-age and set up a security perimeter. Decon-tamination of the site begins quickly. A de-contamination bath is set up to decontaminateanyone who may have been contaminated bythe substance. The local police department issubsequently contacted, then Urgence-Santé,



which provides ambulance service and contactsthe emergency rooms at the local hospitals. Italso contacts the regional health board, whichtakes charge of the overall emergency opera-tions.

A2.8.2 Procedure carried out by a special-ized carrier

First-line support is provided by the companyspecializing in on-the-road incidents, which ishired by the carrier. As soon as an accident oc-curs, the driver, if able to call the company,does so by dialing a special number at his orher disposal. The accident is managed by thespecialized company. It provides the equipmentand trucks required for decontamination. It alsocontacts the local police departments and hos-pitals. This emergency procedure must neces-sarily be approved by the CNSC through theissuance of a radioactive materials transport li-cense. If the driver is injured to the point wherehe or she is unable to contact the company, thegeneral emergency procedure is begun as soonas the call to 911 is made.

A2.9 LEVELS OF RESPONSIBILITY

The production site is responsible for the ra-dioactive material until it arrives at its destina-tion. Upon its arrival, the receiver contacts thedistribution site in order to release the latterfrom this responsibility.

The carrier is responsible for providing thetype A package, taking the radioactive materialto its destination and taking charge of theemergency procedures in the event of an acci-dent.

The receiving site is responsible for receivingthe package and the procedure for opening it,as adopted by the CNSC. The essential steps ofthis procedure are as follows:

� Check the package for any damage.

� Measure the radiation level at the surface ofthe package. If it exceeds 2 mSV/hr, itshould be assumed that there is a leak.

� Measure the radiation level at 1 m from thesurface of the package in order to check thetransport index.

� Perform a dry-smear test on the surface ofthe package, then on the internal surface inorder to determine if there are any leaks.

� Report any of the following abnormalitiesto the CNSC and the distribution site:

� Discovery of a leak in the package.

� Radiation level exceeding 10mSV/hr on the surface of the pack-age.

� Radiation level exceeding 200µSV/hr at 1 m from the surface ofthe package.

� Nonfixed contamination exceeding3.7 Bq/cm_.

The distribution centre must act promptly todetermine the cause of the incident and correctthe situation before the next shipment. The re-sults of the inquiry and any corrective proce-dure must be sent to the CNSC as soon as pos-sible.


Appendix 3: The geographic distribution of PET centres95

Appendix 3

The Geographic Distribution of PET Centres



APPENDIX 3: GEOGRAPHIC DISTRIBUTION OF PET CENTRESTable 11: Geographic distribution of PET centres

(1999 to 2001, depending on the source*)

CONTINENT,COUNTRY,REGION OR PROVINCE

NUMBER

OF PETCENTRES

NUMBER OF

SCANNERS:DEDICATED

(VS. TOTAL)

NUMBER OF

SCANNERS WITH

CYCLOTRON

(CYCLOTRON

ALONE)

SCANNERS PER 1MILLION

POPULATION

NUMBER OF PET SCANS

REIMBURSED PER

100,000 POPULATION

BY PUBLIC SYSTEMS

(JULY 98 - JUNE 99)AFRICA

Saudi Arabia 1NORTH AMERICA

U.S. 75 (120 coincidence) 0.4 17.8 Canada (6): 8 0.2 Québec 0.3 (Québec) 3.3 Montréal 1 1 Sherbrooke 1 1** Ontario 0.3 Hamilton 1 1 Ottawa 1 1 Toronto 1 1 British Columbia 0.3 Vancouver 1 1

Alberta (2001)ASIA

Japan 20 37 0.3 China 1 12 0.01 Korea 1 Israel 1 Taiwan 1

EUROPE

Germany 22 35 (73) 16 (2) 0.9 UK 6 7 (11) 5 0.2 Belgium 7 5 (9) 5 0.9 Russia 1 2 (9) 2 0.1 Italy 7 8 (9) 7 (1) 0.2 France 4 (7) 4 0.1 Austria 2 2 1 Czech Republic 1 1 1 Denmark 2 2 2 14.2 Finland 4 2 2 3.6 Hungary 1 1 1 Holland 1 3 2 0.3 Norway 1 Spain 2 3 2 (Basque Country) 6.5

(Madrid) 0.4 Sweden 4 2 2 Switzerland 3 3 17.2

OCEANIA

Australia 4 9.7 New-Zealand 0.2* Sources: Tashiro et al., 2001; Ghersi et al., 2000; Adams et al., 1999; and Beanlands et al., 1999.** One additional scanner was installed at this centre in February 2002.



Table 12: Clinical use of PET at CHUS

For the period from May 15, 2000 to May 11, 2001: 1,253 scans

Lung cancer: 33.4%Lymphoma: 21%Breast cancer: 6.6%Brain studies: 6.4%ENT cancer: 5.4%Cancer of the colon and rectum: 4.9%Gynecologic cancers: 4.9%Cardiac studies*: 2.5%Melanoma: 2.3%Testicular cancer: 1.8%Other cancers: 10.8%

* A sizeable proportion (about 50%) of cardiac viability studiesare performed in accordance with a research protocol (PARR-2study [Beanlands, 2000-2001]) and are not included under theheading of clinical use.

Table 13: Number of clinical scans and waiting times at CHUS

For the period from May 15, 2000 to May 11, 2001: 1,253 scans

Mean waiting time*: 46 daysMedian waiting time*: 32 days

*Achieved by very effective priority management

Waiting list as at June 1, 2001: 330 patients

Table 14: Source of patients at CHUS

Hospitalized at CHUS: 172 (13.7%)Hospitalized elsewhere: 57 (4.5%)Outpatients (Eastern Townships): 705 (56%)Referred from outside the E.T.: 303 (26%)



Figure 3: Growth in the demand for PET at CHUS (Sherbrooke)

0

20

40

60

80

100

120

140

160

180

D-98

M-9

9J-

99S-9

9D-9

9M

-00

J-00

S-00

D-00

M-0

1

Date

# o

f re

qu

ests

Requested

Done


Appendix 4: Location of Imaging Equipment in Québec101

Appendix 4

Location of Imaging Equipment in Québec


Appendix 4: Location of Imaging Equipment in Québec103

APPENDIX 4: LOCATION OF IMAGING EQUIPMENT (OTHER

THAN PET EQUIPMENT)

Table 15: Equipment other than PET equipment by region (Québec)*

REGION

CT SCANNERSMRI

SCANNERS**

NUCLEAR

MEDICINE

SCANNERS

01- Lower St. Lawrence3 1 4

02- Saguenay - Lac Saint-Jean 5 1 4

03- Québec 7 2 18

04- Mauricie and Central Québec 5 1 12

05- Eastern Townships 2 1 10

06- Montréal-Centre 28 11 62

07- Ottawa Valley 2 1 5

08- Abitibi-Témiscamingue 3 1 2

09- North Shore 2 3

11- Gaspé Peninsula and the Magdalen Islands 3 1

12- Chaudière-Appalaches 4 1 7

13- Laval 1 4

14- Lanaudière 2 1 2

15- Laurentians 2 3

16- Monteregia 8 1 10

TOTAL 77 22 147

* A detailed breakdown is provided on the following page.** The MRI scanner will be a mobile one and will serve

Region 08. In addition, in the short term, there are plansfor 7 other scanners, which will bring the total numberto 29.


Appendx 4: Location of Imaging Equipment in Québec104

Table 16: Imaging equipment in Québec[Source: MSSS, December 2000; + revisions 03/01]

Location of equipement NuclearCT scanners MRI scanners* medecine

FACILTY Region scannersCENTRE HOSPITALIER DE MATANE 01 1CENTRE HOSPITALIER RÉGIONAL DE RIMOUSKI 01 1 1 1CENTRE HOSPITALIER RÉGIONAL DU GRAND-PORTAGE 01 1 3CENTRE HOSPITALIER DE DOLBEAU 02 1CENTRE HOSPITALIER DE JONQUIÈRE 02 1COMPLEXE HOSPITALIER DE LA SAGAMIE 02 1 1 4PAVILLON DE L'HÔTEL-DIEU D'ALMA 02 1HÔTEL-DIEU DE ROBERVAL 02 1CHAUQ - HÔPITAL ENFANT-JÉSUS 03 2 1 2HÔPITAL LAVAL 03 1 4CHUQ - CHUL 03 1 4CHUQ - HÔTEL-DIEU DE QUÉBEC 03 1 3CHUQ - HÔPITAL SAINT-FRANCOIS D'ASSISE 03 1 1 3CHAUQ - HÔPITAL SAINT-SACREMENT 03 1 3CENTRE HOSPITALIER RÉGIONAL DE LA MAURICIE 04 1 2CHETR - HÔPITAL SAINTE-MARIE 04 1 1 3CHETR - HÔPITAL ST-JOSEPH 04 1 3HÔPITAL STE-CROIX 04 1 2HÔTEL-DIEU D'ARTHABASKA 04 1 2CHUS - BOWEN BRANCH 05 1 5CHUS - FLEURIMONT BRANCH 05 1 1 5CHUM - HÔPITAL NOTRE-DAME 06 2 1 4CENTRE HOSPITALIER DE LACHINE 06 1C.H. ANG. - VERDUN BRANCH 06 1 3CHUM - HÔPITAL HÔTEL-DIEU 06 2 1 11MUHS - MONTREAL CHILDREN'S HOSPITAL 06 1 1MUHS - MONTREAL GENERAL HOSPITAL 06 3 1 4MUHS - ROYAL VICTORIA HOSPITAL 06 2 8MUHS - MONTREAL NEUROLOGICAL HOSPITAL 06 1 1ST. MARY'S HOSPITAL 06 1 1CENTRE HOSPITALIER FLEURY 06 1HÔPITAL MAISONNEUVE-ROSEMONT 06 2 1 5CHUM - HÔPITAL ST-LUC 06 2 2 5HÔPITAL DU SACRÉ-COEUR DE MONTRÉAL 06 2 1 5LAKESHORE GENERAL HOSPITAL 06 1 2HÔPITAL JEAN-TALON 06 1HÔPITAL SAINTE-JUSTINE 06 1 1 2HÔPITAL SANTA CABRINI 06 1 1MONTREAL HEART INSTITUTE 06 1 5SIR MORTIMER B. DAVIS JEWISH GENERAL HOSPITAL 06 2 1 4C.H. DES VALLÉES DE L'OUTAOUAIS (CHRO) 07 1 1 3C.H. DES VALLÉES DE L'OUTAOUAIS (GATINEAU) 07 1 2MOBILE MRI UNIT*** 08 1CENTRE HOSPITALIER ROUYN-NORANDA 08 1CENTRE HOSPITALIER DE VAL-D'OR 08 1 2CENTRE HOSPITALIER HÔTEL-DIEU D'AMOS 08 1


Appendx 4: Location of Imaging Equipment in Québec105

Table 16: Imaging equipment in Québec (Cont’d)CENTRE HOSPITALIER RÉGIONAL DE SEPT-ILES 09 1 2PAVILLON LE ROYER 09 1 1CENTRE HOSPITALIER DE CHANDLER 11 1CENTRE HOSPITALIER BAIE-DES-CHALEURS 11 1CENTRE HOSPITALIER DE GASPÉ 11 1CENTRE HOSPITALIER DE L'ARCHIPEL 11 1CENTRE HOSPITALIER BEAUCE-ETCHEMIN 12 1 2HÔTEL-DIEU DE LÉVIS 12 1 1 3HÔTEL-DIEU DE MONTMAGNY 12 1CENTRE HOSPITALIER DE LA RÉGION DE L'AMIANTE 12 1 2CITÉ DE LA SANTÉ DE LAVAL 13 1 4CENTRE HOSPITALIER LE GARDEUR 14 1CENTRE HOSPITALIER RÉGIONAL DE LANAUDIÈRE 14 1 1 2CENTRE HOSPITALIER SAINT-EUSTACHE 15 1HÔTEL-DIEU DE ST-JÉRÔME 15 1 3C.H. RÉGIONAL DU SUROÎT À SALABERRY-DE-VALLEYFIELD 16 1 1CENTRE HOSPITALIER ANNA-LABERGE 16 1CENTRE HOSPITALIER DE GRANBY 16 1CENTRE HOSPITALIER PIERRE-BOUCHER 16 1 2HÔPITAL CHARLES-LEMOYNE 16 1 1 2HÔTEL-DIEU DE SOREL 16 1 1HÔPITAL DU HAUT-RICHELIEU 16 1 2RÉSEAU SANTÉ RICHELIEU - YAMASKA 16 1 2Total 77 22 146


Appendix 5: Description of assessment reports107

Appendix 5

Description of Assessment Reports



APPENDIX 5: DESCRIPTION OF ASSESSMENT REPORTS

A5.1 COUNCIL OF MEDICAL IMAGING

The Council of Medical Imaging, an Ontarioorganization, conducted a literature review onpositron technology [CMI: Beanlands et al.,1999]. The CMI's report was prepared for thegeneral public and describes, in plain language,positron technology and its validated and po-tential applications; gives the arguments in fa-vour of its deployment in clinical practice inOntario; and explains the proposal for a busi-ness plan based on clinical utility and cost-effectiveness evidence for providing Ontarionswith PET services. The report analyzes whatPET's impact would be on the health of thepopulation in question, with regard to improv-ing the standard of care and cost-effectiveness,for applications in oncology, cardiology andneurology. The CMI states that its main objec-tive would be to initiate dialogue between theinterested parties on the best way for Ontario tobenefit from technological advances in bio-medical sciences.

A5.2 MINNESOTA HEALTH TECHNOLOGY

ADVISORY COMMITTEE

In March 1999, the Minnesota Health Technol-ogy Advisory Committee (MHTAC) publisheda report on positron emission technology thatconcerns oncological applications only. Thereport's objectives were to evaluate, by meansof a literature review, the efficacy of PET indiagnosing and monitoring cancer patients; tocompare PET with other options; to assess theimpact of PET on patient management and theultimate effects on patient health; to determinethe cost-effectiveness of PET in the treatmentof cancer and to evaluate the effect of usinghybrid PET technologies on cancer. TheMHTAC's observations concerned brain tu-mors; cancers of the head and neck, thyroid,urinary tract and kidneys, lung, breast, esopha-

gus, pancreas, ovaries, prostate and testicles;colorectal, neuroendocrine and gastrointestinalcancer; and melanoma and lymphoma.

A5.3 MANAGEMENT DECISION AND RE-

SEARCH CENTER, DEPARTMENT OF

VETERANS AFFAIRS -TECHNOLOGY

ASSESSMENT PROGRAM (VA-TAP)In 1996, then in 1998, the VA-TAP evaluatedthe experience of the Veterans Health Admini-stration (VHA) with PET centres and con-ducted a systematic review of the PET litera-ture [Flynn and Adams, 1996; Adams andFlynn, 1998]. The first assessment reported: 1)the results of systematic reviews of the clinicalapplications of FDG-PET for certain types ofcancer (head and neck cancer, the staging oflung cancer, the evaluation of solitary pulmo-nary nodules, breast cancer, colorectal cancer),and for Alzheimer's disease, two important dis-eases in veterans; and 2) results of surveys andaudits of VHA PET centres concerning the useof PET, the centres' operations, and researchactivities. The VA-TAP concluded that theVHA should maximize its involvement ratherthan set up other PET centres (implementationof a multicentre VHA PET registry, support forprospective research etc.). In 1998, the VA-TAP reevaluated the VHA's PET centres usingdata from the PET registries and conducted asystematic review of the PET literature pub-lished between September 1996 and December1998 for the above-mentioned cancers andAlzheimer's disease.

A5.4 INTERNATIONAL NETWORK OF

AGENCIES FOR HEALTH TECHNOLOGY

ASSESSMENT

Given the increasing interest in PET world-wide, the INAHTA produced, in 1999, a reporton the use of PET [Adams et al., 1999] and its



coverage in various countries and a synthesisof the assessments of PET conducted by INA-HTA members and three private American or-ganizations. The report examines all PET sys-tems (full-ring conventional scanners, newpartial-ring scanners and digital SPECT scan-ners modified for positron emission imaging).In order to document the use of PET and itscoverage, the OSTEBA conducted a surveyamong 12 INAHTA members (excluding thosein the United States) and 8 participants in theHTA Europe project that are not members ofthe INAHTA, in order to obtain their annualestimates of PET research and clinical applica-tions at private and public facilities and to ob-tain information on the reimbursement of PETin 1997. In 1999, the OSTEBA conducted an-other survey among 31 INAHTA members onthe availability of public coverage of clinicaluses of PET between July 1, 1998 and June 31,1999. A different approach had to be used forthe United States because of the heterogeneityof American health-care systems.

Each INAHTA member was surveyed to obtaincomplete evaluations of the clinical use ofPET. Three private American organizationsalso took part in the project: the Blue CrossBlue Shield Association Technology Evalua-tion Center, the Emergency Care and ResearchInstitute (ECRI) and HAYES Inc.

A5.5 BLUE CROSS BLUE SHIELD ASSO-

CIATION TECHNOLOGY EVALUATION

CENTER

The Blue Cross Blue Shield Association Tech-nology Evaluation Center has published as-sessments of FDG-PET for certain applica-tions, including colorectal cancer, lymphoma,melanoma, pancreatic cancer, lung cancer,head and neck cancer, epilepsy and coronaryartery disease. Their reports evaluate the avail-able evidence in order to determine if the useof FDG-PET in managing patients with these

types of cancer can improve health outcomes.To this end, the authors examine the subjectaccording to the following criteria:

1) The technology must have been approvedby government regulatory agencies (e.g.,the FDA).

2) The scientific evidence must permit con-clusions concerning the technology's ef-fects on health outcomes.

3) The technology must improve the finalhealth outcome.

4) The technology must be as beneficial asthe recognized alternatives.

5) The improvement must be easy to repro-duce outside the study.

A5.6 MEDICARE SERVICES ADVISORY

COMMITTEE (AUSTRALIA)In 2000, the MSAC published a reviewaimed at determining the conditions in whichcoverage of PET in Australia's public health-care system should be supported of the avail-able evidence on the safety, efficacy and cost-effectiveness of PET. To this end, a committeeconsisting of experts in the fields of nuclearmedicine, surgical oncology and cardiologywas formed to evaluate the evidence in the lit-erature and to provide expert opinions. Thestudy inclusion criteria published by the VA-TAP were used, together with criteria for pa-tient selection and for blinded comparisons,among others.

A5.7 HEALTH CARE FINANCING AD-

MINISTRATION (UNITED STATES)

In July 2000, the HCFA received, from Drs.Michael Phelps and Sam Gambhir, a requestfor acceptance of global coverage of FDG-PET. A list of 22 diseases was included withthe request, and, because of the large amount ofevidence submitted by the PET community, the



HCFA asked the Agency for Health Researchand Quality (AHRQ) for assistance. TheAHRQ had the PET submission checked by theEvidence-based Practice Centre (EPC), whichconcluded that the PET coverage request hadnot been submitted in accordance with thestandards for systematic reviews, but that itwas a large compilation of papers on the sub-ject. Questions were raised as to the studiesthat should not have been included with the re-quest and as to various errors identified in thedata taken from the studies for the purpose ofpreparing the summary tables. The HCFA de-termined that separate systematic reviews ofthe literature on FDG-PET were necessary inorder to issue an appropriate coverage policy.

The HCFA therefore added to this coverage re-quest PET assessments published in 2000 bythe Blue Cross Blue Shield Association, theCommonwealth Review of Positron EmissionTomography and other analyses concerninglung and esophageal cancer from the request-ers' submission. The following diseases wereidentified for the use of FDG-PET, based onthese assessments and the analysis of the pa-pers: lung cancer, colorectal cancer, lym-phoma, esophageal cancer, melanoma, headand neck cancer, coronary artery disease andepilepsy.

A5.8 COMITÉ D’ÉVALUATION ET DE

DIFFUSION DES INNOVATIONS

TECHNOLOGIQUES (CÉDIT)

In February 1999, CÉDIT published a progressreport on its positron emission tomographyproject [CÉDIT: Baffert et al., 1999], withguidelines for strategic medical decision-making in the context of Assistance-publique -Hôpitaux de Paris (AP-HP). The guidelines donot constitute a validated assessment outsidethis context. In October 1997, CÉDIT recom-mended to the AP-HP the creation of a PET

centre for managing cancer patients. In 1998, aseries of steps were undertaken to set up a PETcentre. Patients management protocols will beimplemented for the following uses: lung can-cer, gastrointestinal cancers, lymphoma andENT cancers. CÉDIT is planning medico-economic studies in the areas where the litera-ture was insufficient and a clinical researchprogram for uses such as bile duct cancer,melanoma and pediatric cancers. A scientificcommittee was formed and charged withevaluating the patient management protocolsand the scientific protocols and with imple-menting them. CÉDIT also plans to set up anorganizing committee to supervise the centre'soperating conditions. After an evaluation pe-riod, their observations will be used to workout the details for opening centres at other hos-pitals [CÉDIT, 1999].

A5.9 NATIONAL HEALTH SERVICES -RESEARCH & DEVELOPMENT (UNITED

KINGDOM)

As part of the NHS R&D's technology assess-ment program, Robert and Milne's recommen-dations [1999] concerning the research priori-ties in the assessment of positron emission to-mography were published. The specific objec-tives of the 3-month project were to review theknowledge regarding the clinical applicationsof PET (oncology, cardiology and neuropsy-chiatry) by consulting the literature on the topicand to identify, by means of a 3-step Delphistudy, the key issues for evaluative research onPET in the United Kingdom. The results of asystematic review published by the VA-TAP inthe United States in 1966 were used as a start-ing point for the literature review and were up-dated by means of searches in the MEDLINEand Cochrane databases. The results of thisanalysis provide an up-to-date overview of thepotential clinical applications of PET in theNHS.



A5.10 ANDALUSIA HEALTH TECHNOLOGY

ASSESSMENT AGENCY (AHTAA)

In February 2000, the AHTAA published a re-port intended as a synthesis of studies of the ef-ficacy and effectiveness of positron emissiontomography. The report is based on systematicreviews and assessment reports. The authorsdid a search in the INAHTA's databases, theCochrane Library and the NHS-CRD economicevaluation databases, with emphasis on the re-cent publication of a joint INAHTA project[1999], the result of a collaborative effort be-tween various agencies in the network that fo-cused on a synthesis of previous works. One ofthese, a recent (1998) systematic review by theDepartment of Veterans Affairs - TechnologyAssessment Program, was considered to be ofgood methodological quality. It was decided toconduct a search in the primary sources onlyfor the uses that had not been covered by theprevious reviews and those for which it wasthought necessary to have better information,such as the use of PET in melanoma and lym-phoma.

The report is centred on decision making inAndalusia's health-care system with regard tothis technology, which is in its initial phase andwhich will require a more thorough examina-tion in the future.


Appendix 6: Search, selection and evaluation of studies113

Appendix 6

Search, Selection and Evaluation of Studies



APPENDIX 6: SEARCH, SELECTION AND EVALUATION OF STUDIES

A6.1 SEARCH

Language was not an exclusion factor, al-though only articles with an abstract in Englishor French were selected automatically. Articlesin Spanish or German without an abstract inEnglish or French were selected for future con-sideration, if warranted.

Retrospective literature search

PubMed (1999-2000), Cochrane and Inter-net

Descriptors: tomography, emission-computedfluorodeoxyglucose F18 tomography, emission-computed, single-photon: Expressions: PET,SPECT (tomography and emission computed).

The descriptors in PubMed were generatedfrom many different references, which weregrouped together for certain types of studies(review, meta-analysis, etc.). The expressionsinitially used were matched with expressionsconcerning cost and efficacy (cost OR costs ORcompar* OR efficien* OR "comparativestudy"[MeSH Terms] OR "cost benefit analy-sis"[MeSH Terms] OR "efficiency"[MeSHTerms] OR "costs and cost analysis"[MeSHTerms]).

In addition to the MEDLINE searches, a fewsearches were done in Embase and Cancerlit.Also, more selective identification wasachieved, thanks to the assistance of the advi-sory committee's members, who informed us orobtained relevant publications. The referencesidentified were grouped together in databasescreated with ProCite Version 5 for Windows.

Profiles (December 2000 - February 2001)

1. Current Contents (Clinical Medicine): PETor FLUORODEOXYGLUCOSE or (TO-MOGRAPHY and EMISSION) or SPECT orF-18.

2. PubMed (Index Medicus): ("tomography,emission computed"[MeSH Terms] OR "to-mography, emission computed, single pho-ton"[MeSH Terms] OR (Tomography ANDemission computed) ) OR "positron emissiontomography"[Title Word] OR "pet"[TitleWord]OR "spect"[Title Word] OR"fdg"[Title Word][ALL].

3. Cochrane: Descriptors equivalent oridentical to those mentioned above wereused to query the databases available onCD-ROM or on the Internet.



A6.2 SELECTION

Table 17: Study selection criteria (inclusion and exclusion)TYPE

OF PUBLI-CATION

SIZE OF

STUDY

POPULATION

STUDIES

INCLUDED

STUDIES EXCLUDED STUDY DESIGN AND

METHODOLOGY

COMMENTS

CRITERIA Language* ≥ 10 humansubjects (noanimal studies)with the diseaseof interest

Studies usingt h e r a d io-pharmaceuti-cal2-[18F]fluoro-2-D-glucose(FDG)

Duplicate studies or studiesreplaced by other, more re-cent studies (at the samehierarchical level andwhose objectives were thesame) by the same institu-tion.

Studies not presentingenough information to as-sess the comparability ofthe cases and controlgroups, or the details of theimaging protocol, to de-termine if the analysis ofthe PET data was donevisually or quantitatively,or to determine the type ofquantitative analysis per-formed.

Patient selection criteriaclearly stated.

All consecutive patientseligible, based on the in-clusion criteria, were in-cluded.

Independent, blinded com-parisons conducted with areference standard.

PET scan results had no in-fluence on the decision toperform or not perform thereference standard.

The methods for perform-ing the test are described inenough detail for replica-tion.

• Language was not an exclusion factor, although only articles with an abstract in English or French were automatically selected. Articles in Spanishand German without an abstract in English or French were selected for future consideration, if warranted.

A6.3 EVALUATION OF DIAGNOSTIC TESTS

The problem of evaluating diagnostic tests, es-pecially in medical imaging, raises a number ofissues that should be addressed at the outset.

The assessment of a new diagnostic tool shouldbe sufficiently detailed to determine the extentto which and at what cost it permits a more ac-curate diagnosis or more accurate staging ofthe disease, improves patient management andclinical outcomes, and provides information orresults comparable to those of a reference stan-dard while at the time reducing costs and in-conveniences (see Fryback and Thombury'smodel, 1991, which is cited by many evaluat-ors).

Formal protocols for evaluating medical im-aging modalities are generally limited to com-

paring the new modalities’ potential accuracy(sensitivity and specificity) with that of con-ventional signal detection-type modalities[Swets and Pickett, 1982]. A discriminationprocess is begun in standardized conditions,once the diagnosis is made, in order to measurethe accuracy with which the various readersdetect the presence or absence of the diseasefrom the images generated by the modalitiesbeing compared. The rigorous but often arti-ficial selection of cases of the “disease" andof "absence of the disease" is made independ-ently of the imaging modalities being studied.

These protocols, like the "culture" of the scien-tific basis for the assessment and comparisonof the performance of real-time imaging tests,are neither very well developed nor standard-ized. As a result, the research syntheses arebased on individual studies of varying quality



and with vague inclusion criteria, the resultsapply only to a single imaging modality, andthe selection techniques make comparisons dif-ficult. Although precautions have been ex-pressed by a number of authors [Begg, 1983;1986; 1987; 1988a; 1988b; 1989; Feinstein,1989; and McMaster University Group:http://hiru.mcmaster.ca/default.htm], the qual-ity of assessment reports on diagnostic tests ismarkedly lower than that of assessment reportson therapeutic modalities. Furthermore, Sackett[1991], in his book on clinical epidemiology,and the McMaster University Group, in severalarticles on the critical analysis of published pa-pers, had attempted to make reader-evaluatorsaware of this, but the first checklists for authorsof assessments of diagnostic tests have barelyjust been developed [STARD, 2001].

The first prospective comparisons of diagnosticperformance based on clinical readings of di-agnostic images, which were conducted in thelate 1980s, indicated beforehand the proceduresto be used to arrive at a valid result [Neal,1995, 1996; Gatsonis, 1995; Baum et al.,1995]. Initially, the prospective clinical studieswere conducted by ad hoc groups in universityradiology departments. It was not until 1998that the U.S. National Cancer Institute fundedthe first network of such departments for pro-

spectively evaluating new imaging technolo-gies in an expeditious and rigorous manner[Hillman et al., 1999: American College of Ra-diology Imaging Network (ACRIN) http:www.acrin.org/].

The complete assessment of given diagnostictest’s performance requires at least one or, ide-ally, several pairs of sensitivity and specificitymeasures. In the past, few studies reportedcomplete ROC (receiver operating characteris-tic) curves or simultaneous comparisons. Suchweaknesses mean that the accuracy figures re-ported in a study are a function not only of theimaging technique in question, but also of test"positivity" criteria (subjective), of the casemix and of selective checks. Meta-analyses[Irwig et al., 1995; Walter et al., 1999] of sev-eral studies did not fully take the effect of thesethree additional, nonstandardized factors intoaccount. Thus, the comparison of diagnostictest performance is much more prone to distor-tions than that of simultaneous randomized,comparative trials of different treatments.

For these reasons, no assessment including aformal meta-analysis of individual studies hasso far been attempted, except to indicate theobserved sensitivity and specificity rates. Thisreport follows suit.



Table 18: Methodological quality of diagnostic accuracy and diagnostic thinking efficacystudies

A Studies with broad generalizability to a variety of patients and no significant flaws in research methods.

- ≥ 35 patients with the disease of interest and ≥ 35 patients without the disease (since this number yields 95% Cls whoselower limit excludes 0.90 if Se =1).

- Patients drawn from a clinically relevant sample (not filtered to include only severe disease) whose clinical symptomsare completely described.

- Diagnoses defined by appropriate reference standard.

- PET (positron emission tomography) studies technically of high quality and evaluated independently of the reference di-agnosis.

B Studies with a narrower spectrum of generalizability and with only a few methodological flaws that are well de-scribed (and whose impact on the conclusions can be assessed).

- ≥ 35 cases with and without the disease of interest.

- More limited spectrum of patients, typically reflecting referral bias of university centres (more severe illness).

- Free of other methodological flaws that promote interaction between test results and disease determination.

- Prospective studies still required.

C Studies with several methodological flaws.

- Studies with small sample size.

- Incomplete reporting.

- Retrospective studies of diagnostic accuracy.

D Studies with multiple methodological flaws.

- No adequate reference standard for diagnosis.

- Test results and determination of final diagnosis not independent.

- Source of patient cohort could not be determined or was obviously influenced by the test results (workup bias).

- Opinions without substantiating data.

Source: Flynn and Adams, 1996.



Table 19: Assessing the quality of individual studies:Classifications of study designs and levels of evidence

(when high-quality meta-analysis/overviews are not available)

I Randomized trials with a low false-positive (alpha) and a low false-negative (beta) error rate (high power)

- Positive trial with statistically significant treatment effect (low alpha error rate).

- Negative trial that was large enough to exclude the possibility of a clinically important benefit (low beta error rate/highpower, i.e. had a narrow confidence interval around the treatment effect, the lower end of which was greater than theminimum clinically important benefit).

- Meta-analysis can be used to generate a pooled estimate of treatment efficacy across all high-quality, relevant studiesand can reveal any inconsistencies in results.

II Randomized trials with a high false-positive (alpha) and/or high false negative (beta) error rate (low power)

- Trials with interesting positive trend that is not statistically significant (high alpha error rate).

- Negative trials but possibility of a clinically important benefit (high beta error rate/low power, i.e., very wide confidenceintervals around the treatment effect).

- Small positive trials with wide confidence intervals around the treatment effect, making it difficult to accurately assessthe magnitude of the effect.

- When Level II studies are pooled (through quantitative meta-analysis), the aggregate effects may provide level I evi-dence.

III Nonrandomized, concurrent cohort comparisons between contemporaneous patients who did and did not(through refusal, noncompliance, contraindication, local practice, oversight, etc.) receive treatment.

- Results subject to biases.

- Level III data can be subjected to meta-analysis, but the result would not shift these data to another level and is not usu-ally recommended.

IV Nonrandomized, historical, cohort comparisons between currently followed patients who received treatment(as a result of local policy) and former patients (from the same institution or from the literature) who did not(since, at another time or in another institution, different treatment policies prevailed).

- Results subject to biases, including those that result from inappropriate comparisons over time andspace.

V Case series without control subjects.

- May contain useful information about clinical course and prognosis but can only hint at efficacy.

Source: Flynn and Adams, 1996.


Appendix 7: Positions according to the clinical use of PET121

Appendix 7

Positions of Current and Potential Users, Assessment Agenciesand Reimbursement Organizations According to the Clinical

Use of PET



APPENDIX 7: POSITIONS OF CURRENT AND POTENTIAL USERS, AS-SESSMENT AGENCIES AND REIMBURSEMENT ORGANIZATIONS AC-

CORDING TO THE CLINICAL USE OF PET

The assessment procedure chosen by AÉTMISled to a rearrangement of the data. The reportssubmitted between June and September 2000 tothe FMSQ's Positron Technology Committee(PTC) served as the starting point. The reportby Ontario's Council of Medical Imaging(1999) was included with these reports to con-stitute the group called "current and potentialusers". For Québec, the active participation ofthe team from CHUS, which already has acomplete PET unit (cyclotron and scanner), inthe drafting of the report by the Association desmédecins spécialistes en médecine nucléaire duQuébec (AMSMNQ) explains the term "cur-rent" in the name of this first group.

The second group consists of the main organi-zations (public and private) that make PETcoverage policies. In its latest report (Decem-ber 2000), the HCFA responds to a re-quest submitted by the ICCPET (renamed theAcademy of Molecular Imaging on November29, 2000) for global coverage of PET in on-cology, but does not provide a detailed accountof the request. The HCFA's recommendations,published on December 15, 2000, will none-

theless have a determining influence in theUnited States. The Australian committee[MSAC: Ghersi et al., 2000] explains the basisof its conclusions, and its report will serve asthe main comparator. In addition, and as theonly example of a private insurance company,the position of the Blue Cross and Blue ShieldAssociation was included in this group.

A third group consists of conclusions and rec-ommendations taken from reports published byassessment agencies. Reports from other agen-cies are described in Appendix 5. Their conclu-sions or recommendations are not stated heresystematically, unless they are noteworthy.

Upon examining the decision-making proc-esses leading to the coverage of PET scans, onecan situate the positions of reimbursement or-ganizations halfway between the structured re-ports of assessment agencies and the expecta-tions expressed by current or potential users.The coverage of PET scans could, in fact, serveas an indicator of the acceptance of this tech-nology by various health-care systems.



A7.1 ONCOLOGY

A7.1.1 Lung cancer

Table 20: Positions – Lung Cancer

CURRENT AND POTENTIAL USERS POLICYMAKERS → COVERAGE ASSESSMENT AGENCIES

FMSQ

(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC

(AUSTR.)2000

BLUE

CROSSBLUE

SHIELD

2000

MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

INITIALEVALUATION OFTHE PULMONARY

NODULE

� � � � � � � � � � �

INITIAL STAGINGOF NSCLC

NM NM NM NM � �

(stagingand prog-

nosis)

� � NM � �

DETECTING

METASTASES:

MEDIASTINAL

STAGING

�

(normalCT scan)

� � � �

(+ con-current

CT)

� NM � � � NM

DETECTINGDISTANT ME-TASTASES

�

(exceptfor brainmetasta-

ses)

NM �

(before tho-racotomy)

NM NM � NM � NM � NM

MONITORING

RESPONSE TOTHERAPY

�

(radio-therapy,chemo-therapy,surgery)

�

(che-motherapyand ra-dio-ther-apy)

NM �

(chemo-therapy)

� � NM NM NM �

(+ prog-nosis ofthera-peutic re-sponse)

NM

DETECTING

RECURRENCE ORRESIDUALTUMORS

� � �

(when con-ventionalimaging re-sults areequivocal)

� �

(restag-ing)

� � � NM � NM

� Clinical use recognized; � Clinical use not recognized; � Potential use (further research necessary); NM: Not mentioned.

Current and potential usersThe members of the FMSQ's PTC and of theCouncil of Medical Imaging (Ontario) considerthat PET offers several clinical advantages over

conventional diagnostic tests. Accordingly,PET is useful in the clinical situations encoun-tered most often when managing a patient witha pulmonary nodule or proven lung cancer: 1)



the initial investigation of the pulmonary nod-ule; 2) the evaluation of the mediastinum whena diagnosis of neoplasia is made; 3) theevaluation of the usual metastatic sites (stagingof distant metastases) when a diagnosis of neo-plasia is made; and 4) the reevaluation of thechest when recurrent disease is suspected [La-casse, 2000; CMI: Beanlands et al., 1999].

According to the AMSMNQ, the systematicuse of PET during preoperative evaluationswould lead to the identification of patients withinoperable lung cancer and would result in asubstantial reduction in hospital costs byavoiding surgery that would prove traumatic orunnecessary to the patient. The use of PET inevaluating pulmonary nodules (instead oftransthoracic biopsies and surgical excision)would also reduce costs. PET performs betterthan conventional investigational tools (e.g.,CT-guided transthoracic biopsy) and does notcause any complications that require hospitali-zation [AMSMNQ, 2000].

For the ARQ, the role of PET is now well-established in lung cancer. It can help distin-guish between a benign nodule and a malignantone, completes the staging and contributes tobetter patient selection for surgery. It also per-mits a better follow-up after chemotherapy andradiation therapy [ARQ, 2000].

The clinical uses of PET mentioned by theCMI are the same as the preceding ones. Itadds that studies specifically concerning thecosts associated with PET in lung cancershould be carried out. Even though there are nosuch studies, it is estimated that PET can gen-erate several million dollars in savings a year inlung cancer for Ontario [CMI: Beanlands et al.,1999].

Assessment agenciesThe INAHTA and the VA-TAP recognizePET's potential in the evaluation of non-small-cell lung cancer. However, their conclusions

differ with regard to the staging of this type ofcancer. The INAHTA recognizes PET's poten-tial for this application, but the VA-TAP con-cludes that the evidence does not support thisuse. Both organizations call attention to theimportant need to conduct further research byway of rigorous studies comparing PET withgamma coincidence cameras, partial-ring PETscanners and conventional PET with alternativestrategies [Flynn and Adams, 1996; Adams andFlynn, 1998].

The MHTAC recognizes that numerous studieshave shown that FDG-PET has superior effi-cacy to other imaging techniques, especiallycomputed tomography, in differentiating ma-lignant pulmonary lesions from benign pulmo-nary lesions. It concludes that FDG-PET is use-ful for diagnosing lung cancer (primary or me-tastatic disease) and for detecting recurrenceafter treatment.

Coverage policymakersPET has been covered by Medicare since 1998for the diagnosis of pulmonary tumors, theevaluation of solitary pulmonary nodules andthe staging of non-small-cell lung cancer. InDecember 2000, the HCFA concluded thatthere was enough evidence to support PET forthe detection of residual or recurrent tumors(restaging). However, certain conditions mustfirst be met. For the evaluation of a pulmonarynodule, a CT scan must have shown the pres-ence of a primary tumor; the concurrent resultsof conventional imaging must be included inthe request for reimbursement; and, when PETis combined with computed tomography andperformed serially, it is not reimbursed withinthe 90 days following negative PET findings.As for the initial staging of NSCLC, a pathol-ogy report indicating the presence of a primarytumor must be submitted, and the PET resultsfor the entire body must be coordinated withthe results of a concurrent CT scan and of afollow-up by lymph node biopsy (the lymphnode biopsy is not reimbursed if the CT and



PET scan findings are negative. It is reim-bursed only in the context of a positive CTscan and a positive PET scan, a negative CTscan and a positive PET scan or a positive CTscan and a negative PET scan).

The Blue Cross Blue and Shield AssociationTechnology Evaluation Center (BCBSA-TEC)determined, in its qualitative review, that FDG-PET meets its methodological criteria for itstwo main uses in lung cancer (the staging ofnon-small-cell lung cancer and diagnosing asolitary pulmonary nodule), provided that thePET findings can result in a change in how thepatient is managed [Adams et al., 1999].

The MSAC (Australia) considers that PET ismore accurate than the other, conventional im-aging techniques in detecting lung cancer me-tastases before surgery and recognizes its in-creased efficacy when combined with conven-tional imaging. Despite the evidence of supe-rior efficacy to that of conventional imaging,the MSAC notes that it is difficult to quantifythis improvement because of the variable qual-ity of the studies reviewed, for although it isclear that PET can provide useful prognosticinformation that can lead to a reevaluation oftreatment in patients with non-small-cell lungcancer, there are no data for assessing the im-pact of this reevaluation on the final outcome[MSAC: Ghersi et al. 2000].



A7.1.2 Colorectal cancer

Table 21: Positions – Colorectal CancerCURRENT AND POTENTIAL USERS POLICYMAKERS → COVERAGE ASSESSMENT AGENCIES

FMSQ

(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC

(AUSTR.)2000

BLUE

CROSSBLUE

SHIELD

2000

MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

DIAGNOSIS

DETECTION OF THEPRIMARY TUMOR

NM�

NM

NM

NM

NM

NM�

NM

NM

NM

NM

NM

NM

�

NMNM NM

NM

NM

NM

DETECTING ME-TASTASES IN THE

CONTEXT OFRECURRENCE:

HEPATIC

EXTRAHEPATIC

�

�

�

NM

NM

NM

NM

NM

NM

�

NM

NM

�

�

�

�

�

�

�

�

�

�

NM

NM

NM

�

NM

�

�

�

NM

NM

NM

ASSESSING OPER-

ABILITY

�

(by de-tecting

metasta-ses)

NM � NM NM � �

(duringstaging +postop.

follow-up)

� � � NM

EVALUATING REGIONALLYMPH NODES

� NM NM NM NM NM NM NM NM NM NM

EVALUATING RESPONSE

TO THERAPY

NM NM NM NM NM NM NM � � NM NM

DETECTING RE-CURRENCE

�

(+ CTscan, if

necessary)

�

(syst.screen-

ing, sup-plmented

by CTand/orMRI)

� � �

(+ CT scan,if neces-

sary)

� NM � � � �

(as an addi-tion to con-ventionalimaging)

DIFFERENTIATING A

POSTOPERATIVE SCARFROM OPERABLERECURRENCE

NM NM � NM � � � NM � NM NM

DETERMINING THE

LOCATION OF RE-CURRENT TUMORS INTHE CONTEXT OF A

RISING CEA LEVEL

�

(when re-sults ofmedicalimagingare nega-tive)

NM � NM �

(an ele-vated CEAshould notbe the onlyfactor forevaluatingrecurrence)

NM NM NM � NM NM


Current and potential usersBoth for the members of the FMSQ's PTC andthe authors of the CMI's position paper, therole of PET in detecting the primary tumor in

colon cancer is not defined, with colonoscopyremaining the procedure of choice for evaluat-ing the colon and performing biopsies of suspi-cious lesions. For the preoperative evaluation



of regional lymph nodes, the sensitivity of PET(as well as that of computed tomography) is toolow. It is for the detection of hepatic metastasesthat PET is most useful in colon cancer, as itenables one to resect these metastasesor to avoid major surgery in the presence ofgeneralized disease. As for detecting hepatic orextrahepatic recurrence and differentiating scartissue from local recurrence, PET has beenshown to be more sensitive than computed to-mography. However, it should be noted thatPET can yield false-positive results during thesix months following radiation therapy, be-cause of the inflammatory reaction [CMI:Beanlands et al., 1999; Laplante, 2000; Lé-tourneau, 2000]. In the ARQ’s opinion, PET isthe best technique for detecting recurrences ofcolon cancer. It thus supports the use of PETfor systematically screening for recurrent dis-ease, with the addition of computed tomogra-phy or magnetic resonance imaging in order tobetter define the extent of the recurrence andguide the choice of treatment. This would re-duce the costs associated with patient follow-up [ARQ, 2000]. The AMSMNQ states thatPET is generally not indicated for systemati-cally monitoring cancer, but that in the contextof an elevated CEA and of clinical suspicion ofrecurrent colorectal cancer, PET proves quiteuseful for determining the location of the recur-rence, for assessing its extent and for differen-tiating a recurrent tumor from postoperativescar or for assessing the tumor’s resectability.The CMI (Ontario) adds that there are notenough data to determine the cost-effectivenessof PET for colon cancer in Ontario.

Coverage policymakersSince 1999, Medicare (U.S.) has been coveringPET for determining the location of suspectedrecurrences of colorectal tumors in the contextof an elevated CEA, in the following condi-tions: the request for reimbursement mustdocument the previously treated colorectalcancer and the results of a concurrent CT scan(or other diagnostic technique). In 2002, the

HCFA reevaluated its position and extendedthe indication by lifting the restriction of anelevated CEA and now provides coverage fordiagnosis and staging. It recommended cover-age of PET when used to differentiate postop-erative scars from recurrent colorectal tumorsand for detecting hepatic and extrahepatic me-tastases during the initial staging of colorectalcancer before any decision has been made as tohow the patient will be managed [Tunis et al.,2000].

The MSAC recognizes that PET's sensitivity issuperior to that of computed tomography indetecting local recurrence and hepatic metasta-ses but also states that it is difficult to differen-tiate between recurrence and postirradiation in-flammation. It notes that little is known abouthow outcomes are affected by changing man-agement in patients who have abnormalitiesthat are detectable only on PET. It concludesthat the increased sensitivity of PET in detect-ing extrahepatic metastases has a major poten-tial impact on the decision to perform or not toperform surgery, but that the effects of this de-cision on survival or quality of life have stillnot been assessed.

The Blue Cross and Blue Shield Association(BCBSA) considers that the results are compa-rable from one study to another and that the fi-nal outcome is definitely improved through theuse of PET in detecting metastases and assess-ing tumor resectability during initial staging orpostoperative follow-up. As for differentiatingbetween postoperative scar and recurrent tu-mor, the association does not support the use ofPET.

Assessment agenciesLike most other organizations, the VA-TAPrecognizes PET's potential in colorectal cancerbut notes that the evidence is insufficient tosupport its use in managing patients with thistype of cancer [Flynn and Adams, 1996; Ad-ams and Flynn, 1998].



The MHTAC believes that if other studies canconfirm the current findings, FDG-PET couldprove to be a valuable tool for diagnosing, pre-operative staging, monitoring response to ther-

apy and monitoring for recurrent disease. Thepotential role of PET when combined withconventional imaging to confirm the presenceof recurrence after treatment is recognized inthe INAHTA's report.

A7.1.3 MelanomaTable 22: Positions - Melanoma


FMSQ(PTC)2000

ARQ

2000AMSMNQ

2000CMI

1999

HCFA (US)2000

MSAC(AUSTR.)

2000

BCBSA2000 MHTAC

2000AHTAA VA-TAP

1998INAHTA

1999

DIAGNOSIS NM NM NM NM � NM NM NM NM �

INITIAL STAGING NM � NM � � NM � NM NM �

DETECTING ME-

TASTASES

WHEN A DIAGNOSIS OF

AN ADVANCED LESION ISMADE (CLARK III ANDIV)

EXTRANODAL ME-TASTASES

DURING POST-TREATMENT FOLLOW-UP

LYMPH NODE ME-TASTASES

�

NM

NM

�

�

NM

NM

NM

NM

NM

�

�

NM

NM

NM

NM

NM

NM

NM

�

�

NM

NM

NM

�

�

NM

NM

NM

�

�

NM

�

�

�

NM

NM

NM

NM

NM

�

�

NM

NM

�

NM

NM

NM

NM

NM

NM

NM

NM

NM

NM

POSTOPERATIVEEVALUATION

NM NM NM NM NM � NM NM NM NM NM

EVALUATING RE-

CURRENCE

NM NM � NM �

(as analterna-tive to agallium

scan

NM � NM � NM NM


Current and potential usersComputed tomography and magnetic resonanceimaging are not sensitive or specific in deter-mining the location of melanoma metastases.FDG-PET is highly sensitive. The evaluationof the sentinel lymph node by PET is, however,debated, with preference given to sentinellymph node biopsy as a more sensitive andmore specific method [Létourneau, 2000; Lap-lante, 2000]. The Council of Medical Imaging

(Ontario) also mentions this debate, explainingthat sentinel lymph node biopsy may initiallyseem less expensive than PET. However, anAmerican analysis showed that sentinel lymphnode scintigraphy actually increases the costsassociated with staging melanoma, given thehigh cost of operating rooms and because it re-quires several services, including nuclearmedicine, surgery, and pathology personnel[CMI: Beanlands et al. 1999].



For melanoma, the AMSMNQ supports PETonly for evaluating metastases when a diagno-sis of an advanced lesion is made and forevaluating recurrent disease. The ARQ doesnot explicitly express its opinion regarding theuse PET in melanoma. However, it does brieflymention its potential for becoming the test ofchoice for staging this disease.

Assessment agenciesDespite the keen interest in using PET in mela-noma in several American health-care systems,the INAHTA and MHTAC both note that theliterature reviews do not enable them to clearlyestablish PET's role in the management ofmelanoma. The VA-TAP does not mentionmelanoma in its analysis.

Coverage policymakersSince 1999, the HCFA has recommended cov-erage of PET for evaluating recurrent mela-noma, prior to resection, in the following con-ditions: 1) a diagnosis of melanoma; 2) inclu-sion of findings of concurrent conventional im-aging in the claim; 3) a PET scan must be usedas an alternative to a gallium scan; 4) limitationon use: coverage for PET scans is allowed nosooner than 50 days after the last PET scan or agallium scan; and 5) full-body PET scans arecovered only once during a 12-month period,

unless there is a documented medical need todetermine the location of a recurrent tumorduring this period. In its December 2000 re-port, the HCFA recommended that Medicareadd coverage of PET for the evaluation of me-tastatic lesions during staging. It does not sup-port coverage of PET for the evaluation oflymph nodes [Tunis et al., 2000].

The MSAC (Australia) concludes that PET ismore accurate than conventional imaging indetecting metastatic lesions, but that it isclearly less effective than sentinel lymph nodebiopsy for diagnosing micrometastatic disease.PET's role in managing patients with earlymetastatic disease is not clear. Long-term, ran-domized, comparative studies would need to beconducted before any conclusions regardingthis matter can be drawn.

The BCBSA concludes that the results are con-cordant from one study to another. It statesthat, in the evaluation of patients with mela-noma, outcomes would undoubtedly be im-proved with the use of PET to detect extranodalmetastases during initial staging or the post-treatment follow-up. However, it does not sup-port the use of PET for detecting lymph nodemetastases in patients who are candidates for asentinel lymph node biopsy.

A7.1.4 Head and neck cancer



Table 23: Positions - Head and Neck Cancer


FMSQ

(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC

(AUSTR.)2000

BCBSA

2000 MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

IDENTIFYING ANUNKNOWING PRIMARYTUMOR WITH ME-

TASTASES

�

(Cervicalmetasta-

sis)

NM NM NM �

(Cervicalmetasta-sis: whenthe pri-

mary tu-mor can-

not beidentifiedwith con-ventionalimaging)

NM � NM � � NM

EVALUATING THEPRIMARY TUMOR

� NM NM NM NM NM NM �

(PETsuperiorto MRI

butequiva-lent toCT)

� � NM

PREOPERATIVESTAGING

NM NM NM � NM NM NM NM � NM NM

EVALUATINGREGIONAL LYMPH

NODES

� NM NM NM �

(initialstaging ofcervicallymph

nodes inthe con-text of

metasta-ses)

NM � NM � NM NM

EVALUATING RESPONSE

TO THERAPY

� NM NM NM NM NM NM NM � � NM

DETECTING POST-TREATMENT RE-CURRENCE OR RE-

SIDUAL TUMORS

� � NM � � NM � � �

(caseswith

negativefindings

onCT/MRI)

� NM

DISTINGUISHINGBETWEEN POSTOP-

ERATIVE SCAR ANDPERSISTENT TUMOR

NM NM NM � NM NM NM NM � NM NM


Current and potential usersThe members of the FMSQ's PTC believe thatPET does not offer any advantage over con-ventional imaging with regard to evaluating the

primary tumor in head and neck cancer. How-ever, it would seem to be of benefit in patientswith a metastatic cervical tumor of unknownorigin, but further studies are necessary to draw



such a conclusion. The use of PET to evaluateregional lymph nodes and response to treat-ment remains to be clarified. However, PET isuseful for identifying posttreatment recur-rences, with better sensitivity and specificitythan computed tomography and magnetic reso-nance imaging [Laplante, 2000; Létourneau,2000]

The CMI (Ontario) does not mention metastatictumors of unknown origin but considers thatPET is a useful tool for the preoperative stag-ing of head and neck cancer and for detectingposttreatment recurrence. PET is ideal for dif-ferentiating between postoperative scars andresidual tumors. The CMI also mentions asmall number of studies which have shown thatPET is superior to MRI in detecting recurrentENT tumors.

The ARQ indicates that studies are underwayfor the purpose of determining the role of PETin ENT tumors and that it could potentially be-come the test of choice for detecting recur-rences of such neoplasms.

Assessment agenciesThe MHTAC recognizes that PET’s efficacy issuperior to that of magnetic resonance imaging.However, for head and neck cancer, computedtomography is equivalent to PET.

Despite its assessment of PET's potential fordetermining the location of the primary tumor,evaluating response to therapy or differentiat-ing between postirradiation inflammation and

posttreatment recurrence, the VA-TAP main-tains that the evidence is insufficient to supportthe use of PET in patients with head and neckcancer. Blinded, prospective, comparativestudies are necessary to assess the impact thatPET would have in such patients [Flynn andAdams, 1996; Adams and Flynn, 1998].

Coverage policymakersThe HCFA maintains that the evidence is suffi-cient for the coverage of PET costs in cases ofcervical metastases if conventional imaging hasfailed to detect the primary tumor. In caseswhere PET detects a tumor confirmed by bi-opsy to be the primary tumor, management ofthe patient would be modified, which would re-sult in a decrease in the morbidity associatedwith unnecessary surgical treatment or radia-tion therapy. However, the HCFA states that,despite the evidence of beneficial effects asso-ciated with modifying the management of thesepatients, the impact of this modification on thefinal outcome has not yet been determined.Furthermore, the HCFA considers that PETshould be reimbursed for the initial staging ofcervical lymph nodes when there are metasta-ses and for the detection of recurrent or resid-ual tumors. Coverage therefore applies to thediagnosis and staging of head and neck cancer(except for tumors of the central nervous sys-tem and thyroid) [Tunis et al., 2000].

The MSAC (Australia) makes no mention ofhead and neck cancer.



A7.1.5 Lymphoma

Table 24: Positions - LymphomaCURRENT AND POTENTIAL USERS POLICYMAKERS → COVERAGE ASSESSMENT AGENCIES

FMSQ (PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC

(AUSTR.)2000

BLUE CROSS

BLUESHIELD2000

MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

STAGING � � NM � � NM � � � NM NM

RESPONSE TOTHERAPY NM � NM � NM NM NM � NM NM NM

POSTTREATMENTFOLLOW-UP � NM NM NM NM NM � NM NM NM NM�Clinical use recognized; �Clinical use not recognized; � Potential use (further research necessary); NM: Not mentioned.

Current and potential usersIn lymphoma, PET has comparable sensitivityto that of computed tomography but has supe-rior specificity. Furthermore, computed tomo-graphy yields a high number of false-positiveresults. The advantages of PET are a one-dayprocedure with high resolution, better dosime-try, less intestinal activity and the fact that onecan perform a quantitative analysis. In addition,FDG-PET can be used to evaluate posttreat-ment residual lesions with great accuracy[Laplante, 2000].

The CMI (Ontario) states that staging lym-phoma usually involves multiple procedures,including bone marrow aspiration and a biopsy,a CT scan of the abdomen, pelvis and chest,gallium imaging and possibly a bone scan, allof which involve costs in excess of $3,000.These numerous procedures can be replacedwith a single PET scan. Furthermore, PET of-ten reveals a more advanced stage of diseaseand results in more aggressive management.PET is as useful for evaluating response totherapy as it is for staging. The CMI mentionsan American study [Young, 1997] in which,with the use of PET before treatment and againafter two rounds of chemotherapy for evaluat-ing the intermediate response to therapy, themortality rate due to lymphoma decreased byone half compared to the mean national rate.

The CMI also cites a European group whichmaintains that the therapeutic response can beevaluated by PET after only seven days oftreatment. The CMI supports the use of FDG-PET as the main tool for staging all types oflymphoma and for evaluating response to ther-apy.

The ARQ cites the same studies as the CMI,and its position is similar, although less ex-plicit. The AMSMNQ also supports the use ofPET for initial staging and for differentiatingbetween a residual lymphoma and a fibroticmass.

Assessment agenciesThe MHTAC states that studies comparing 18F-FDG-PET with alternative techniques obtainsuperior results with PET compared to com-puted tomography, digital 99mTc-MIBI SPECT,and 111In-somatostatin scintigraphy for detect-ing treated and untreated lymphoma, but thatthe evidence is limited to a few small trials. Forthis reason, the MHTAC cannot draw any con-clusions at this time with regard to the efficacyof PET in evaluating malignant lymphoma.

The VA-TAP does not mention lymphoma inits analysis, and the INAHTA only notes thatno review based on the data gathered from



1977 to 1988 was able to determine the role ofPET in this application.

Coverage policymakersIn 1999, the HCFA recommended coverage ofPET for diagnosing and staging lymphoma(Hodgkin's and non-Hodgkin's), with the fol-lowing stipulations: a diagnosis of lymphomamust have been made; the findings of concur-rent conventional imaging must be includedwith the claim; the PET scan must be used asan alternative to a gallium scan; there is alimitation on use, i.e., reimbursement for PET

scans is allowed no sooner than 50 days afterthe last PET scan or gallium scan; and full-body PET scans are covered only once during a12-month period, unless there is a documentedmedical need to determine the location of a re-current tumor during this period. This positionwas supported by the BCBSA, which statesthat there is enough evidence to support the useof PET in this application [Tunis et al., 2000].

The MSAC (Australia) does not mention lym-phoma.

A7.1.6 Breast cancer

Table 25: Positions – Breast CancerCURRENT AND POTENTIAL USERS POLICYMAKERS → COVERAGE ASSESSMENT AGENCIES

FMSQ(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA(US)2000

MSAC(AUSTR.)

2000

BLUE CROSS

BLUESHIELD2000

MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

DETECTING THEPRIMARY TUMOR

(SPECIFIC PATIENTS)

�

(as anaddi-

tion tomam-

mogra-phy)

NM � � NM NM NM NM � � NM

STAGING � � � � � NM � � � �

(recur-rent tu-mors)

NM

DETECTING LYMPH NODEMETASTASES

�

or�

�

(+ internalmammary

chains)

�

(as a sub-stitute foraxillary

dissection)

NM NM NM NM � � � NM

DETECTING DISTANTMETASTASES

� NM NM NM NM NM NM NM � NM NM

RESPONSE TO THERAPY � � NM � NM NM � � NM � NM

DETECTING RE-CURRENCE

NM � NM NM NM NM NM � � NM

�Clinical use recognized; �Clinical use not recognized; � Potential use (further research necessary); NM: Not mentioned.

Current and potential usersThe members of the FMSQ's PTC consider thatPET's role in detecting the primary breast tu-mor is often limited to specific patients, such asthose with dense breasts or fibrocystic disease,those who have undergone a first biopsy, sur-

gery or radiation therapy, and those with breastimplants, or when the mammography findingsare equivocal. FDG-PET is inferior to search-ing for sentinel lymph nodes and to biopsy indetecting metastatic axillary lymph nodes be-cause it does not detect micrometastases, but it



is useful for evaluating an axillary mass sus-pected of being breast cancer and can even ob-viate the need for axillary dissection. It is supe-rior to mapping in detecting osteolytic bonemetastases, although it may miss certain os-teoblastic lesions. PET could also prove usefulin evaluating response to therapy [Laplante,2000; Létourneau, 2000].

The CMI's thinking is along the same lines. Itnotes that PET plays little or no role in the di-agnosis of breast cancer, with the possible ex-ception of patients with breast implants andpatients at high risk, such as those whosebreasts are too dense for mammography. Itstates that PET plays a key role in initial stag-ing and evaluating response to therapy.

The ARQ does not mention the detection of theprimary tumor in specific patients. It does,however, mention the important role that PETplays in initial staging, evaluating response totherapy and detecting axillary lymph node me-tastases. It attaches special importance to thedetection of metastases in the internal mam-mary chains, since early detection and thetreatment of these metastases improves sur-vival in these patients.

The AMSMNQ supports the following uses ofPET in breast cancer: detecting recurrence,evaluating mammary masses, and evaluatinglymph node metastases as a substitute for axil-lary dissection.

Assessment agenciesThe MHTAC concludes that the preliminarydata seem to favour the use of FDG-PET fordifferentiating benign tumors from malignanttumors during the initial staging of breast can-cer and for evaluating axillary lymph node in-

volvement. It notes, however, that there are toofew data on the utility of PET in evaluating theresponse to breast cancer treatment, but that,despite the paucity of available data, FDG-PETor 11C-MET PET could be useful in this appli-cation by showing a response to therapy earlierthan conventional imaging. However, the sam-ples were small. Further research is necessaryto confirm the efficacy of PET in imagingbreast cancer.

The VA-TAP mentions a number of potentialuses of PET for imaging breast cancer: thenonsurgical evaluation of breast disease; stag-ing recurrent disease; quantifying the tumorglycolytic rate as a prognostic factor; monitor-ing response to therapy; selecting patients foraxillary dissection and preoperative therapy;screening in certain subgroups of women (e.g.,those with breast implants, prior radiation ther-apy, multiple breast masses and negative bi-opsy results, or severely fibrocystic breasts).Despite PET's high potential in breast cancer,prospective, blinded, comparative studies arenecessary to better define its role in this appli-cation in relation to other imaging modalities[Flynn and Adams, 1996; Adams and Flynn,1998].

The INAHTA states that no review has accu-rately determined the role of PET in patientswith breast cancer [Adams et al., 1999].

Coverage policymakersFor this application, the HCFA referred thecoverage decision to the MCAC DiagnosticImaging Panel and will generate internally anew request for a national coverage decision.

The MSAC (Australia) does not mention breastcancer.

A7.1.7 Prostate cancer

Table 26: Positions – Prostate Cancer




FMSQ(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC(AUSTR.)

2000

BLUECROSS

BLUE SHIELD

2000

MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

� NM NM NM NM NM � � NM NM NM�Clinical use recognized; �Clinical use not recognized; � Potential use (further research necessary); NM: Not mentioned.

Current and potential usersThe members of the FMSQ's PTC believe thatFDG has a limited ability to detect primarycarcinoma of the prostate and to differentiatebetween a malignant tumor and benignprostatic hyperplasia. They recognize the utilityof FDG-PET in investigating recurrent carci-noma of the prostate. Other studies are neces-sary [Laplante, 2000].

The ARQ, AMSMNQ and CMI do not mentionprostate cancer.

Assessment agenciesThe MHTAC states that, although 18F-FDG hasbeen used in certain cases of prostate cancer,radiotracers other than 18F-FDG may be moreeffective. The data are presently insufficient todraw any conclusions concerning the role ofPET in prostate cancer. The INAHTA and VA-TAP do not mention this type of cancer.

Coverage policymakersPET is not covered for prostate cancer in theUnited States or Australia.



A7.2 NEUROLOGY

A7.2.1 Dementia and Alzheimer's disease

Table 27: Positions – Dementia and Alzheimer’s DiseaseCURRENT AND POTENTIAL USERS POLICYMAKERS → COVERAGE ASSESSMENT AGENCIES

FMSQ(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC(AUSTR.)

2000

BLUE CROSSBLUE SHIELD

2000VA-TAP

1998AHTAA INAHTA

1999

ALZHEIMER’S DISEASE � NM � � � NM ��/� �/� �/�

DEMENTIA � NM � � � NM � NM NM NM�Clinical use recognized; �Clinical use not recognized; � Potential use (further research necessary); NM: Not mentioned.

Current and potential usersThe AMSMNQ states that FDG-PET can beused to evaluate patients with dementia ormemory loss (Alzheimer's disease, Parkinson'sdisease, etc.), although there are few treatmentsfor these diseases. PET can be used to confirmthe degenerative process even before the onsetof conclusive clinical symptoms. Early diagno-sis enables patients and their families to planthe patient's environment and to provide thenecessary home maintenance resources.

The members of the FMSQ's PTC committeenote that, because of the differences betweenthe impairments in Alzheimer's disease andvascular dementia, PET can be used to targetareas of metabolic dysfunction, confirm the di-agnosis, rule out a differential diagnosis of de-pression and evaluate therapy [Carmant, 2000].

The CMI (Ontario) states that there are nononinvasive diagnostic tests to confirm a diag-nosis of Alzheimer’s disease before an autopsy.The FDG-PET has been proposed as a diag-nostic tool, but the CMI concludes that its util-ity has not been clearly determined. It alsopoints out that, once effective treatments havebeen developed for this disease and for otherdiseases associated with dementia and memoryloss, there will ensue a significant increase in

the number of requests for brain PET scans.

The ARQ does not mention the use of PET indementia or memory loss.

Assessment agenciesThe INAHTA states that the main role of PETin Alzheimer's disease would be the differentialdiagnosis vis-à-vis other diseases that are treat-able or reversible. Furthermore, although thereis still no treatment for Alzheimer's disease,psychosocial techniques and pharmacologictreatments for slowing the progression of thedisease are now available and can improvethese patients' quality of life. The assessmentscompiled by the INAHTA confirm that the ac-curacy of PET in the differential diagnosis ofAlzheimer's disease is comparable or superiorto that with other imaging techniques (e.g., CT,MRI, digital SPECT and EEG) but that itsquality is nonetheless low. The value attachedto a diagnosis of Alzheimer's disease in termsof patient management or the improvement inclinical outcomes has not been studied. This iswhy PET's potential in the diagnosis of thisdisease should be considered in light of thefacts that there is no treatment for the diseaseand that other amply documented diagnostictechniques are similar to PET from the stand-point of efficacy [Adams et al., 1999].



According to the AHTAA, although PET mayprove useful in contributing to the diagnosis ofAlzheimer's disease, the current lack of cleartherapeutic options for improving the prognosisand the lack of suitable scientific evidenceconcerning the value of PET prevent its use inclinical practice. A highly accurate diagnostictest should be demonstrated first in the contextof epidemiological research and in the evalua-tion of potential therapies. Furthermore, the useof PET should be clearly situated in relation toall other available diagnostic tests (clinical,epidemiological and genetic) for this disease.

The VA-TAP states that the existing evidenceargues against the clinical use of PET for diag-

nosing Alzheimer's disease until more effectivetreatments and risk modification interventionsare validated and until reliable predictive val-ues are obtained from an ongoing Europeanmulticentre PET study [Flynn and Adams,1996; Adams and Flynn, 1998].

Coverage policymakersFor this application, the HCFA referred thecoverage decision to the MCAC DiagnosticImaging Panel and will generate internally anew request for a national coverage decision.

The MSAC does not mention Alzheimer's dis-ease.

A7.2.2 Refractory epilepsy

Table 28: Positions – Refractory EpilepsyCURRENT AND POTENTIAL USERS POLICYMAKERS → COVERAGE ASSESSMENT AGENCIES

FMSQ

(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA (US)2000

MSAC

(AUSTR.)2000

BCBSA VA-TAP1998

AHTAA INAHTA1999

DETERMINING THELOCATION OF EPI-LEPTOGENIC FOCI

� � � � � � � NM � �/�


Current and potential usersThe members of the FMSQ's PTC consider thatthe use of interictal PET for localizing epilep-togenic foci is now confirmed and widelypracticed at specialized centres [Couillard,2000]. Its advantage over other diagnostic mo-dalities is that it can be performed between sei-zures, unlike digital ictal SPECT, for example,which must be performed several times. Fur-thermore, PET can reveal areas of cortical dys-plasia that cannot be visualized with magneticresonance imaging, delineate the area of epi-leptic dysfunction in conjunction with an EEGand digital SPECT, substitute for preoperativefunctional localization tests, especially in pedi-atric patients, and be used for the preoperativeevaluation. According to the PTC, PET can

also obviate the need for costly investigationswith implanted or deep electrodes [Carmant,2000].

The AMSMNQ supports the use of PET in theevaluation of epilepsy during the interictalphase for isolating an epileptogenic focus witha view to surgery in patients who are poorlycontrolled with medications [AMSMNQ,2000]. It notes that the combined use of PETand MRI permits accurate localization of epi-leptogenic foci and can obviate the need for in-vasive monitoring with deep electrodes, a la-bour-intensive and expensive procedure whoseresults are difficult to interpret [AMSMNQ,2000].



The ARQ and the CMI (Ontario) briefly men-tion PET’s utility in investigating certain epi-leptic patients whose epileptogenic foci cannotbe localized with computed tomography ormagnetic resonance imaging. Apart from PET,the placement of electrodes, a procedure thatrequires brain surgery, is the only means of lo-calizing epileptogenic foci. However, this useis limited to ultraspecialized centres for thetreatment of epilepsy [ARQ, 2000; CMI:Beanlands et al., 1999].

Assessment agenciesThe INAHTA concludes that the quality of theevidence on which the efficacy of interictalPET has been determined for epilepsy is insuf-ficient. The assessments reviewed by the INA-HTA suggest that the diagnostic accuracy ofinterictal FDG-PET is comparable or superiorto that of other methods of localizing epilepto-genic foci, but the INAHTA maintains that theavailable evidence is still insufficient to rec-ommend substituting PET for other, invasivediagnostic modalities and nonexistent for rec-ommending its use in patients with nontempo-ral lobe epilepsy. It recognizes that PET couldbe beneficial in a minority of patients whoseepilepsy is difficult to control, but its impact onthe management of epileptic patients, the final

outcome of the disease and the costs is still notknown [Adams et al., 1999].

The VA-TAP does not mention epilepsy.

Coverage policymakersThe MSAC (Australia) recognizes the use ofPET in the preoperative evaluation of individu-als with medically refractory epilepsy in thecontext of a comprehensive epilepsy program,where the information obtained from a standardassessment, including seizure semiology, EEGand MRI, is inconclusive. However, it notes thelack of information on false-negative resultswith PET (specifically, in patients with nega-tive PET findings who are eligible for surgeryand patients with false-positive findings whoare actually inoperable).

The HCFA recommends coverage of PET forthe preoperative evaluation of patients withmedically refractory epilepsy.

The BCBSA-TEC observes that FDG-PET im-aging for the purpose of localizing epilepto-genic foci and assessing their resectability inpatients with refractory epilepsy meets its ac-ceptance criteria [Adams et al., 1999].



A7.2.3 Brain tumors (glioma)

Table 29: Positions – Neuro-oncology (brain tumors, mainly glioma)


FMSQ

(PTC)2000

ARQ2000

AMSMNQ2000

CMI1999

HCFA

(US)2000

MSAC

(AUSTR.)2000

BCBSA MHTAC2000

AHTAA VA-TAP1998

INAHTA1999

PREOPERATIVEEVALUATION

� NM NM NM NM � � NM NM �

DEFINING TUMORHISTOLOGY

NM NM NM NM NM NM � � NM NM �

GRADING TUMORS NM NM NM NM NM � � NM NM NM NM

PATIENT MAN-AGEMENT

� NM NM NM NM � � � NM NM �

PROGNOSIS NM NM NM NM NM � � NM NM NM NM

DETECTING CNS* AND

NON-CNS METASTASES

NM NM NM NM NM NM � � NM NM �

FOLLOW-UP(DIFFERENTIATINGBETWEEN POST-

OPERATIVE SCAR ANDPOSTTREATMENTRECURRENCE)

� �

(incombi-nationwithMRI)

� �

(incombi-nationwithMRI)

NM � � � � NM �

EVALUATING THEMALIGNANTTRANSFORMATION OF

A LOW-GRADEGLIOMA

NM NM �

(as asubstitutefor sur-gery)

NM NM � � � NM NM �

�Clinical use recognized; �Clinical use not recognized; � Potential use (further research necessary); NM: Not mentioned.• Central nervous system.

Current and potential usersThe members of the FMSQ's PTC support theuse of PET prior to surgery for central nervoussystem (CNS) tumors, specifically, differenti-ating between radionecrosis and tumor recur-rence in cases of glial tumors; determining thelocation of functional areas in the cortex priorto surgery; and a systematic investigation incases of a brain lesion of undetermined nature[Couillard, 2000].

The ARQ and the CMI (Ontario) maintain thatFDG-PET in combination with magnetic reso-nance imaging is the examination of choice,when following patients with a brain tumor, fordifferentiating scar from posttreatment recur-rence. The CMI considers that PET should beused in conjunction with computed tomogra-

phy and magnetic resonance imaging in suchcases.

The AMSMNQ states that PET is effective indifferentiating between a residual or recurrentbrain tumor and radionecrosis and for evaluat-ing the malignant transformation of a low-grade glioma.

Assessment agenciesThe MHTAC states that 18F-FDG-PET has po-tential for brain tumor imaging but that itsclinical application has not yet been estab-lished, since this technique is unable to definebrain tumor histology. In addition, further re-search would be necessary in order to assessthe role of PET in detecting CNS and non-CNSbrain metastasis, differentiating malignant from



nonmalignant lesions, detecting recurrent dis-ease in patients who have undergone intensiveradiation therapy, and evaluating brain tumorsin pediatric patients. Due to the paucity of dataon radiotracers other then 18F-FDG, furtherstudy is required to validate the use of PETwith other radiotracers to evaluate brain tu-mors.

The INAHTA points out the contradictionsbetween the different assessments regardingthis application. The BCBSA concluded thatthe evidence was insufficient to determine theeffect of PET on outcomes in uses concerningbrain tumors; the CAHTA [1993] concludedthat PET's diagnostic performance in differen-tiating radionecrosis from recurrence or a per-sistent tumor is better than that of conventionalimaging techniques; and the AHTAA notedthat, despite its apparent superiority to mag-netic resonance imaging in this application,PET is inferior to digital SPECT. Given thesecontradictions, the INAHTA concludes that theclinical impact of PET has not been docu-mented and that the overall quality of the avail-

able evidence is low. Further research is neces-sary [Adams et al., 1999].

The VA-TAP does not mention brain tumors.

Coverage policymakers

The HCFA does not mention brain tumors.

The MSAC (Australia) maintains that the evi-dence is insufficient to conclude that PET issuperior to digital SPECT in differentiatingbetween radionecrosis and recurrent glioma. Itstates that the information on PET's potential toalter the management of patients with glioma isvaluable, but that there are practically no dataon the impact of this modification on theirmorbidity, mortality or quality of life, althoughit is reasonable to expect improvements inthese parameters. Long-term outcome trialsshould provide this information. As for tumorgrading, evaluating the malignant transforma-tion of glioma and the preoperative evaluation,the MSAC recommends that exhaustive, sys-tematic reviews be undertaken.



A7.3 CARDIOLOGY

Table 30: Positions - CardiologyCURRENT AND POTENTIAL USERS POLICYMAKERS →

COVERAGE

ASSESSMENT

AGENCY

FMSQ(PTC)2000

ARQ2000

AMSMNQ2000

OCMI1999

HCFA (US)2000

MSAC(AUSTR.)

2000INAHTA

1999

MYOCARDIAL VIABILITY

INITIAL EVALUATION OF PATIENT ANDSELECTION FOR BYPASS

� � �

(together witha resting per-fusion study)

� �

(in place ofSPECT or

after aSPECT scanthat is unre-

vealing)

� �

ASSESSING PATIENTS FOR HEART TRANSPLANT � NM NM NM NM NM NM

POSTOPERATIVE EVALUATION NM � NM NM NM NM NM

CORONARY PERFUSION

DIAGNOSIS AND PROGNOSIS OF CORONARYARTERY DISEASE

� � � � � �

(cost-effective-

ness not de-termined)

�

MONITORING RESPONSE TO THERAPY � � NM � NM NM NM

DETECTING DIMINISHED CORONARY RESERVEIN PATIENTS WITH ISCHEMIA THAT IS NOT

ASSOCIATED WITH CORONARY ARTERYDISEASE

� NM �

(in the con-text of

equivocal re-sults with

myocardialperfusion

scintigraphy)

� NM NM NM


Users in Québec, cardiology (expert opinions):Dr. Peter Bogaty, 2000.The tabled report, as it appears from the scien-tific literature that was examined, concludesthat PET has proven to be useful in evaluatingmyocardial viability; selecting, for revasculari-zation, patients with compromised left ven-tricular function; and selecting patients for aheart transplant [Bogaty, 2000 in Comité sur latechnologie du positron, 2000].

The value of PET for studying myocardial per-fusion has been demonstrated in the followinguses: the diagnosis and prognosis of coronaryartery disease, the diagnosis of coronary arterydisease in subsets of patients who are prone tofalse-positive results with conventional nuclear

medicine techniques (obese patients, womenwith atypical symptomatology, etc.), evaluatingthe progression or regression of coronary arterydisease under treatment (e.g., with lipid-lowering agents), and detecting diminishedcoronary reserve in patients with an ischemicsubstrate that is not associated with coronaryartery disease (hypertrophic cardiomyopathy,cardiac syndrome X, etc.), since no other diag-nostic modality is currently able to provide in-formation with as much accuracy.

Thus, there are many recognized and potentialuses of PET. In addition, PET will, in the nextfew years, be used frequently in cardiology re-search. The report calls attention to the veryshort half-lives of the radiotracers used in car-



diology and to the need to install cyclotrons atcardiology centres so as not to compromise theuse of PET in this area.

Users in Québec, radiology (expert opinions):AMSMNQ and ARQ, September 2000.PET should be the best technique for detectingischemic but viable myocardium and evaluat-ing myocardial perfusion, thus permitting bet-ter patient selection for revascularization. PETis considered an imaging modality with an es-tablished role. Steps should therefore be takento ensure its accessibility by the Québec popu-lation, especially in cardiology [AMSMNQ,2000; Bourgouin, 2000].

User outside Québec (expert opinion): the CMI(Ontario), 1999.FDG-PET imaging has proven utility in evalu-ating myocardial viability for:

� selecting patients with reduced cardiacfunction for revascularization; and

� selecting candidates for a heart transplant.

Myocardial perfusion studies with PET haveproven utility in the following applications:

� the diagnosis and prognosis of coronary ar-tery disease.

� the diagnosis of coronary artery disease inthe subset of patients prone to false-positiveresults with conventional nuclear imaging.

� evaluating the progression or regression ofdisease following pharmacologic treatment.

� detecting a decrease in coronary reserve inpatients with ischemic disease that is notdue to coronary atherosclerosis.

Based on the incidence of coronary artery dis-ease and left ventricular dysfunction in Ontarioand on the data for the recognized applicationsof PET, this test would probably be performedin 10,000 patients per year in Ontario, witheach undergoing one or more scans. The CMIrecommends that when deciding on the loca-

tion of PET centres, patients with heart diseasebe considered one of the two most importantpriorities, the other being cancer patients[Beanlands et al., 1999].

Assessment agency: INAHTA, 1999This was a collaborative effort concerning thecurrent PET use and PET coverage policies inmember countries of the INAHTA, and it in-cluded a synthesis of the technology assess-ments of PET by INAHTA members and threeprivate American organizations (Blue Crossand Blue Shield Association TechnologyEvaluation Center, Emergency Care and Re-search Institute, and HAYES, Inc.). The syn-thesis involved 31 assessments from 13 organi-zations [INAHTA: Adams et al., 1999]. Mostof these assessments were systematic, qualita-tive reviews.

The INAHTA's conclusions concerning cardi-ology are as follows:

PET provides better-quality images than con-ventional imaging. The metabolic informationyielded by PET can improve patient selectionfor revascularization and increase the likeli-hood of successful surgery. The use of PETcould reduce costs by avoiding unnecessaryangiography or revascularization in certain pa-tients.

� For myocardial perfusion studies, PET per-forms better, but how much so in relation todigital thallium SPECT is not clear. Theextent of its contribution to patient man-agement is not clear either. PET is more ex-pensive than other noninvasive techniquesand has still not replaced coronary angi-ography for evaluating coronary artery dis-ease. For patients at intermediate risk (25 to50% probability of having a 50% or greaterstenosis in the left main artery or a greaterthan 70% stenosis in another artery), PET isnot a cost-effective alternative to directlyperforming coronary angiography or other



noninvasive tests, such as stress ultrasoundor SPECT. There are not enough retrospec-tive data to determine the cost-effectivenessof PET in diagnosing coronary artery dis-ease.

� For assessing myocardial viability and/orpredicting the risk of cardiac events, mostassessments have shown that PET has com-parable sensitivity and superior specificityto other techniques, although there havebeen few studies and although the method-ology used often lack robustness. As for im-proving the probability of better postrevas-cularization outcomes and the realizableeconomic gains, there are insufficient datato confirm that PET has a favourable cost-effectiveness ratio.

� As for monitoring the effectiveness oftreatment in coronary patients with hyper-tension or cardiomyopathy, the evidence isinsufficient. This application is still beingresearched.

Assessment agency: Alberta Heritage Founda-tion for Medical Research (AHFMR), 1999.The AHFMR is a member assessment agencyof the INAHTA that provides information topolicymakers in the field of health at the local,regional, national and international levels.

The conclusions of the AHFMR's report[Cowley et al., 1999] are as follows:

� The review of the available literature showsthat the role of the different methods for as-sessing myocardial viability in daily clinicalpractice is not clear.

� PET and dobutamine stress echocardiogra-phy provide the same level of diagnostic ef-ficacy, but the evidence is limited.

� There is a small amount of methodologi-cally weak evidence that PET has some pre-dictive value concerning the clinical out-come of patients who undergo this test.

PET is a promising imaging technique, but theevidence of its clinical benefits is still insuffi-cient. The comparison with other techniquesfor assessing viability is limited. In Alberta, theuse of these methods for assessing myocardialviability should be used in prospective studieswith a long-term follow-up.

Coverage policymaker: MSAC, March 2000[Ghersi et al., 2000].The MSAC specifically examined the role ofPET in evaluating coronary artery disease inpatients with left ventricular dysfunction whohave had a previous digital SPECT scanshowing uncertain myocardial viability or thatthe myocardium is not viable.

The report [MSAC: Ghersi et al., 2000] pre-sents the criteria for selecting articles from theliterature, which led to the review of 103 refer-ences. Additionally, 126 abstracts and 33 arti-cles were examined.



It seems that the sensitivity and specificity ofPET are superior to those of single-positronemission computed tomography. AlthoughPET provides better diagnostic accuracy, thisimprovement is difficult to quantify because offlaws in the study designs.

The MSAC concludes that presently, there isnot enough evidence to draw any firm conclu-sions about the clinical efficacy and cost-effectiveness of PET in this application. Inmost cases, PET is added to other diagnosticmodalities. Further assessments of this tech-nology are needed. The committee recom-mends that FDG-PET be covered on an interimbasis for the evaluation of patients withischemic heart disease with left ventricular dys-function who are being considered for revas-cularization and in whom evaluations of myo-cardial viability using conventional techniqueshave yielded negative results. However, thecommittee stipulates that clinical and/or eco-nomic data from MSAC-approved prospec-tively designed studies must be provided to acentral body so as to permit more long-termdecisions to be made about the role of FDG-PET in clinical practice.

Coverage policymaker: HCFA, 2000 [Tunis etal., 2000]The analysis of the problem in cardiology isbased on the work presented in the Australianreport entitled "Commonwealth Review ofPositron Emission Tomography", drafted byGhersi et al. [1999], who reviewed 33 articles

on the efficacy of PET as a diagnostic test.None of the articles met the predefined criteriafor methodological quality. As regards the im-pact of FDG-PET findings on patient health, incases where the SPECT scan is negative andthe PET scan is positive, there are insufficientliterature data to state that a change in patientmanagement would result in improved healthoutcomes. In cases where the SPECT scan ispositive but questions remain as to the appro-priateness of revascularization, PET could be apromising test, based on the literature exam-ined.

Medicare covers rubidium-82 PET for evalu-ating myocardial perfusion at rest or withpharmacological stress in the context of man-aging patients with known or strongly sus-pected coronary artery disease, when PET isused as a replacement for SPECT or when theSPECT findings are equivocal.

Medicare (U.S.) does not cover PET scans forscreening for coronary artery disease, regard-less of the number or importance of the pa-tient’s risk factors.

In December 2000, Medicare also began cov-ering FDG-PET for evaluating myocardial vi-ability following a positive SPECT scan, butwhen there is doubt as to myocardial viabilityif revascularization is being considered.

Coverage for this test in other applications wasreferred for study to an advisory committeeconsisting of nuclear imaging experts.


Appendix 8: Selected studies on PET147

Appendix 8

Selected Studies on PET

APPENDIX 8: SELECTED STUDIES ON PETA8.1 LUNG CANCER

Table 31: Selected studies on PET (oncology)

Study SiteLanguageandsource

Size ofpatientsample

Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods

Patientswith/with-out disease

Sensitivity

Erasmuset al.,2000

Lungcancer

English

AJR

25 patients(18 men and7 women;age: 37 to86 years).

No Visualanalysis ofaxial andcoronalplane im-ages.

PrimaryNSCLCand pleu-ral effu-sion.Retro-spectivedesign:1993-1999.

Alleligi-ble pa-tientsin-cluded.

Not men-tioned.

PETscans as-sessed byexperi-encedobserv-ers, CTscans bythoracicradiolo-gists.

Thoracente-sis or pleu-ral biopsy.Blinded toPET results.

Scannerwith anaxialfieldview of15.2 cm.Imagingof chestand up-per ab-domenper-formed 1hr aftertheadmin. of10 mCiof FDG.Imagesrecon-structedand datacorrectedfor scat-ter andrandomevents.

In detect-ing pleurametasta-ses: PET95% (CI:77-100%)(21/22).

Accuracyof 92%(CI: 74-99)(23/25)

Guptaet al.,2000

Lungcancer

English

CHEST

118 cases,but 54 un-derwentsurgical testand wereincluded inthe analysis(73 men;age: 35-84).

No Visualanalysis inaxial, cor-onal andsagittalviews.Quantitativeanalysis(SUV: stan-dardizeduptakevalue). AnFDG uptakeof 4-5 wasclassified asmalignant,1-3 as be-nign.

Patientswithproven orsuspectedNSCLCwho werecandi-dates forsurgery.Con-secutivecases.

Alleligi-ble pa-tientsinclud-ed, butonly54 ofthe118 pa-tientswere in-cludedin theanaly-sis.

Yes.PETscansread bynuclearphysi-cians, CTscans byradiolo-gists.

Mediastino-scopy orbroncho-scopy fol-lowed bythora-cotomy.

Whole-bodyFFD-PET.Axialview of14.6 cmin 3-4bed posi-tions.Scanningper-formed60 minafter theadmin. of10 mCiof FDG.Patientsfasted forat least4 hrs.Data re-con-structedand par-tial vol-ume cor-rected.

Staging ofmediasti-nal lymphnodes:PET<1 cm:82%(15/17)1-3 cm:100%(32/32)> 3 cm:100%(4/4)

Overall

PET:96%(51/53)

Accuracy94%(158/168)CT: 68%(36/53)Accuracy66%.

Vanuy-tselet al.,2000

Lungcancer

English

RadiationTherapyand On-cology

105 patients(age and sexdistributionnot given).

Yes.Over-lap-pingofcaseswithtwo

Visualanalysis intransaxial,sagittal andcoronalplanes.Quantitative(GTV)

Patientswith op-erableNSCLC(detaileddescrip-tion).Prospec-

All pa-tientsin-cluded, butnot allwere in-

Yes.PETscansread bytwo nu-clearphysi-cians, CT

Mediastino-scopy andthora-cotomy.

CTI/Siemens scan-ner withan axialfieldview of10.1 cm.Fasting

Lymphnodestaging:PET72%(64/89)CT47%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

previ-ouslypub-lishedstud-ies.

analysis. tive de-sign.

in-cludedin theanaly-sis.

scans bya chestradiolo-gist.

for atleast 6hrs.Scanningin twobed posi-tions,with at-tenuationcorrec-tion(max.555 MBqof FDG)60-70min. afteradmin. ofFDG.

(42/89)PET + CT78%(69/89)AccuracyPET: 95%(943/988)CT: 89%(887/988)(p< 0.001)

Higashiet al.,2000

Lungcancer

English

NuclMedCommun

35 patients(age: 46-84,including 14men).

No Visualanalysis andsemi-quantitative(SUV)analysis ofFDG up-take.

Pulmo-nary ade-nocarci-noma(Jan 1995to Nov1998).Retro-spectivedesign.

Alleligi-ble pa-tientswere in-cludedin thestudyandanaly-sis.

Not men-tioned.PETscanswere readby two oftheauthors.

Thora-cotomy

PETcameraPET withan axialfieldview of10 cm.AverageFDGdose:157MBq.Scanningper-formed40 min.afteradmin. ofFDG in 2bed posi-tions.Imagesrecon-structedusingmeasuredattenua-tion.

No data

Marom etal.,1999

Lungcancer

English

Radiol-ogy

100patients,age: 25-83(with 58men).

No Visualanalysis andquantitative(SUV)analysis ofFDG up-take.

Patientswithnewly di-agnosedcancer orsuspectedfrom ra-diologi-cal study(Nov1995 toJuly1997).Retro-spectivedesign,consecu-tive pa-tients.

39 pa-tientsex-cludedforvari-ous rea-sons,suchaspoor-qualityPETscans.

Yes.PET as-sessed bynuclearmedicinephysi-cians, CTby expe-riencedchest ra-diolo-gists.

Needle bi-opsy, medi-astinoscopy,thoraco-scopic bi-opsy. Testnot influ-enced byPET results.

Fastingfor atleast 4hrs be-fore theadmin. of5.365kBq/kgof FDG.Scanningper-formed30 min.after in-jection. 2bed posi-tions,with at-tenuationcorrec-tion. Im-ages re-

Lymphnodes(N3):PET: 92%(22/24)

CT: 25%(6/24)(P= 0.005

Pulmonarymetasta-ses:PET: 94%(17/18)

Accuracy:98%(98/100)



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

con-structed.SUV cal-culatedby di-vidingthe meanactivityin the le-sion bythe doseof FDGadminis-tered perkg ofbodyweightafter at-tenua-tion.

CT: 78%(14/18)

Accuracy:91%(91/100)

Bénard etal.,1999b

Lungcancer

English

J NuclMed

28 patientsfollowed,but analysisconcernedonly 17 ofthem (14men; age:48-78).

Yes.Samepatientpopu-lationas in aprevi-ousstudy.

Visualanalysis andquantitative(SUV)analysis ofFDG uptakein the tu-mor.

Patientswithproven orsuspectedmesothe-lioma(Sep1995 toMay1997).Refer-encegiven fordetailsand de-scriptionof popu-lation.

All 28pa-tientsin-cluded, butthequan-titativeanaly-sis in-volvedonly17 ofthem.

Not men-tioned.

Histology PENN-PET240Hscanner.Fastingfor atleast 4hrs.4.218MBq/kgof FDG,withscanningper-formed60-90min. afterinjection.Attenua-tion cor-rectionand im-age re-construc-tion.

No datagiven.

Vansteenkiste etal.,1998

Lungcancer

English

J ClinOncol

68patients,age: 40-83.Sex distri-bution notgiven.

No?(casesin thisstudyover-lappedwiththosein an-otherstudy).

Visualanalysis oftransaxial,sagittal andcoronal im-ages andquantitative(SUV)analysis ofFDG up-take.

Patientswithproven orsuspectedNSCLCoperableafterstagingwithconven-tionalimaging,includingCT (Sep1995 toJan1997).Prospec-tive de-sign.

Not alleligi-ble pa-tientswereen-rolledin thestudybe-causeofsched-ule con-flictsandlack ofcon-sent.

Yes.PET as-sessed by2 nuclearmedicinephysi-cians, CTby 2chest ra-diolo-gists.

Invasivemediastinalstaging.

PETcamerawith anaxialfieldview of10.1 cm.10 mCiof FDGadmin.afterfastingfor atleast 6hrs.Scanning60 min.afteradmin. ofFDG.Recon-structionand at-

Locallyadvanceddisease(N2/N3):PET + CT93%(26/28)Accuracy:94%(64/68)CT: 75%(21/28)Accuracy:68%(46/68)

Individualmetastaticlymphnode sta-tions:PET + CT89%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

tenuationof im-ages.

(42/47)Accuracy:98%(678/690)CT: 47%(22/47)Accuracy:92%(638/690)

Pieter-man et al.,2000

Lungcancer

English

N Engl JMed

102 patients(including88 men);age: 25-77.

No Visualanalysis ofhot spot.

Patientswith po-tentiallyoperableNSCLCwho werebeingevaluatedwithconven-tionalimaging(Sep1996 toDec1998).Prospec-tive de-sign,consecu-tive pa-tients.

Alleligi-ble pa-tientswere in-cluded.

PET andCT as-sessed ina blindedfashionby 2 ob-servers.

Histopa-thology –nodal ex-aminationwith cervi-cal medi-astinoscopy,parasternalmediasti-notomy andexploratorythora-cotomy.

Scannerwith anaxialfieldview of10.8 cm.Fastingfor atleast 6hrs be-fore theadmin. of370 MBqof FDG.Scanningbegan 90min afterinjection.Imagescorrectedfor at-tenuationand recon-structed.

Mediastinalmetastases: Based on32 patientPET: 91%(CI: 81-100)Accuracy:87% (CI:80-94)

CT: 75%(CI:60-90Accuracy:69%(CI: 60-78)PET + CT94% (CI:86-100)Accuracy:88%(CI: 82-94)

Distantmetastase: Based on17 patientPET: 82%(CI: 64-100)OverallPET: 95%(CI: 88-100)

Barkheetet al.,2000

Lungcancer

English

Clin NuclMed

10 cases.Age and sexdistributionnot given.

No Visualanalysis

Patientswith can-cer(March1996 toJuly1997).Retro-spectivereview.


No men-tion ofthosewhoevaluatedthe im-ages orwhetheror notthey wereblinded.

Histopa-thologicalnodal sam-pling.

Scannerwith anaxialfieldview of16.2 cm.7 to 10mCi ofFDGadmin.after 12-hr fast.Whole-bodyscanning45-60min afterinjection.Images

No data



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

recon-structedusingHannfilter.Not allimagescorrectedfor at-tenua-tion.

Robertset al.,2000

Lungcancer

English

Ann Tho-rac Surg

100 pa-tients; ageand sex dis-tribution notgiven.

No Visualanalysis inaxial, cor-onal andsagittalplanes.Semiquan-titative(SUV)analysis ofFDG up-take.

PatientswithearlyNSCLCreferredfor PETscan andwho hadpatho-logicalconfir-mation(Jan 1995to April1999).Retro-spectivereview.


PET im-ages readby nu-clearphysi-cians butno men-tion ofblinding.

Histopa-thology –mediastino-scopy, me-diasti-notomy orthoracotomy

Scannerwith anaxialfieldview of16.2 cm.Fastingfor atleast 4hrs be-fore theadmin. of0.143mCi/kgof FDG.6 bed po-sitions.Whole-bodyscanning30-45min afterinjection.Imagesrecon-structedwith/without at-tenuationcorrec-tion.

Mediasti-nal me-tastases:

PET87.5%(21/24)

Accuracy90%(90/100)

Hara etal., 2000

Lungcancer

English

J NuclMed

29patients;age: 40-83(including19 men).

No Semiquan-titative(SUV)analysis ofFDG uptakein the tu-mor.

Patientswith bi-opsy-provenNSCLCand me-diastinallymphnodemetasta-ses re-garded asNO, N1or N2 byCT. Ret-rospec-tive de-sign.


Yes. PETimagesread by 2blindedradiolo-gists.

Histopa-thologicalnodal sam-pling.

Scannerwith 6-mm spa-tial resolu-tion. 12-hr over-nightfastingand in-jection of370 MBqof FDG.Scanningfromneck toliver 40min afterinjection.6 bed po-sitions.Attenua-tion cor-rectionby com-bining

Detectionof medi-astinallymphnode me-tastases:

PET-FDG75%(68/90)Accuracy96%(251/261)CT19%(17/90)Accuracy94%(245/261)11C-cholinePET100%(90/90)Accuracy97%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

transmis-sion andemissiondata in acom-puter.

(253/261)

Weberet al.,1999

Lungcancer

English

Eur JNuclMed

27 patients;mean age:62 ± 9 (26men).

No Visualanalysis andsemiquan-titative(SUV)analysis ofFDG uptakein the pri-mary tumorand in-volvedlymphnodes. Im-ages inter-preted oncomputerscreen usinglinear grayscale.

Patientswith lungcancer orinterme-diatepulmo-narynodules.Retro-spectivedesign.


Yes.PETscansread by2 experi-encedobserv-ers, CTby 2 ra-diolo-gists.

Histopa-thology (bi-opsy or tho-racotomy).

Full-ringscannerwith anaxialfieldview of16.2 cm.Admin.of 185-270 MBqof FDGafter4 hrs offasting.PET im-ages re-con-structedwith (ac)andwithout(nac) at-tenuationcorrec-tion us-ing pe-nalizedleast-squarealgo-rithm.

Detectionof lymphnode me-tastases:PET (ac)100%(11/11)

PET (nac)91%(10/11)

CT91%(10/11)

CGC73%(8/11)

Vesselle et al.,2000

Lungcancer

English

ClinicalCancerResearch

39 patients.Age and sexdistributionnot given.

No Quantitativeanalysis oftumor FDGuptake,quantitatedwith themaximumpixel stan-dardizeduptake value(SUV).

Patientswith op-erableNSCLCwho un-derwentresectionor surgi-cal bi-opsy(Feb1998 toJune1999).Prospec-tive de-sign.


No men-tion.

Surgicalstaging(broncho-scopy andmediastino-scopy withor withoutthora-cotomy).Immuno-chemistryfor Ki-67(prolifera-tion indexmarker).Specimenswere re-viewed forcellular dif-ferentiation(poor, mod-erate, good)and tumortype.

Whole-bodyFDG. 12-hr fast.Admin.of 7-11mCi ofFDG.Scanningin thethoracicplane 45min afterinjection.Imagescollectedin 2 di-men-sions.Data re-con-structedand cor-rectedusingstandardfilteredback-projec-tion.

No datagiven.

Berlan- Lung English 50 patients: No Visual Patients All Yes. PET Mediastino- Scanner Nodal



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

gieri etal., 1999

cancerCardio-thoracicSurgery

37 men;age: 41-78.

analysisgrade on a5-pointscale.

suspectedof havingNSCLCwho re-mainedsurgicalcandi-dates af-ter con-ventionalimagingandstaging.Con-secutivecases.

eligi-ble pa-tientswere in-cluded.

assessedby a nu-clearmedicinephysi-cian, CTby ablindedradiolo-gist.

scopy orthora-cotomy.

with anaxialfieldview of10 cm.Admin.of400 MBqof FDG.Thoraxand up-per ab-domenemissionscan ac-quiredafter 45minutes.Attenua-tion cor-rectionmade.Recon-structionwithHanningfilter.

staging:

PET80%(16/20)

CT65%(13/20)

Saunders et al.,1999

Lungcancer

English/

Ann Tho-rac Surg

97 patients:64 men;age: 36-77.

No Visualanalysis inthe tran-saxial, cor-onal andsagittalslices, andquantitative(SUV)analysis oftissue up-take.

Patientswith sus-pected orprovenlung can-cer re-ferred forpossiblesurgery(Nov1992 toJuly1995, andfolloweduntilJuly.1996).

Alleligi-ble pa-tientswere in-cludedin thestudy,butonly84 ofthe 97pa-tientswere in-cludedin theanaly-sis.

Yes. PETassessedby 2 nu-clearmedicinephysi-cians.

Mediastino-scopy orthora-cotomy.

ECAT951/31Rscanner.6-hr fast.Admin.of 350MBq ofFDG.Imagesof thechest ac-quiredafter 81min. At-tenuationcor-rected,and im-ages re-con-structed.

Mediasti-nal stag-ing: (N2 andN3):

PET70.6%(12/17)*

CT20%(3/15)*

Magnaniet al.,1999

Lungcancer

English

J Cardio-vascSurgery

28 patients:26 men;age: 50-75.

No Visualanalysis ofFDG uptakecomparedwith bloodpool activ-ity.

PatientswithprovenNSCLCwaitingfor sur-gery.Retro-spectivedesign.


PET andCT as-sessed byinde-pendentobserv-ers.

Broncho-scopy orneedle bi-opsy as wellas mediasti-noscopy orthora-cotomy.

Whole-bodyscannerwith anaxialview of15.4 cm.10-minemissionscan 50min afterinj. of370 MBqof FDG.Tran-saxial

Mediasti-nal lymphnodestaging:

PET67 %(6/9)CT66%(6/9)PET + CT78%(7/9)



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

imagesrecon-structedwithHannfilter andcorrectedfor at-tenua-tion.

Dhitalet al.,2000

Lungcancer

English

Eur JCardio-thoracicSurg

97 cases,but only 77were in-cluded inthe analysis(53 men;age: 36-77).

Yes Visualanalysis

Patientswith sus-pected orprovenlung can-cer clini-callyconsid-ered op-erable(Nov1992 toJuly1995).Retro-spectivedesign.

Of theinitial97 pa-tients,only77were in-cludedin theanaly-sis.

2 blindednuclearmedicinephysi-cians.

Histologicaldiagnosis bybroncho-scopy,transtho-racic needlebiopsy orthora-cotomy.

ECAT951/31Rscanner.Fastingfor 6 hrs.Admin.of 350MBq ofFDG.Thoracicemissionbegan 81min later.Attenua-tion cor-rected,and im-ages re-con-structed.

No data

A8.2 COLORECTAL CANCER



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

Valk, etal. (1999)

Colo-rectalcancer

English

ArchSurg

115 con-secutive pa-tients for di-agnosis orstaging ofrecurrentcolorectalcancer.

No Visualanalysis inaxial, cor-onal andsagittalplanes.

All pa-tientswith con-firmed orsuspectedrecurrentcolorectalcancer(Oct1992 toMay1996).

21 ptsex-cludeddue tolack ofa crite-rionstan-dard. 6fol-lowedforlessthan 1year, 6lost tofol-low-up, 5died, 4treatedwith-outfurthervali-dation.

CT andPET in-terpretedblindlyand readtogetherby 1 or 2investi-gators.

PET wasperformedafter the CT(between 0-56 days).

Yes.ECATEXACT921scanner.Patientsfasted forat least 4hrs priorto injec-tion ofFDG,5.29MBq/kg(0.14mCi/kg).4 minutesper bedposition.Scanningbegan 30minutesafter in-jection.

With dis-ease:Liver: 57Pelvis: 31Abdomen:28Retroperi-toneum: 12Lungs: 17Other: 12Total: 157

Withoutdisease:Liver: 58Pelvis: 84Abdomen:87Retroperi-toneum:103Lungs: 98Other: 104Total: 534

PETLiver:95%Pelvis:97%Abdomen79%Retrop-eritoneum100%Lungs:94%Other:100%Total:93%

CTLiver:84%Pelvis:68%Abdomen46%Retrop-eritoneum58%Lungs:94%Other:33%Total:



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

69%

Imdahl etal. (2000)

Colorectalcancer

English

Langenbeck’s ArchSurg

71 patients(77 investi-gations)with sus-pected re-currence,metastasesor elevatedCEA.

No By 2 inde-pendent in-vestigators.Lesionsclassified asmalignantby focallyincreaseduptake ex-ceeding thenormal lim-its in the re-spective ar-eas or astandardizeduptake value(SUV) > 4.

Unclear Un-clear

Interpre-tationswereblindedto theclinicaldata andthe re-sults ofthe con-ventionalimaging.Scansread by 2inde-pendentinvesti-gators.

Comparisonwith CT,ultrasono-graphy,MRI andchest x-ray.

> 6 wksafter thelast che-mother-apy ormorethan 3monthsafter thecomple-tion ofradiationtherapy.Patientsfasted for12 hrs.Bladdercatheterwas in-sertedunlesspatientdid notgive con-sent.Sie-mens/CTI ECATEXACT921/31scanner.Scanningdone 90minutesafter in-jection of350 ± 50MBq incubitalvein.

Liver me-tastases in28 patients(43 le-sions).Pulmonarymetastasesin 17/77investiga-tions.

Detectionof localrecur-rence:PET: 92%CT: 88%MRI: 83%Detectionof hepaticmetasta-ses:PET:100%CT: 87%MRI:100%

Zhuanget al.(2000)

Liverme-tasta-sesfromcolo-rectalcancer

English

NuclearMedicineCommu-nications

80 patients No Not explic-itly stated.Assumedvisual.

Con-secutivepatientsrecruitedretrospec-tively.

96 con-secu-tiveFDG-PETpa-tientswithcon-firmedcolo-rectalcan-cer. 16ex-cluded(lackof con-currentana-tomi-calimag-ing,nega-

Notstated.

Results ofPET and CTwere com-pared withthose ofsurgicalpathologyand clinicalfollow-up.

Scanningwith C-PETcamerawithin 8weeks ofconven-tionalimaging.Fastingfor atleast 4hrs priorto injec-tion of2.516MBq/kgof FDG.Imageacquisition began40 minpostinjection. At-tenuationcorrec-tion us-

28 withhepaticmetastases,52 withouthepaticmetastases.

Detectionof hepaticmetasta-ses: 100%

CT: 71.4%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

tivefind-ings,lessthan 6months offol-low-up,deathor lostto fol-low-up).

ing a cae-sium-137pointsource.Imagesrecon-structedusing theorderedsubsets(expecta-tionmaximi-zation).

Staib etal. (2000)

Recurrentcolorectalcancer

English

AmericanJournalof Sur-gery

100 patientswith his-tologicallyconfirmedcolon orrectal can-cer.

No Recon-structed im-ages wereassessedvisuallywithoutquantifica-tion of FDGuptake.

Prospec-tively re-cruitedfrom1994 to1998.

Pa-tientswithuncon-trolleddia-betesoracuteinflam-mationwereex-cluded.

Blindedto thespecificresults ofconven-tionalimagingbut withknowl-edge ofclinicaldiagnosisand indi-cations.

CT, liverultrasound,and carci-noembry-onic antigen(CEA).

SiemensECATEXACTHR+scannerand Sie-mensECAT8/12scanner 1hr afterthe in-jection of408 +10.5(mean +SE) MBqof 18F-FDG. Noattenua-tion cor-rection.

With dis-ease:Liver me-tastases: 33Lung me-tastases: 17Rectal re-currence:20Colon re-currence: 3

Withoutdisease:Liver me-tastases: 67Lung me-tastases: 83Rectal re-currence:35Colon re-currence:42

Detectionof malig-nancy:PET: 98%87 CTscans:91%98 CEAassays:76%Detectionof livermetasta-ses:PET:100%68 ultra-sounds:87%Local re-currence:PET: 96%

Will-komm etal. (2000)

Recur-rentcolo-rectalcancer

English

Journalof Nu-clearMedicine

28 patientswith sus-pected re-currence (15men, 13women).

No 2 nuclearmedicinespecialistsunaware ofthe resultsof the con-ventionalimagingstudies. PETfindingsclassified as“malig-nancy-typical”when anti-body uptakewas mark-edly higherthan liveruptake.

Followedfor 6 to19months.

Inclu-sionandexclu-sioncriterianotclearlystated.

Blindedto the re-sults ofconven-tionalimaging.

Immu-noscintigra-phy (CEA-scan).

Siemens/CTIECATEXACTscanner.12-hr(over-night)fast. Ap-prox.250-370MBq ofFDG in-jectedandflushedwith 20mL ofsaline.Patientsasked todrink 1 Lof waterandvoidedprior toexamina-tion.

With dis-ease:Local re-currence: 9Hepaticmetastases:9

PET:Local re-currence:100%

CEA:Local recurrence:89%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclu-sion

Blindedreading?

Referencetest

PETmethods


Sensitivity

Scansobtained45-60min afterinjection.

A8.3 MELANOMA



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria

Inclusion Blindedreading?

Refer-encetest

PETmethods


Sensitivity

Paquet etal. (2000)

Me-tastaticmela-noma

English

Derma-tology

24 patients(28 assess-ments)

No Visual in-terpretationin coronalsagittal andtransversalplanes.

No Unclear Unclear CT,MRIand ul-tra-sound.Corre-lationwith his-tologi-cal con-firma-tion(n=11)andclinicalfollow-up(n=17).

150-300mBq in-jected ac-cordingto bodyweight,and scanper-formed60-90minutesafter in-jection.UGMPennPET 240H scan-ner.

Not given. Diagnosticaccuracyof 80%.

Eigtvedet al.,.(2000)


English

Eur. J.Nucl.Med.

38 patients,stage II(n=27) or III(n=11).

No Visual in-terpretationby 3 blindedobservers,who re-corded thesites of in-creased up-take.

Yes Consecu-tive pa-tients (Dec1993 toOct 1995)after re-section of amelanoma.

Yes PET re-sultscom-paredwith theresultsof clini-cal ex-amina-tion andof otherimagingmeth-ods (CTof chest

GE Ad-vancePETscanner.Fastingfor atleast 6hrs priorto injec-tion(mean:357MBq).Scanningbegan

Histologi-cal diagno-sis ob-tained in29/38 pa-tients (25malignanttumors).25 patientswith ma-lignant le-sions andinvolve-ment ofregional

PET: Allsites: 97%Intraab-dominalsites:100%Pulmo-nary/intra-thoracicsites:100%Conven-tionalmethods:All sites:



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

or ab-domenif me-tastasesweresus-pectedin thosere-gions).An ultra-soundwas per-formedif theclinicalexamyieldedsuspi-ciousfindingsin theregionsof in-terest.

35-40minuteslater.AfterPET, abiopsy orexcisionof the tis-sues wasdone, ifpossible.

lymphnodes,11 withmalig-nancy inotherlymphnodes,6 withpulmo-nary/intra-thoracic le-sions, and4 with in-tra-abdominallesions.

62%Intraab-dominalsites:100%

Crippa etal. (2000)


English

J. Nucl.Med.

38 patientswith previ-ous historyof cutane-ous mela-noma and acurrentclinical di-agnosis ofnodal me-tastases.

No Visualanalysis by2 blindednuclearphysicians.Images withat least onesite of FDGuptake wereconsideredpositive formelanoma.Retrospec-tive analysisto evaluateeach lymphnode basinand deter-mine thenumber ofsites ofFDG up-take.

Yes Currentclinical di-agnosis(throughphysicalexam, ul-trasound orCT) ofnodal me-tastases.

Compari-son withhistologyresults.

PETscandone 1-3 daysbeforesurgery

GE 4096WBscanners.Injectionof a meandose of496 MBqof FDG.Scanningdone 1-3daysprior tosurgery.Patientsfasted forat least 5hrs.Meanglucoselevel be-fore PETwas 84mg/dL.

With dis-ease: 35lymphnode ba-sins.

Withoutdisease: 19negativefor lymphnode ba-sins.

Lymphnode ba-sins: 95%

Number ometastasesfound(comparedto histol-ogy)<5 mm:23%6-10 mm:83 %11-15mm:100%16-20mm:100%21-25mm:100%> 25 mm:100%Total:66%

Acland etal. (2000)


English

J. Am.Acad.Derma-tol.

54 patientsreferred byphysicianswith variouscriteria forPET (notconsistent).

On-goingstudy.

Visual as-sessment byexperiencednuclearmedicinephysicians(coronal,sagittal andtransaxialplanes).

Retro-spectivefromPET scandatabase.All pa-tientswithhistologi-cal diag-nosis ofmela-nomawho had

No Compari-son withhistologyor clini-cal pro-gressionof thedisease.Furtherstudy todo directcompari-son withsentinel

N/A ECAT951Rwhole-bodyscannerin ex-tended 2-dimen-sionalmode. 6-hr fast.Bloodglucoselevels

With dis-ease:17 patientswith stageI3 patientswith stageII14 patientswith stageIII.

Withoutdisease:

Overall:87%Stage Idisease:50%Stage IIdisease:33%Stage IIIdisease:93%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

under-gonePET.

node bi-opsy.

meas-ured.Scanningbegan 50minutesafter theintrave-nous in-jection of350 MBqof FDG.

27 nega-tive forstage I.6 negativefor stageII.2 negativefor stageIII.

Krug etal. (2000)


English

Acta Ra-diologica

94 patientsexaminedretrospec-tively,

No Evaluatedby a radi-ologist, anuclearmedicinespecialistand a der-matologist.

Yes. Allmela-nomapatientsreferredbetweenJune1995 andMarch1999.

FDG-PETfor malig-nant mela-noma. Oneexclusiondue to hy-perglyce-mia.

Confir-mationwithhistologyor otherimagingresults.

Maxi-mumtime of2 weeksbe-tweenPETandotherexami-nations.

ECATEXACTscanner(Sie-mens/CTI). 300-400 MBqwas in-jected asper eachprotocol.Maxi-mum al-lowabletimebetweenPET andother ex-amina-tions be-ing com-paredwas 2weeks.

N/A Could notbe calcu-lated be-cause thenumberswere toosmall.

Wagneret al.(1999)

Mela-noma

English

J ClinOncol

70 patients(89 lymphnode ba-sins).

No By the samenuclearmedicinespecialist.

Patients(> 18years ofage) withinvasivecutane-ous mela-noma.

Exclusioncriteria:ocular ormucosalmelanoma;clinicalevidenceof regionallymphnode me-tastases, ordistantmetastases;palpablelymphade-nopathy;infectionor inflam-mation inthe re-gionallymphnode ba-sins; priorexcision >4 cm;lymphnode dis-sections,skin grafts,tissuetransfers,or flapsthat can

Blindedinvesti-gator in-terpretedrecon-structedimagesand as-signedeachlymphnode ba-sin at riskfor beingdefinitelypositive,probablypositive,uncer-tain,probablynegative,or defi-nitelynegative.

Preop-erativelym-phaticmap-ping,SNBproce-duresandsurgicalspeci-mens.

Pre-operativewhole-bodyPET.SiemensECAT951/31RPETscanner.Two im-agingprotocols(patients1-24: 10mCi ofFDG in-jected 30min priorto scan;patients25-74: 60min afterinjec-tion).

Results for89 basins.

Sentinelnode bi-opsy fordetectingoccult re-gionallymphnode me-tastases:94.4%PET:16.7%



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

alter thelymphaticdrainagepatternfrom theprimarytumor siteto the re-gionalnode ba-sins; preg-nancy orbreast-feeding;prior ma-lignancy;and allergyto isosul-fan bluedye or toFDG.

Jadvar etal. (2000)

Melanoma

English

Clin. Nu-clearMed

38 patientsstudied ret-rospec-tively.

No Visual (ax-ial, coronaland sagittalplanes) on acomputer bya single, ex-periencedobserveraware ofclinicalhistory andradiologyresults.

All pa-tientswith pri-marycutane-ous mela-noma (≥1 mm indepth)evaluatedbetweenJune1995 andAug1997.

Not stated. CT per-formedwithin 9weeksafterPET withno inter-veningthera-peuticinterven-tion.

CT re-sultsavail-able for21 pa-tients.

ECATEXACTscanner(CTI,Knox-ville,TN). Im-agingwas done40-60min afterthe intra-venousinjectionof 10-15mCi(370-555MBq) ofFDG. 6bed posi-tions.

N/A Not stated

Tyler etal. (2000)

Mela-noma

English

Cancer

95 patients(106 PETscans).

No FDG-PETactivity wasassessed in-dependentlyby 2 experi-enced ob-servers aspositive(activity >background)or negative(activity ≤back-ground). Fi-nal determi-nation byconsensus.

Patientswithclinicallyevidentstage IIIlymphnodeand/or in-transitmela-noma.

Consecu-tive pa-tients at theDuke Uni-versityMedicalCentreMelanomaClinic.

PETstudieswere in-terpretedinde-pendentlyof CTwithoutanyknowl-edge ofclinicalor pa-thologyresults.

CT, bi-opsy ofarea ofin-creaseduptakeon PET.

Fastedfor 4 hrs.GE Ad-vancemedicalscanner.FDGdoserange of10-20mCi.

165 areaswith mela-noma.69 nega-tive areas.

PET:87.3%

Dietleinet al.(1999)

Mela-noma

English

NuclearMedicineCommu-nications

68 patientswith ad-vancedmelanoma.

No PET, ultra-sound andradiologicaldata inter-preted bydifferent in-vestigators.

91 PETscansdone in68 pa-tientsbetweenJune1995 andOct

Not stated. Notstated.

CT No stan-dardprotocolwas es-tablishedfor theextent ofthe im-aging or

N/A FDG de-tectedfewerpulmonaryand he-patic me-tastasesand fewercerebral



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

1997. 15had 2PETscans,and 4 had3.

the tim-ing afterthe in-jection.All camefrom dif-ferent in-stitutions,and allinstitu-tionsusedECATEXACTscanners(Sie-mens/CTI). 300-400 MBqof FDGwas in-jected ineachprotocol.

foci butmorelymphnode andbone me-tastasesthan con-ventionalradiologyor CT.

A8.4 HEAD AND NECK CANCER



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

DiMartinoet al.(2000)

Headandneckcancer

English

Arch Oto-laryngolHeadNeckSurg

50 patientswith pri-mary or re-current headand neckcancer.

No Imagingprocedureswere per-formed 1-20days prior tosurgery.

Nonran-domized,com-parativestudyfrom Oct1, 1997to Nov30, 1998.

Exclusioncriteria:diabetesmellitusand a his-tory ofacute orchronic in-flamma-tory dis-ease.

Notstated.

Ultra-sound,CT,histo-pathol-ogy ofbiopsyspeci-mens.

PET per-formedwith anECATEXACT922/47scannerafter aminimum12-hrfast. 212± 59MBq ofFDG wasadminis-tered in-tra-venously.Scanningbegan45-60min later.

WITHCANCERPET, CT,panendo-scopy:Primary:37Recurrent:8

Ultra-sound:Primary:23Recurrent:6

WITH-OUTCANCER:PET, CT,panendo-scopy:Primary:13Recurrent:5

Ultra-sound:Primary: 4Recurrent:2

PET:Detectionof primarytumor:95%Recurrentcarcino-mas:100%

CT:Detectionof primarytumor:68%Recurrentcarcino-mas: 62%

Panendo-scopy:Detectionof primarytumor:95%Recurrentcarcino-mas:100%

Ultra-sound:Detectionof primarytumor:74%Recurrent



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

carcino-mas: 67%

Lowe etal. (2000)

Recur-rentheadandneckcancer

English

J. Clin.Oncol.

44 patientswith stageIII and IVhead andneck cancer.

No Assumed tobe visual.

Patientswere partofneoadju-vant or-gan-preserva-tionstudy thatincludedchemo-therapy,radio-therapyand sur-gical sal-vage.

Unclear CT doneat 2 and10months.Pathol-ogy re-sults ob-tained forsome.

Physi-calexam(PE)and cor-relativeimaging(CT).

ECAT951/31PETscanner(Siemens). PETdone at 2and 10monthsaftertherapy.All pa-tientsfasted be-fore PETstudies.Glucoselevelsmeas-ured.

16/30 pa-tients hadrecurrenceafter thefirst year.5 weredetected byPET only,4 by PETand con-ventionalimagingonly, 5 byPE andPET, and 2by PE,PET andconven-tional im-aging.

PET:100%PE: 44%Correla-tive im-aging:38%

Lonneux(2000)


English

The La-ryngo-scope

44 patients No Quantita-tive, usingSUV (= ac-tivity in the9 maximalpixels inµCi/mL) ÷(injecteddose ÷ leanbody mass).

Prospec-tive in-clusionof pa-tientswithclinicalmanifes-tations ofrecur-rence(pain,palpablemass,bleeding,dysphonia).

All pa-tients re-cruitedprospec-tively.

CT +MRIdonewithin 2weeks.

None ECATEXACTHR (Sie-mens/CTI). Pa-tientsfasted for6 hoursbeforeinjectionof 185-370MBq.Scanningbegan 45min later.

22/38 withdisease.

PET: 95%CT +MRI: 73%

Jungehul-sing et al.(2000)

Un-knownpri-marytumorwithheadandnecklymphnodemani-festa-tion

English

Oto-laryngolHead andNeckSurgery

27 patientswith noprimary tu-mor foundafter con-ventionaldiagnosticprocedures(from origi-nal 723).

No Visual. PETimages wererecon-structedwith filteredback-projectionand dis-played incoronal,sagittal andtransaxialprojectionswith thecommercialsoftwareMPITool.

27 pa-tientswith noprimaryidenti-fied, May1994 toJuly1998.

All pa-tients di-agnosedwith ma-lignantdisease atan ENTclinic wereeligible(n=723).

PETfindingswere cor-roboratedby fine-needleaspirationcytology,biopsy orsurgery.

N/A Siemens/CTIECATEXACTwhole-bodypositronscanner.Standarddose of370 MBq(10 mCi)was in-jectedafter abloodspecimenobtainedfor bloodglucosedetermi-nation.Patientsfasted forat least 6hrs.Scanningbegan60-90

N/A Could notbe calcu-lated dueto smallsamples.



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

min afterinjection.2 bed po-sitions.

Bohusla-vizki etal. (2000)

Un-knownpri-mary

English


53 patients No Visual in-spection.

Patientswithmetas-tatic cer-vical ade-nopathyor ex-tracervi-cal me-tastaseswere in-cludedbetweenJan 1997and Jan1999 af-ter anextensivebut in-conclu-sive di-agnosticworkup.

53 patientswith me-tastasesfrom un-knownprimarytumor fromJan 1997 toJan 1999.

No.Other re-sultswereused tointerpretthe PETresults.

Clini-cal,surgicalandhisto-patho-logicfindingswereused. Inpatientswith sus-pectedlungtumors,chestCT andsubse-quentbiopsieswereper-formedto evalu-ate thePETfind-ings.

Completepatienthistoryandphysicalexam(includ-ing chestx-ray)were per-formedprior toPET, pa-tientsfasted forat least 6hrs, andscanningbegan 60min afterthe in-jection of370 MBqof FDG,using anECATEXACT47 scan-ner.

With dis-ease: 20(10 lung, 8head andneck, 1breast, 1ileocolonicarea).

PET:37.8% trupositives,22.2%false posi-tives

Perie etal. (2000)

Un-knownpri-mary

English

Ann OtolRhinolLaryngol

4/60 pa-tients withuntreatedhead andnecksquamouscell carci-noma in-cluded in aprospectivestudy ofFDG.

No N/A March toOctober1998

N/A N/A N/A N/A N/A N/A

Lassen etal. (1999)

Un-knownpri-mary

English

EuropeanJournalof Cancer

20 patients No Visual in-terpretationof PET bynuclearmedicinespecialistswith PETexperience.Semiquan-titativeanalysis ofFDG uptakewas not per-formed.

Referredto Co-penhagenUniver-sity Hos-pital inApril1996 toSep1997.

Patientsaged 18 to75 with bi-opsy -provenmetastaticdisease andunknownprimarytumor fol-lowingphysicalexam, x-ray and/orCT androutinelaboratorytests.

At thetime ofthe visualinterpre-tation,correla-tive in-formationon his-tologyand loca-tion ofmetas-tatic le-sions wasavailable.

De-pendedon his-tology.

GE Ad-vancePETscanner.Patientsfasted forat least 6hours. 40min afterinjectionof 350-400 MBqof FDGin a cu-bitalvein. Thepatientemptiedbladder,then wasplaced onthe PET

PET sug-gestedprimarysite in 13patientsand wasconfirmedhistologi-cally (orby theclinicalcourse ofthe dis-ease) in 9.

Data notgiven.



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

scannerbed. Allthe PETscanswere per-formedwithin 4weeksafter ini-tial diag-nosticproce-dures.

Farber etal. (1999)


English

Laryngo-scope

28 patientswho re-ceived ra-diation ther-apy forsquamouscell carci-noma of thehead andneck.

No Qualitativeanalysis ofcorrectedand non-correctedimages, by2 readers.

Unclear Unclear Yes CT andMRI

PENN-PET240HPETscanner.FDG in-jectedinto a pe-ripheralvein at adose of0.114µCi/kg.Patientthenplaced ina lyingpositionand in-structednot totalk for15 minbeforeand 30-45 minafter theinjection.Scanned~60 minafter up-take.

14 withdisease.14 withoutdisease.

PET: 86%CT andMRI: 71%

Tiepolt etal. (2000)

Thy-roidcancer

English

Annals ofNuclearMedicine

31 patients No The PETimages wererecon-structed byfilteredback-projection.Uncorrectedimages wereused forcomparingthe resultsobtainedwith a dedi-cated PETscanner andon a coinci-dence cam-era and readseparatelyby 2 blindedreaders.

Notstated.

Patientswith thy-roid cancerwho hadundergonea thyroi-dectomyand at least2 radioio-dine treat-ments witha positiveI31I whole-body scan(n=22)and/or highthy-roglobulinlevels(n=27).

2 physi-cianswereblindedto the re-sults ofthe coin-cidencegammacamerawhenreadingthe PETscans.

Coinci-dencegammacamera(PETwasactuallythe ref-erencetest).Coinci-denceimagingis acost-effec-tive al-terna-tive.

ECATEXACTHR+ (Sie-mens/CTI, Knox-ville,Tenn.)withBGOdetectors.PET andcoinci-denceimagingwere per-formedon thesameday. Pa-tientsfasted forat least 4

118 lesionsidentifiedin 31 pa-tients.

Coinci-dence im-agingcomparedwith PET:69%Concur-rence was96% in le-sions > 1.5cm and62% inthose be-tween 1and 1.5cm. Le-sions < 1cm couldnot beidentifiedwith thecoinci-dence



Dupli-catestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-encetest

PETmethods


Sensitivity

hrs andhad nor-malbloodglucoselevels.

camera.Identicalstagingobtainedin 84%.

A8.5 LYMPHOMA



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Jerusalem et al.(1999)

Hodg-kin’s andnon-Hodg-kin’s Lym-phoma

English

Blood

54 pa-tients in-cluded(19 HD,35 NHL)

No Qualita-tiveanalysiswithoutcorrec-tion(transver-sal, cor-onal andsagittalplanes).

Unclear.Patientsrecruitedprospec-tivelyfromJune1994 toFeb1998.

Patientswithclinicallyprogres-sive dis-ease afterchemo-therapywere ex-cluded.

No None. CTwas onlyper-formedafter ab-normalPETfindings.

PENN-PET 240-H scan-ner(UGM,Philadel-phia,PA). 6-8mCi ofFDG wasadminis-tered in-tra-venously,andscanningbegan60-90min later.Patientsfasted forat least 6hrs priorto scan-ning.

N/A Positivepredictivevalue ofPET vs.CT (100%(6/60 vs.42%(10/24)).

Buch-mann etal.(2001)

Lym-phoma

English

Cancer

52 pa-tients (27HD, 25NHL)

No Regionsof FDGuptakeclassifiedby site,intensity,size andshape.Any siteexhibit-ing in-creaseduptakewas con-sidered asuspectedlym-phoma.

Con-secutivepatientswith ahistologi-cally con-firmeddiagnosisof un-treatedmalig-nant HD.

Untreateddiseases,withhistologi-callyprovendiagnosisof HD orNHL,Oct 1996to July1998.Diabeticsexcluded.

PETscans in-terpretedby 2 ex-periencednuclearmedicinephysi-cians, CTscans by2 inde-pendentradiolo-gists, allin ablindedfashion.

Refer-ence testsper-formed 4weeksbefore orafter PETand CT.Discrep-anciesbetweenPET andCT veri-fied bybiopsy,MRI orclinicalfollow-upwithinthe fol-lowing 4to 24months.

Patientsfasted forat least12 hrs.Meandoses of390 MBqof 18F-FDG-PET and20 mg offu-rosemideinjectedintrave-nously.Scanningbegan60-90min later.

25 addi-tional le-sions de-tected byPET.Nodal:124 truepositivesidentifiedby PET, 0false posi-tives, 1false nega-tive and655 truenegatives.103 truepositivesby CT, 22false nega-tives, 7false posi-tives and648 truenegatives.Extra-nodal:PET: 24true posi-tives, 349

PET:Nodal:99.2%Extra-nodal:100%Supra-diaphrag-matic:99.1%Sub-diaphrag-matic:100%CT:Nodal:83.2%Extra-nodal:80.8%Supra-diaphrag-matic:80.3%Sub-diaphrag-matic:91.2%



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

true nega-tives, 0false nega-tives, and 3false posi-tives;CT: 20true posi-tives, 348true nega-tives, 4false posi-tives, and 4false nega-tives.

Hueltenschmidt etal.(2001)

Lym-phoma

English

Cancer

81 pa-tients

No Any siteexhibit-ing FDGuptake inthe re-gion ofinterestwas con-sideredrepre-sentativeof lym-phomamanifes-tation.

81 pa-tientswith HDunder-went 106PETstudies.25 initialstagings,63 evalua-tions ofthe re-sponse totherapyand 18suspectedrecur-rences.

Retro-spectiverecruit-ment(Aug1996 toApril2000).

Visualanalysisof PETimagesby 3 ex-periencednuclearmedicinephysi-cianswho hadknowl-edge ofthe pa-tients’clinicalhistory.

Conven-tionalimaging(mainlyCT andsome-timesMRI),biopsyand/or avery de-tailedclinicalfollow-up.

ECATEXACT47 scan-ner (Sie-mens/CTI, Knox-ville,TN). Pa-tientsfasted for12 hrs.Normalbloodglucoselevels con-firmed inall pa-tientsprior toscanning.MeanFDGdose: 370MBq.

63 PETstudies forrestagingin 51 pa-tients. PETscanspositive in21/63cases andnegative in42/63cases.

RestaginggroupPET: 95(95% CI:89-100)Conventionalimaging:95 (95%CI: 89-100)Recurrence group:PET: 91(95%CI:78-100)Conventionalimaging:91 (95%CI:78-100)

Jerusa-lem et al.(2000)

Lym-phoma

English

Haema-tologica

28 pa-tients

No PET im-ages in-terpretedby a phy-sician inthe Divi-sion ofNuclearMedicineand re-viewedby an in-vestiga-tor. Anyregionexhibit-inghigher-than-back-grounduptakeand/orexcretionwas con-sideredpositivefor the

Con-secutivepatientswithhistologi-cally con-firmedNHL andwho weresched-uled forchemo-therapywere in-cludedfromMay1994 toMarch1997.

Not men-tioned.

None.Pre- andpost-treatmentevalua-tion.

Whole-bodyscan witha PENNPET 240-H camera45-90minutesafter theintrave-nous in-jection of200-300MBq of18F-FDG.Patientsfasted forat least 6hrs priorto scan-ning.

5/28 pa-tients withincreaseduptake.PET nega-tive in23/28 pa-tients.

All 5 pa-tients withand 7/21withoutresidualabnormaluptakerelapsedor repro-gressed.



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

presenceof tumor.

Spaepenet al.(2001)

Lym-phoma

English

Jounralof Clini-cal On-cology

93 pa-tients

No Negativedefinedas noevidenceof dis-ease,positiveas a siteof uptakeor anydiffuseregion ofincreasedactivity.

93 con-secutivepatientswithhistologi-cally con-firmedNHL.Retro-spectiveanalysis.

June1995 toSep1993.Con-secutivepatients.

Clinicalinterpre-tation by2 investi-gatorsblindedto theclinicaldata orCT find-ings.

Conven-tional di-agnosticmethodsbeforeand aftertherapy(CT,MRI andbiopsy).

CTI/SiemensECAT931scanner.Patientsfasted forat least 6hrs, andbloodglucoselevelmeasuredprior toscanning.Dose of370-555MBqadminis-tered in-tra-venously.

26 positivescans (per-sistent ab-normaluptake) in14/26 pa-tients.Only PETdetectedpersistentdisease.67 withnegativescans(completeremission).Only 11out of 63relapsed.

Tatsumiet al.(2001)

Lym-phoma

English


30 pa-tients

No Regionsof FDGuptakeclassifiedaccordingto site,intensity,size,shapeand lat-eralasymme-try. Anysite ex-hibitinghigher-than-back-groundFDGuptakeand thatwas notlocated ina physio-logicalregion ofuptakewas con-sideredpositivefor thepresenceof lym-phoma.

Prospec-tive re-cruit-ment.

Untreatedor recur-rent NHLcon-firmed bybiopsy.

All thepatientswererandom-ized, andthe re-sultswere readby 2blindednuclearmedicinephysi-cians.

CT scansobtainedwithinthe 2weeksfollowingFDG-PET.

Whole-bodyscan per-formedwithHead-tomeV/SET2400Wcamera.Fastingfor atleast 4hrs.Scanningbegan 1hr afterthe in-jection of370 MBqof FDG.PETstudywithdual-headgammacamera(hybrid)equippedfor coin-cidenceemissionper-formedon sameday.

206 dis-ease sites.Hybrid anddedicatedscannersdetected159 sitesand 179sites, re-spectively.CT and67Ga: 164.

HybridPET:77.2%PET:86.9%CT and67Ga:79.6%.

A8.6 BREAST CANCER



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Cook andFogel-man(1999)

Breast –skeletalmetasta-ses

English

Seminarsin Nu-clearMedicine

< 10 N/A N/A N/A N/A N/A N/A N/A N/A N/A

Crippa etal. (1997)

Breastcancermetasta-ses

English

Tumori

66 pa-tientswho un-derwentaxillarynode dis-sectionand 16who didnot.

No Increaseduptakewas con-sideredabnormaland in-terpretedas patho-logic.

Notstated.

Notstated.

Not men-tioned.

PET re-sultswerecomparedwith thepathol-ogyfindingsat sur-gery.

PETstudiesper-formed1-7 daysprior tosurgeryusing aGE 4096WB Plusscanner.Scanningbegan45-60min afterinjectionof FDG.

31 withaxillarymetastases,52 withoutaxillarymetastases.

PET in thedetectionof axillarymetasta-ses:84%Palpablenodes vs.non-palpablenodes:92%(11/12) vs79%(15/19)

Cowen(1998)

Breastcancermetasta-ses

French

Can-cer/Radiotherapie

908 pa-tientswho re-ceivedsurgeryand ra-diother-apy butno che-mother-apy.

No No Patientswere di-videdinto twogroups,based onthe tumorresectionmargins.

Patientsselectedretrospec-tively,1980 to1995.

N/A Pathol-ogy.

N/A N/A N/A

Noh et al.(1999)

Breastcancer

English

Eur JSurg

8 patientswithbreastimplants(of 59who un-derwentPET) (6patientshad par-affin im-plants,the other2 siliconeim-plants).

No Stan-dardizeduptakevalue(SUV).Relativeuptakevaluesexpressedas a per-centageof thebaselinePETscan.

All ex-amined ata localhospitaland sentto SeoulNationalUniver-sity Hos-pital

Patientswho un-derwentPETfromJune1995 toNov1997.

PET im-ageswere as-sessedand docu-mentedbeforethe surgi-cal andhistopa-thologyresultswereavailable.

Physicalexam,mam-mogra-phy andtumorhistol-ogy.

ECATEXACT37 scan-ner. 370MBq ofFDG wasinjectedintrave-nously ineach pa-tient 60min priorto scan-ning.

N/A N/A

Crippa etal. (1998)

Breastcancer

English


68 fe-malespatients(age: 29to 84).

No Imageswereconsid-eredpositivein thepresenceof in-creasedlocalizedFDGuptake inrelationto the sur-roundingtissue.Semiqua

Con-secutivepatientssched-uled forsurgeryand axil-larylymphnode dis-section.

All con-secutivepatientsincluded.

PET in-terpreta-tionsblindedto thehistopa-thologyfindingsat sur-gery.

Pathol-ogy re-portswere thebasis forthe finalclassifi-cation ofthe nod-ules.

400 MBqof FDGwas in-jectedinto aveincontralat-eral tothe tumorside, pa-tientsfasted forat least 5hrs andall hadnormalglucoselevels.

27 withaxillarymetastases,45 withoutaxillarymetastases.With N0:10WithoutN0: 26With N1a: 8WithoutN1a: 13With N1b-2:9WithoutN1b-2: 6

PET indetectingaxillarymetasta-ses: 85%Clinicalaxillarystage ofthe N0 pa-tients:70%N1a

patients:85.5%N1b-2

patients:100%



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

ntitativeanalysisof FDGuptakewas per-formedby gener-ating pa-rametricimages ofstan-dardizeduptakevalues(SUVs)in whichthe con-centra-tion ofradioac-tivity wasdividedby theratio oftotal ad-minis-tered ac-tivity tobodyweight.

PET per-formed1-7 daysprior tosurgerywithGE4096WB Plusscanner.

Raylmanet al.(2000)

Breastcancer

English

Med.Phys.

No sub-jects(simu-latedbreasttissue).

N/A N/A N/A N/A N/A N/A N/A N/A N/A

A8.7 PROSTATE CANCER



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitiv-ity

Seltzer,et al.(1999)

Prostatecancer

English

Journalof Urol-ogy

45 pa-tientswith anelevatedPSAwerestudiedfollowingprostatectomy, ra-diationtherapyor cryo-surgery.

No Visualanalysis.PET andmono-clonalantibodyscanwere in-terpreted,on differ-ent days,by con-sensus of2 nuclearmedicinephysi-cians,who werethereforenotblindedto oneanotherbut whowere

Aug1996 toJan 1998

Patientswith anelevatedPSA.

CTreadingswereblindedto the re-sults ofPET andmono-clonalantibodyscan.PET andmono-clonalantibodyscanwereblindedto CT butnot to-tally toeachother.

CT. In-terpretedby geni-tourinaryradiolo-gist.

Patientsfastedprior toscanning.12-15mCi ofFDGgiven toimageglucosemetabo-lism. ASiemensHR 961or 962scannerwas used.

N/A Data notgiven.



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitiv-ity

blindedto the CTfindings.

A8.8 MISCELLANEOUS USES (NOT EXAMINED IN THIS REPORT)



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods

Patientswith/with-out dis-ease

Sensitivity

Lips etal. (2000)

Thyroidcancermetasta-ses

English

Nether-landsJournalof Medi-cine

Case re-ports of 4patients.

No N/A N/A N/A N/A N/A N/A N/A N/A

Barkheetet al.(1999)

Benignesophag-eal dis-ease

English

ClinicalNuclearMedicine

Case re-ports of 3patients.


Flamenet al.(2000)

Esophagealcarcinoma

English

Journalof Clini-cal On-cology

74 pa-tientswith eso-phagealor gastro-esophag-eal carci-noma.

No Blindedvisualanalysisof the tran-saxial,coronaland sag-ittalplanes.High-resolu-tion dis-playmonitor.

All re-ferred forevalua-tion ofthe oper-ability ofan eso-phagealtumor.

Exclu-sion cri-teria:previoustreatmentfor eso-phagealcancer,diabetes,pulmo-nary in-flamma-tory dis-ease orinoper-abilityfor medi-cal rea-sons.

Yes All pa-tients hadan ultra-sound ofthe neck,a bariumx-ray oftheesopha-gus, abroncho-scopy, aspiral CTscan ofthe chestand ab-domenand trans-esophag-eal endo-scopicultra-sound(EUS).

CTI/Siemens931/08/12scanner.Imagetransmis-sion be-fore theinjectionof 6.5MBq/kgof FDG;max. of555MBq.Scanningbegan 60min afterinjection.5 bed po-sitions.

With dis-ease: 34patients,stage IVdisease.

Withoutdisease:40 patients

Detectionof stage IVdisease in74 patientsPET: 74%Ultrasound41%CT: 42%CT + ultrasound: 47%

All com-pared to thdiagnosticstandard:histology olymphnodes withmalignanttumorsPET: 39%Ultrasound63%CT: 22%CT + ultrasound: 54%

Meltzeret al.(2000)

Esophag-eal can-cer

English


47 pa-tients re-ferred forinitialstagingof eso-phagealcancerprior tomini-mally in-vasive

No Visualanalysisby 2 nu-clearmedicinespecial-ists. Di-vergentevalua-tions.

New di-agnosisof eso-phagealcancer,Univer-sity ofPitts-burgh.

PatientsevaluatedbetweenJuly 1995andMarch1998. 67eligible,10 ex-cludedbecauseof ad-

Yes. 20PETstudies inother pa-tientswith none-sophag-eal tho-raciccancerselected

CT dataevaluatedby expe-riencedradiolo-gistsblindedto all theclinicaldata, ex-cept thediagnosis

Patientsfasted forat least 4hrs. In-jection of7 mCi ofFDG.Scanningbegan45-60min later.ECAT

Primarytumor.With: 47Without:20

Nodalstaging:With:+: 35-: 36Without:

PET find-ings:Primarytumor+ equivo-cal find-ings: 87%- equivocafindings:79%Nodalstaging



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

surgicalstaging.

vanceddisease,and 10other ex-cludedfor tech-nical rea-sons.

for inclu-sion intheanalysisof thePETfindings.Readingblindedto clini-cal data.

of eso-phagealcancer.

ARTscanner.

+: 12-: 11

Distantmetasta-ses:With: 10Without:36

+ equivo-cal find-ings: 43%- equivocafindings:39%Distantmetastases+ equivo-cal find-ings: 70%- equivocafindings:30%

Choi etal. (2000)

Esopha-gus(squamous cellcarci-noma)

English


48 pa-tientswith a di-agnosisof his-tologi-callyprovenprimaryesophag-eal carci-noma (45men and3women).

No Visualanalysis.

All con-secutivepatientseligible.

61 con-secutivepatientswith a di-agnosisof his-tologi-callyprovenprimaryesophag-eal carci-noma(Feb1997 toDec1998). 13patientsexcludedbecausethey didnot un-dergo anesophagectomy.

Yes. ThePETfindingswere in-terpreted,by con-sensus,by 2 nu-clearmedicinephysicianblindedto theCT, en-doscopicultra-soundand his-tologyresults.

CT (HiS-peed Ad-vantagescanner).Imagesinter-pretedbeforesurgeryby a ra-diologistblindedto thePET, en-doscopicultra-sound(EUS)and his-tologyresults.EUS per-formed inall but 3of thepatients(couldnot toler-ate theproce-dure) andinter-preted bya gastro-enterolo-gistblindedto thePET, CTand his-tologyresults.EUS in-completein 12/45patients.

AdvancePETscanner(GEMedicalSystems).Scanningbegan 45min afterthe in-jection of370 MBqof FDG.Imagesrecon-structedwithoutattenua-tion cor-rection.

100 lymphnodegroupswith me-tastases.282groupsnegative.

Region (+)chest: 62abdomen:30neck: 8

Region (-)chest: 181abdomen:85neck: 16

PET: 57%(for deter-mining ifthere ismetastasisto individ-ual lymphnodegroups)CT: 18%DifferencebetweenPET andCT:p<0.001

Anatomicaregion ofmetastases(PET):Thoraciclymph nodgroups:66.1%Abdominalymph nodgroups:46.7%Cervicallymph nodgroups:37.5%

Stumpeet al.(2000)

Soft tis-sue andbone in-fections

English

Eur JNuclMed

39 pa-tientswith sus-pectedinfections

No Visualanalysisby 2blindednuclear

Yes Patientsreferredfor evalua-tion of

36 pa-tients hadCT orMRIwithin 3

No GE Ad-vancePETscanner.Patients

With dis-ease:24 in softtissues.16 in

Patientswith softtissue dis-ease: 96%Bone in-



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

(cause ofinfection:patientswere ei-ther nottakingtheir an-tibioticsor not re-spondingto ther-apy).

medicinephysi-cians.

suspectedinfection.

days afterPET, orcompari-son madewithhistologyfindings.

fasted forat least 4hrs andreceivedan intra-venousinjectionof 300-400 MBqof FDGthat hadbeenproducedin-house30-40min ear-lier usinga 17.8-MeV cy-clotron.

bone.40 pooledwith dis-ease.

Withoutdisease:3 in softtissues.9 in bone.12 pooledwithoutdisease

fections:100%Pooled: 9%

Schulteet al.(2000)

Skeletaltumors

English


202 pa-tientswith pri-marybone tu-mors.

No Qualita-tive andsemi-quantita-tiveanalysisof the ar-eas ofelevatedFDGuptake.T/Bs(tumor-to-back-groundratios)obtainedby 2 phy-siciansblindedto eachother andto theclinicaldata andaveragedfor sta-tisticalanalysis.

Ongoingprospec-tive studythat be-gan in1993.

Patientswith radio-graphicevidenceof an ac-tive be-nign le-sion, ag-gressivebenignlesion ormalig-nant pri-marybone neo-plasm.

Blindedto clini-cal data.

Inci-sional,exci-sional orneedlebiopsyper-formedwithin 8days afterPET.

ECAT931-08-12scanner(Siemens/CTI).Patientsfasted forat least 8hrs, andbloodglucosemeasuredand re-cordedbeforeinjectionof FDG.Dose de-pendedon bodymass:120-300MBq.

With ma-lignantdisease:115With be-nign dis-ease: 87

PET: 93%(Using aT/B cut-oflevel of3.0).


Tuber-culosis

English


Case re-ports oftwo pa-tients.



Breast in-fectionand in-flamma-tion

English


Case re-ports oftwo pa-tients.


Table 32: Selection of studies on PET (neurology)

A8.9 ALZHEIMER’S DISEASE



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Reimanet al.(2001)

Alz-heimer’sdisease

English

Proc NatlAcad SciUSA

4 het-erozy-gotes and4 noncar-rier con-trolsmatchedfor sex,age andeduca-tion. 2yearslater, 10heterozy-gotes and15 non-carriercontrols.

No N/A Volun-teers re-cruitedby news-paper ad-vertise-ments.Had afamilyhistory ofAlz-heimer’sdiseasewith atleast one1st-degreerelativeaffected.

N/A N/A N/A ECAT951/31scanner.Intrave-nous in-jection of10 mCiof 18-FDG.

N/A N/A

Kawanoet al.(2001)

Alzheimer’sdisease

English

Dementiaand Geri-atricCognitiveDisorders

26 pa-tientswithslowlypro-gressingmemoryprob-lems.

No PET usedto meas-ure theregionalcerebralmetabolicrate ofglucose(rCMRglc).

Probabil-ity orpossibil-ity of Alz-heimer’sdisease.Patientsadmittedto hospi-tal forapprox.oneweek.

N/A N/A N/A N/A N/A N/A

A8.10 REFRACTORY EPILEPSY



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Dupontet al.(2000)

Epilepsy English

ArchNeurol

30 pa-tients

No Calcula-tion oftheasymme-try index.

30 con-secutivepatientswithtemporallobe epi-lepsywho hadunder-gone sur-gery andwho haddifferentpostop-erativehealthout-comes.

Not men-tioned.

None:surgery

ECAT953/31Bscanner(CTI/Siemens). 31transver-sal sec-tions ofthe brain30 minafter theintrave-nous in-jection ofFDG at ameandose of29.6 x107.

2 yearsafter sur-gery:14 sei-zure-free,10 mostlyimprovedand 6 withpersistentseizures.



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Muzik etal. (2000)

Epilepsy English

Neurol-ogy

10 pa-tients

No Semiau-tomatedsoftwarefor de-finingabnormalsuprat-entorialcorticalareas inPET im-ages.

Mostlypediatricpatientswith re-fractoryex-tratempo-ral epi-lepsy.

Notstated.

Not men-tioned.

FMZ-PET, in-tracranialEEG.

Patientsfasted for4 hrs be-fore thescanning.PET per-formedduringthe in-terictalstate.CTI/SiemensECAT/HRscanner.Scanningbegan 30min afterthe in-jection ofFDG(0.143mCi/kg).

Data notgiven.

Detectionof epileptogenic foci:FMZ at10%threshold:81 ± 9%12%threshold57 ± 10%15%threshold37 ± 12%.FDG at10%threshold53 ± 13%(p=0.06)12%threshold:32 ± 12%(p=0.03)15%threshold:16% ± 6%(p=0.08).Detectionof corticalareas ofseizurespread10%thresholdFMZ: 31 ±9%; FDG:34 ± 9%(p=0.51)15%thresholdFMZ: 19 ±8%, FDG10 ± 6%(p=0.28)Regions ointerictalspiking10%threshold80 ± 7% foFMZ, 57 ±13% forFDG(p=0.15)

Tatlidil etal. (2000)

Epilepsy English

Acta neu-rol. Belg.

35 pa-tients,100 con-trols (7515O-water, 25FDG).

No Mean in-dex ofmetabolicsymme-try cal-culated.Abnor-mal PETfindingsdividedinto 3groups:mild,moderate

35 pa-tientswho hadhad ananteriortemporallobec-tomy forcomplexpartialseizures.

Referredby 3centres.

Not men-tioned.

MRI,electroen-cephalo-graphicvideomonitor-ing withrecordingof sei-zures,and neu-rologicalexamina-

GE/Scanditronixscanner.Bolus of2,405-2,960MBq of15O-wateradminis-tered foreach re-cordingafterblood

Histopa-thologicabnor-malitiesobservedin 30 pa-tients. 15patientshad hip-pocampalsclerosis.After atemporallobec-

PET ofblood flow80%

Patientswith nor-mal MRI:FDG-PET78%PET ofblood flow64%Visualanalysis of



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

and se-vere.

tion. flow im-aging.204 MBqof FDGinjected.

tomy, 20patientswere sei-zure-free,13 showedsignificantimprove-ment and2 did notshow sig-nificantimprove-ment.

MRI: 60%(detectionof lesions)

Ryvlin etal. (1998)

Epilepsy English

Brain

100 pa-tients and12 con-trols.

No Qualita-tiveanalysisreviewedby 2 in-depend-ent PETexperts.

100 con-secutivepatientswith re-fractoryepilepsy.

Prospec-tive in-clusionof 100consecu-tive pa-tients re-ferred tothe centrefor medi-cally re-fractorypartialepilepsy.

Yes. PETinterpre-tationsblindedto theotherdata.

MRI,FMZ-PET,videoECGmonitor-ing.

FMZ-andFDG-PET per-formedon thesame dayin 90 pa-tientswith atime-of-flight de-vice(TTV03,LETI,CENG).FDG-PET per-formed90 minafter thelast in-jection ofFMZ.Methodsdescribedelse-where.

Temporallobe epi-lepsy(n=52):Nonlateralmesio-temporalsclerosisMRI(n=30)Detectionby MRI30/30FDG-PET29/30FMZ-PET30/30

FMZ-PET73%MRI: 66%

Hwang etal. (2001)

Epilepsy English

AJNRAm JNeurol

117 pa-tients(103 un-derwentPET, 93interictalSPECTand 91ictalSPECT.

No Visualandqualita-tive in-terpreta-tion.

Con-secutivepatientswho un-derwentsurgeryfor in-tractableneocorti-cal epi-lepsy.

Reviewof 358medicalrecords,includingthose of117 pa-tientswith neo-corticalepilepsy.

Blindedreinter-pretation.

SPECT,MRI,pathol-ogy.

ECATEXACTscanner.Scanningper-formed60 minafter theinjectionof 370MBq of18-FDGduringthe in-terictalperiod.

117 pa-tients inall (patho-logic di-agnosis:50 tempo-ral, 33frontal, 15occipital,13 parie-tal, 2 hemi-spheric, 4multifo-cal); 77neuronalmigrationdisorder(patho-logic di-agnosis:28 tempo-ral, 25frontal, 10occipital,

Overall ratof correctlocalizationof epileptogenic foci:MRI:59.8%PET:77.7%IctalSPECT:70.3%For patientwith anadequatepostopera-tive followup:MRI:65.5%PET:77.2%IctalSPECT:73.8%Rate of cor



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

10 parie-tal, 2 hemi-spheric, 2multifo-cal), 15tumors(patho-logic di-agnosis:13 tempo-ral, 1 oc-cipital, 1multifo-cal) and25 others(9 tempo-ral, 8frontal, 4occipital,3 parietal,1 multifo-cal).

rect local-ization oftemporalneocorticaepilepsy:MRI:64.0%,PET:86.7%IctalSPECT:0.6%)Extratem-poral:MRI:56.7%,PET:70.7%IctalSPECT63.6%

A8.11 BRAIN TUMORS (MAINLY GLIOMA)



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Bader etal. (1999)

Sus-pectedrecur-rences ofglioma

English

Eur JNuclMed

30 pa-tientswho werepart of alargergroup ofpatientsreferredconsecu-tivelybecauseof sus-pectedrecur-rence orforgradingafter ini-tial treat-ment.

No Visualanalysisin axial,coronaland sag-ittalplaneswith areversedgrayscale anda quanti-tativeanalysisof thePET re-sults se-lected foranalysisof the re-gions ofinterest,based ona stan-dardizedgrayscale.

Patientswith sus-pectedrecur-rence andsched-uled forfurthertreat-ment.Initialtreatmentended atleast 6monthsbeforethe PETstudy.

9 patientswith aninitial di-agnosisof gradeII astro-cytoma,10 withgrade IVglioma,3 withgrade IIoligoas-trocy-toma, 6withgrade IIoligoden-droglio-ma and 2withgrade IIIoligoden-droglio-ma.

2 inde-pendentobserversblindedto theclinicalandhistopa-thologydata,classifiedthe PETandSPECTimages aspositiveor nega-tive.

SPECT,stereo-tactic bi-opsy.

ECATARTcamera(Sie-mens/CTI, USA).12-hourovernightfast.Admin.of 200MBq ofFDG.Tran-saxialimagesrecon-structedwith fil-teredback-projectionand cor-rected forattenua-tion (seeref. fordetails).

29 recur-rences and1 nontu-mor lesion(postop-erativescar).

SPECT100% forgrade IV,86% forgrade III,75% forgrade II.

PET100% forgrade IV,71% forgrade III,50% forgrade II.



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

DeWitteet al.(2001)

High-gradeastrocytoma

English

J Neu-rooncol

91 pa-tients (30withgrade IIImulti-formglioblas-tomas, 61withgrade IVglioblas-tomas).

No Qualita-tiveanalysisusing thefollowinggradingscale: 1,uptakelowerthan inthe con-tralateralwhitematter; 2,uptakeinterme-diatebetweenthe con-tralateralwhitematterand cor-tex; and3, uptakegreaterthan orequal tothat inthe con-tralateralcortex.

Patientswithhigh-gradegliomaprovenhistologi-cally forprogno-sis.

91 pa-tientsstudied,butanalysiscon-cernedonly 26patientsbecauseof deaths(n=50) orfollow-upnot done(n=15).

2 inde-pendentobserversinter-pretedthe PETscansvisually.

None Sie-mens/CTI933/08-12 scan-ner with6.75-mmsections.Attenua-tion cor-rection.IV injec-tion of260 MBqof FDG,withscanningbeginning40 minlater.

8 patients:grade I

42patients:grade II

41patients:grade III

No datagiven.

Derlon etal. (2000)

Oligoden-droglio-mas

English

Eur JNuclMed

47 pa-tientswithhistologi-cally con-firmedoligoden-droglio-mas (27low-grade; 20high-grade).

No Histologyslices re-viewedby thesamepatholo-gist:regionsof inter-est, vis-ualanalysis.

Meantumor/healthytissue ra-tio for11C-METandFDG;min. andmax. ra-tio foreachslice;standarddeviationfor theratio val-ues andthe totalvolume

Notgiven.

CT, MRIand 11C-METPET.

LETITTV03scannerwith hightransaxialresolu-tion, 7planes,attenua-tion cor-rectionbased ontransmis-sion with68Ge. Af-ter June1996:ECATHR+scanner(Sie-mens/CTI).

No datagiven.



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

of the inte-gratedregionsof inter-est forthe dif-ferentslices.

Eary etal. (1999)

Malig-nantbrain tu-mors

English

CancerRes

13 pa-tients

No Patientsunder-wentcloselyspaced18F-FDG/11C-dThd +MRI.Image re-sultscomparedby stan-dardizedvisualanalysis.Gradingby 2 ex-periencedobserv-ers.Qualita-tiveanalysis.

0 = up-take lessthan orequal tothat ofnormaltissues

1 = mini-mallyabnormaluptake

2 = up-takedefinitelyabnormal

Compari-son of re-sults:

dThd vs.FDG

dThd vs.IRM

FDG vs.MRI

Primaryor recur-rent braintumors.Patientsreferredfor PETscanningby theUniver-sity ofWash-ingtonMedicalCentre.

Not men-tioned

2-[C-11]thy-midine(dThd),PET,MRI.

Imagingby 11C-dThdfollowedby 18F-FDGPET.Bloodsample.10 to 20mCi 2-[C-11]dThdinjectedIV for 60secondswith Har-vard infu-sionpump.GE-Advancescanner.Image re-construc-tion by3-D re-projectionalgo-rithms.Withtransver-sal andaxial fil-ters(moredetails inarticle).Imagingby FDG-PET after11C. 10mCi ofFDG ad-minis-tered IVwith Har-vardpump for2 min.Bloodsample(glucose).

4: multi-formglioblas-toma7:anaplasticastrocy-toma1: periph-eralneuro-ectoder-mal tumor1: cysticadenoidcarcinoma

No datagiven.

Nuutinen Astrocyto English 14 pa- No Stan- Newly Yes. MRI/CT 425 MBq Data not Data not



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

et al.(2000)

masInt J Ra-diat On-col BiolPhys

tients:13 withlow-grade as-trocy-tomas, 1withmulti-formglioblas-toma.

dardizeduptakevalue andtu-mor/con-tralateralbrainSUV ra-tios.Visualandquantita-tiveanalyses.All im-agesshowingareas ofincreasedMETuptake inthe tumorregionwereconsid-eredpositiveon visualanalysis.

diag-nosed orrecurrentlow-gradetumors.

Blindedreadingby 2 on-colo-gists/radiologistsguidedby theMRI re-sults orwithknowl-edge ofthe MRI+ CT re-sults aswell asthe clini-cal dataon thepatients.

of 11C-me-thionine.Patientsfasted for5 hrs.ECAT931/08scanner(Siemens/CTI). 156.7-mmslices.Imagesrecon-structedaccordingto MRI.Attenua-tion cor-rectionwith68Ge.Scanningbegan 20min afterinjection.

given. given.

Sato etal. (1999)

Gliomas English

Abstract

Neurox-surg Rev

13 pa-tients

11C-me-thioninePET.

Stokkelet al.(1999)

Recurrentbrain tu-mor

English

Abstract

NuclMed

16 pa-tients

Quanti-tativeanalysiswiththalliumindex and

Sus-pectedrecurrentbrain tu-mor.

SPECT SPECT etPET per-formedon thesame day.Coinci-

12 withrecur-rence.

SPECT:92%PET: 62%(p= 0.023)



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

Commun FDG in-dex.Also,visualassess-ment.

dencedetectioncamera.

Thompson et al.(1999)

Recurrenttumors

English

Abstract

Stereotact FunctNeurosurg

15 pa-tients

14 withdisease, 1without

PET: 43%

Roelckeet al.(1999)

Low-grade as-trocy-tomas

English

J NeurolNeuro-surg Psy-chiatry

30 pa-tients

Visualanalysisof the re-gions ofinter-est/quantitativeanalysisof radio-activityconcen-trationratios intumors(T) overcontralat-eral brain(C)(T/C).

Patientswho didor did notreceiveradiationtherapysubse-quent tofirst tu-mor re-section.

Not men-tioned

FDGcomparedwithMET.

CTIscanner(933/04-16) MET:35 minafter in-jection.FDG: 48min afterinjection.

The twotests wereper-formed 3to 4 hrsapart.

In patientswho hadreceivedradiationtherapy (n=13):MET T/C:1.31 (0.42)

FDG T/C:0.90 (0.16)

In patientswho did noreceive ra-diationtherapy (n=17):MET T/C:1.33 (0.40)

FDG T/C:0.82 (0.10)

Malignantprogression(yes, n=7):MET 1.70(0.64);



Duplicatestudy?

Type ofPETanalysis

Inclusioncriteria


Refer-ence test

PETmethods


Sensitivity

FDG 0.98(0.23)

(no; n=13)MET 1.21(0.21);FDG 0.82(0.08)

A8.12 MYOCARDIAL VIABILITY

Table 33: Selection of studies using a clinical criterion as an assessmentcriterion and studies on changes in management due to PET scan re-

sults (cardiology)

StudyMethodologi-cal quality(grade)

Number ofpatientsAgeMean dura-tion of follow-up

LVEF Assessmentcriterion

Groups Results

Eitzman et al., 1992Retrospective

C 8259 ± 10 years12 months

34 ± 13% Composite(deaths, cardiacarrests,MIs, laterevascularization)

Comparison of patientoutcomes with respectto treatment (medicaltreatment, revasculari-zation) and PET results.

Significantly more deaths in the group with viabltreated medically.

Yoshida et al., 1993Retrospective

C 3554 years3 years

43.6% Death Comparison of patientoutcomes with respectto treatment (medicaltreatment, revasculari-zation) and PET results.

No significant difference in mortality between theMortality lower in the patients with viable myoca

Tamaki et al., 1993Retrospective

C 8458 ± 9 years23 ± 12.7 mos

50 ± 12%(group withoutevents)42 ± 14%(group withevents)

Composite(deaths, MIs,angina, late re-vascularization).

Comparison of medi-cally treated patientswith and without vi-ability.

FDG uptake is a predictor of events.

Lee et al., 1994Retrospective

C 12962 ± 11 years17 ± 9 months

38 ± 16% Composite (CVevents +deaths).


The patients with viable tissue who were treated mischemic events (48%) than the patients with viabrevascularized (8%, p<0.001).Cox analysis: Viability and no revascularization aonly age and EF, not viability, are predictors of d

Di Carli et al., 1994Retrospective

C 9365 ± 10 years13.6 months

28 ± 6% Mortality Comparison of patientoutcomes with respectto treatment (medicaltreatment, revasculari-zation) and PET results.

The patients with viable myocardium who underwhigher 1-year survival rate than those treated med

Di Carli et al., 1995Prospective

C 3666 ± 8 years25 ± 14 mos

28 ± 6% Improvement inheart failuresymptoms.

All the patients wererevascularized.

There was greater functional improvement in the cardium as detected by PET.

Haas et al., 1997Retrospective

B 69PET group: 60± 10 yearsNon-PETgroup: 63 ± 9years12 months

≤ 0.35% Survival at 12months after re-vascularization.

Comparison of patientoutcomes with respectto performing or notperforming PET.35 PET patients34 non-PET patients

12-month survival rate: 97% with PET vs. 79% wIn-hospital mortality: PET, 0; other group, 11.4%

StudyMethodologi-cal quality(grade)

Number ofpatientsAgeMean dura-tion of follow-up

LVEF Assessmentcriterion

Groups Results

Beanlands et al., 1997 B 80 81 pts < 50%41 < 30%

Change in man-agement.

Change in management strategy in 57% of the cathe LVEF was < 30%.

Vom Dahl et al., 1997Prospective

B 16157 ± 9 years29 ± 6 months

45 ± 12% Composite+ changes insymptoms.


Among the patients who were revascularized, onlon PET had a significant decrease in the number Improvement in symptoms in the patients with vivascularized in relation to those who were treated

Pagano et al., 1998 C 35 23.6% Change in All the patients were Correlation between the number of viable segmen

Retrospective 45 to 72 years6 months

LVEF, exercisecapacity, andquality of life.

revascularized. LVEF. No correlation with exercise capacity or q

Di Carli et al., 1998Retrospective

C 9369 yearsMedian fol-low-up of 4years

Median: 25% Survival at 4years.


In patients with viability on PET.4-year survival rate with medical treatment: 30%

Schelbert et al., 1998Retrospective

C 112 evaluatedfor cardiactransplantation

≤ 35% Survival at 5years.

Comparison of patientoutcomes with respectto treatment (medicaltreatment, revasculari-zation, transplantation)and PET results.

5-year survival rate:80% in the group with viable myocardium that w71.4% in the group without viable myocardium th42% in the group without viable myocardium tha

Landoni C et al., 1999

Retrospective

C 241(153 evaluatedby PET)

29.8 ± 6.7% Survival at 30days.

Comparison of out-comes in the revascu-larized patients with re-spect to PET scan re-sults.

30-day mortality rate:0.9% in the PET group.19.8% in the group that was not assessed by PETOnly predictor of perioperative outcome: presenction fraction.

Marwick et al., 1999Prospective

B 6366 ± 9%17 months

28 ± 7% Survival, func-tional capacity,quality of life.

All the patients wererevascularized.

Myocardial viability is a predictor of the improveof quality of life.

Siebelink et al., 2001Prospective

A 10328 ± 1 months

Cardiovascularevent-free sur-vival.

Comparison of thegroups managed in lightof the of PET results vs.SPECT results.

No difference between the two groups.

Table 34: Selection of studies of the use of PET, using an intermediate cri-terion as an assessment criterion (cardiology: functional reversibility of

segmental wall motion)Study Patients (n) LVEF Sensitivity (%) Specificity (%) PPV (%)Tillisch et al., 1986 17 32 ± 14% 95 80 85Tamaki et al., 1989 22 NR 78 78 78Tamaki et al., 1991 11 NR 100 37 80Carrel et al., 1992 23 34 ± 14% 94 50 84Lucignani et al., 1992 14 38 ± 5% 93 86 95Gropler et al., 1992 16 NR 79 83 79Manwick et a., 1992 16 NR 71 76 68Vanoverschelde et al., 1993 12 55 ± 7% 100 - 100Gropler et al., 1993 34 NR 83 50 52Maes et al., 1994 20 48 ± 9% 82 67 75Knuuti et al., 1994 48 53 ± 11% 85 84 70Vom Dahl et al., 1994 37 34 ± 10% 66 77 53Grandin et al., 1995 25 49 ± 11% 88 50 79Tamaki et al., 1995 43 41 88 82 76Gerber et al., 1996 39 33 ± 10% 75 67

Bax et al., 1996 17 36 ± 11% PET: 89Thallium-201: 93Echo stress 85

PET: 77Thallium-201: 43Echo stress: 63

PET:62Thallium-201: 40Echo stress: 49

Vom Dahl et al., 1996 52 47 ± 10% 95 73 68Baer et al., 1996 42 40 ± 13% PET: 96

Echo stress: 92PET: 69Echo stress: 88

PET: 72

Maes et al., 1997 30 46.5% - - PET: 91MIBI SPECT: 82

Wolpers et al., 1997 30 42 ± 11% - - 78Pagano et al., 1998 30 25 ± 7% PET: 99

Echo stress: 61PET: 33Echo stress: 63

PET: 66Echo stress: 68

Fath-Ordoubadi et al., 1998 47 ≤ 30% 63 to 81, depending othe FDG uptake cut-olevel.

Zhang et al., 1999 60 44 ± 15% 76 86 88

Rossetti et al., 1999 17 52.5 ± 7% PET 87 PET 26,3 PET: 45,7MIBI SPECT: 46.5THALLIUM SPECT:47.4

Study Patients (n) LVEF Sensitivity (%) Specificity (%) PPV (%)Kitsiou et al., 1999 26 31 ± 8% 65 to 76, depending o

the FDG uptake cut-olevel.

Pasquet et al., 2000 66Global LV function

28 ± 5% PET: 56Echo stress: 94

PET 64Echo stress: 59

Wiggers et al., 2000 46 35 ± 7% PET: 81Echo stress: 51%

PET: 56Echo stress: 98

McFalls et al., 2000 20Global LV function

≤ 27% 67


Appendix 9: Methodological quality of studies189

Appendix 9

Methodological Quality of Studies



APPENDIX 9: METHODOLOGICAL QUALITY OF STUDIES

Table 35: Methodological Quality of Studies – Lung Cancer

STUDY STUDY FEATURES COMMENTSErasmus et al., 2000 - Small sample (25 cases: 18 men; age: 37 to 86).

- Retrospective design (unit of analysis: the patient).- All eligible patients included in analysis.- Diagnosis defined by appropriate reference standard.

PET procedure adequately described; no mention ofblinded reading.

- PET not compared with conventional imaging.

Grade C- Small sample, no mention of blinded

reading.- No comparison of sensitivity and specific-

ity rates with those of conventional imag-ing methods.

- Retrospective design.

Gupta et al., 2000 - Large sample (118 cases: 73 men; age: 35 to 84years).

- Consecutive cases (unit of analysis: the nodal sta-tion).

- All eligible patients included in the study, but only54 cases included in the analysis.

- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; blinded reading

of images.- PET compared with CT.

Grade B- Certain cases excluded from the analysis

(only 54 of the 118 cases were included).- Resection based on staging by PET + CT.

Vanuytsel et al., 2000 - Large sample (105 cases; sex and age distribution notgiven).

- Retrospective design (unit of analysis: the nodal sta-tion).

- All eligible patients included.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-

ated independently.- PET compared with CT and PET + CT.

Grade B

- Overlapping of cases with those from twopreviously published studies.

Dewan et al., 1997 - Adequate sample (52 cases; age > 30 years; sex dis-tribution not given).

- Retrospective design with consecutive cases (unit ofanalysis: the nodal station).

- All eligible cases included in the study and analysis.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-

ated independently.- PET not compared with conventional imaging.

Grade C- Referred cases from the same institutions

as the cases in a previous study by Dewanet al. (1995).

- No comparison between PET and conven-tional imaging.

- Retrospective design.

Marom et al., 1999 - Large sample (100 cas: 58 men; age: 25 to 83).- Prospective design; consecutive cases (unit of analy-

sis: the patient).- 39 of the 139 the initial consecutive patients excluded

from the study, but the 100 remaining cases includedin the study and analysis.

- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-

ated independently.- PET compared with CT.

Grade A



STUDY STUDY FEATURES COMMENTSPatz et al., 1995 - Adequate sample (42 cases: 26 men; age: 25 to 85).

- Prospective design (unit of analysis: the nodal sta-tion).

- Not all the eligible patients were included in thestudy, but all who were included in the study werealso included in the analysis.



Grade B

- Adequate sample, but small.- Not all the eligible patients were included

in the study.

Vansteenkiste et al., 1998 - Adequate sample (68 cases; age: 40 to 83; sex distri-bution not given).

- Prospective design (unit of analysis: the nodal sta-tion).

- Not all the eligible patients were included in thestudy, but all who were included in the study werealso included in the analysis.


ated independently.- Comparison between PET + CT and CT alone.

Grade B

- Not all eligible patients were included inthe study.

- Comparison between CT and PET + CTbut not with PET alone.

- Overlapping of cases in this study withthose in 1997 study by the same authors.

Vansteenkiste et al., 1997 - Adequate sample (50 cases: age: 40 to 83 sex distri-bution not given).

- Prospective design (unit of analysis: the patient).- Not all the eligible patients were included in the

study.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-

ated independently.- PET compared with CT alone and with PET + CT.

Grade B

- Not all eligible patients were included inthe study.

- Overlapping of cases with those in a sub-sequent study by the same authors(1998).

Pieterman et al., 2000 - Large sample (102 cases: 88 men; age: 25 to 77).- Prospective design, consecutive cases (unit of analy-

sis: the patient).- All eligible cases included in the study and analysis.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-


Grade: A

Roberts et al., 2000 - Large sample (100 cases; age and sex distribution notgiven).

- Retrospective design (unit of analysis: the patient).- All eligible cases included in the study and analysis.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; no mention of

blinded reading.- PET not compared with conventional imaging.

Grade C:

- Retrospective design.- No mention of blinded reading.- PET not compared with conventional im-

aging.- The cases were patients referred to the

study – possibility of selection bias.



STUDY STUDY FEATURES COMMENTSHara et al., 2000 - Small sample (29 cases: 19 men; age: 40 to 83).

- Retrospective design (unit of analysis: the nodal sta-tion).

- All eligible cases included in the study and analysis.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-

ated independently.- FDG-PET compared with CT and C-choline PET.

Grade C

- Small sample.- Retrospective design.

Weber et al., 1999 - Small sample (27 cases: 26 men; mean age: 62 ± 9).- Retrospective design (unit of analysis: the patient).- All eligible cases included in the study and analysis.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-

ated independently.- PET compared with CT and coincidence imaging

(coincidence gamma camera).

Grade C

- Small sample.- Cases with disease only.- Retrospective design.

Berlangieri et al., 1999 - Adequate sample (50 cases: 37 men; age: 41 to 78).- Consecutive cases (unit of analysis: the nodal sta-

tion).- All eligible cases included in the study and analysis.- Diagnosis defined by appropriate reference standard.- PET procedure adequately described; images evalu-


Grade B

- Size of patient sample.

Saunders et al., 1999 - Large sample (97 cases: 64 men; age: 36 to 77).- Consecutive cases (unit of analysis: the nodal sta-

tion).- Diagnosis defined by appropriate reference standard.- All eligible cases included in the study, but only 84

were included in the analysis.- PET procedure adequately described; images evalu-


Grade B

- Not all the cases were included in theanalysis.

- Cases referred to the study (possibility ofselection bias).

Magnani et al., 1999 - Small sample (28 cases: 26 men; age: 50 to 75).- Retrospective design (unit of analysis: the patient).- Diagnosis defined by appropriate reference standard.- All eligible cases included in the study and analysis.- PET procedure adequately described; images evalu-


Grade C

- Small sample.- Retrospective design.



Table 36: Methodological Quality of Studies – Colorectal Cancer

STUDY SITE LANGUAGE

AND SOURCE

INCLUDE

OR

EXCLUDE

COMMENTS METHODOLOGICAL

QUALITY

Valk et al., 1999 Cancer colorectal English

Arch Surg

Include Inclusion/exclusioncriteria clearly stated.PET and CT imagesread at the same timeby 1 or 2 investigatorswith access to the clini-cal data.Ambiguous.

A

Imdahl et al.,2000

Metastases from co-lorectal cancer

English

Langenbeck’sArch Surg

Include Inclusion/exclusioncriteria not clearlystated (consecutive pa-tients?).Circular reference stan-dard.

C

Zhuang et al.,2000

Hepatic metastasesfrom colorectal can-cer

English

Nucl MedComm

Include Direct comparison:PET/surgical pathol-ogy/clinical follow-up– blinded reading of re-sults???? Performedwithin 8 weeks of con-ventional imaging.80 consecutive patients(retrospective).

B

Staib et al., 2000 Recurrent colorectalcancer

English

Am J Surg

Include Prospective recruit-ment.Inclusion/exclusioncriteria clearly stated.PET scans evaluatedindependently, withreaders blinded to re-sults of conventionalimaging but withknowledge of diagnosisand indication.Reference standardsomewhat circular.

B

Willkomm et al.,2000

Recurrent colorectalcancer

English

J Nucl Med

Include Blinded reading ofscans with conventionalimaging.Recurrent disease de-tected in only 9 patients(small sample).Prospective recruit-ment.

C

Table 37: Methodological Quality of Studies - Melanoma

STUDY SITE LANGUAGE

AND SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Paquet et al.,2000

Metastatic melanoma English Include Small sample (24 pa-tients).

D



STUDY SITE LANGUAGE

AND SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Dermatology Comparison betweenPET and conventionalimaging.Blinded reading? Nomention.Inclusion/exclusioncriteria not clearlystated.PET not compared withconventional imaging.

Eigtved et al.,2000

Metastatic melanoma English

Eur J Nucl Med

Include Consecutive patients.PET scans visually in-terpreted. PET com-pared with conven-tional imaging andclinical examination.

C

Acland et al.,2000


J Am AcadDermatol

Include PET compared withhistology.Patients referred forPET for various rea-sons.Patients identified ret-rospectively in a PETdatabase.PET not compared withconventional imaging.

C

Krug et al., 2000 Metastatic melanoma English

Acta Radiologica

Include Consecutive patientsstudied retrospectively.No sensitiv-ity/specificity data.

C

Wagner et al.,1999

Melanoma English

J Clin Oncol

Include Prospective study.Methodology ade-quately described.Inclusion/exclusioncriteria not clearlystated.PET scans interpretedin a blinded fashion bya single investigator.74 patients.

A

Jadvar et al.,2000

Melanoma English

Clin Nucl Med

Include Retrospective chart re-view.Small sample (38 pa-tients).Reading done by a sin-gle observer who wasaware of the clinicalchanges and previousradiographic results.No sensitiv-ity/specificity data.

D

Dietlein et al.,1999

Melanoma English

Nucl MedComm

Include 68 patients (smallnumber in the analysis).No protocol for time orextent of the examina-tion.

D



STUDY SITE LANGUAGE

AND SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

PET scans were read byphysicians differentfrom those who readthe x-rays and ultra-sounds.Inclusion/exclusioncriteria not clearlystated.No independent refer-ence standard.

Crippa et al.,2000


J Nucl Med

Include Inclusion/exclusioncriteria not clearlystated.38 patients followedprospectively.PET scans read in ablinded fashion and re-read retrospectively.No comparison, butsensitivity/specificitywith respect to size.

C

Table 38: Methodological Quality of Studies – Head and Neck Cancer

STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Di Martino etal., 2000

Head and neck can-cer

English

Arch Otolaryn-gol Head NeckCan

Include 50 patients recruitedprospectively.Inclusion/exclusioncriteria clearly stated.Blinded reading? Nomention.Method of interpretingPET scans not ex-plained.

C

Lowe et al.,2000

Recurrent head andneck cancer

English

J Clin Oncol

Include Overlapping of patientswith those of anotherstudy (44).30 patients in thisstudy.Inclusion/exclusioncriteria not clearlystated.Blinded reading? Nomention.Method of interpretingPET scans not ex-plained.PET compared withcorrelative imaging.

C

Lonneux, 2000 Recurrent head andneck cancer

English

Laryngoscope

Include 44 patients recruitedprospectively.Inclusion/exclusion

C



STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

criteria.Blinded reading.PET compared withmorphological imaging.

Jungehulsing etal., 2000

Unknown primarywith manifestation inthe head and necklymph nodes

English

OtolaryngolHead & NeckSurg

Include 27 patients with un-known primary.Inclusion/exclusioncriteria not clearlystated.No mention of blindedreading.

C

Bohuslavizki etal., 2000

Unknown primary English

J Nucl Med

Include Knowledge of clinical,surgical and histopa-thologic findings andcorrelative imagingwere used to assess thePET scan.53 patients evaluatedretrospectively.

D

Perie et al., 2000 Unknown primary English

Ann Otol Rhi-nol Laryngol

Exclude N/A N/A

Lassen et al.,1999

Unknown primary English

Eur J Cancer

Include 20 patients.Knowledge of correla-tive imaging findings,histological findingsand localization infor-mation on metastasesavailable at the time ofPET scan interpreta-tion.

D

Farber et al.,1999

Recurrent head andneck cancer

English

Laryngoscope

Include Retrospective, 28 pa-tients.Two readers visuallyanalyzed the PET im-ages.No mention of blindedreading.

C

Table 39: Methodological Quality of Studies – Lymphoma

STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Jerusalem et al.,1999

Hodgkin’s and non-Hodgkin’s lym-phoma

English

Blood

Include Prospective recruit-ment.CT performed uponabnormal PET findings.PET scans examined bya single investigator.

C

Buchmann et al.,2001

Lymphoma EnglishInclude

52 patientsReference test per- B



STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Cancer formed within 2 weeksfollowing PET.Discrepant results wereverified by biopsy.Blinded reading.

Hueltenschmidtet al., 2001

Lymphoma English

Cancer

Include81 patients.Inclusion/exclusioncriteria not given.Conventional methodsused irregularly.

C

Jerusalem et al.,2000

Lymphoma English

Haematologica

IncludeOnly 28 patients.No comparison. D

Spaepen et al.,2001

Lymphoma English

Journal ofClinical Oncol-ogy

Include93 patients.Pre- and posttreatmentevaluation.Blinded reading.

B

Tatsumi et al.,2001

Lymphoma English

Journal of Nu-clear Medicine

IncludeSmall sample (30 pa-tients).CT performed within 2weeks following PET.Blinded reading.

C



Table 40: Methodological Quality of Studies – Breast Cancer

STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Cook andFogelman 1999

Breast - metastases English

Sem Nucl Med

Exclude N/A N/A

Crippa et al.,1997

Breast - metastases English

Tumori

Include Only 16 cases withoutthe disease.31 axillary lymphnodes + /52 axillarynodes -.Inclusion/exclusioncriteria not given.Patient characteristicsnot given.

C

Cowen, 1998 Breast - metastases French

Can-cer/Radiotherapie

Exclude N/A N/A

Noh et al., 1999 Breast cancer English

Eur J Surg

Exclude N/A N/A

Crippa et al.,1998

Breast cancer English

J Nucl Med

Include Consecutive patientswith palpable nodules;27 nodes +/ 45 nodes -.PET images interpretedin a blinded fashion.PET compared withpathology report.

B

Raylman et al.,2000

Breast cancer English

Med Phys

Exclude N/A N/A

Table 41: Methodological Quality of Studies – Prostate Cancer

STUDY DETECTION

OF:LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Seltzer et al.,1999

Prostate cancer English

J Urol

Include PET images examinedby 2 investigators whowere not blinded toeach other but whowere blinded to thefindings of the mono-clonal antibody scans.Reference standardsomewhat circular.Prospective recruit-ment.

C



Table 42: Methodological Quality of Studies – Other Applications(not examined in this report)

STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Flamen et al.,2000

Esophageal cancer EnglishJ Clin Oncol

Include 74 patients reviewedprospectively.Inclusion/exclusioncriteria clearly stated.The images were inter-preted blinded to pa-tient data.

B

Meltzer et al.,2000

Esophageal cancer EnglishClin Nucl Med

Include 47 patients.PET images reviewedby 2 nuclear medicineclinicians. CT scansread by one radiologistin a blinded fashion.Exclusion criteria.

C

Choi et al., 2000 Esophageal cancer EnglishJ Nucl Med

Include 61 consecutive patientsstudied prospectively.100 N+, 282 N-.PET and CT scans readby clinicians in ablinded fashion, byconsensus.Inclusion criteriaclearly stated.

B

Tiepolt et al.,2000

Thyroid cancer EnglishAnn Nucl Med

Include 31 patients.Reading of coincidencegamma camera scansblinded to PET scans.PET not compared withconventional imaging.

D

Lips et al., 2000 Thyroid cancer me-tastases

EnglishNetherl J Med

Exclude N/A (case reports of 4patients).

N/A

Schulte et al.,2000

Skeletal tumors EnglishJ Nucl Med

Include 202 patients.Inclusion/exclusioncriteria given.Protocol clearly de-scribed.PET images read bytwo independent clini-cians blinded to clinicaldata. Diagnosis deter-mined by surgicalspecimen.PET not compared withconventional imaging.

C

Barkheet et al.,2000

Breast infection andinflammation

EnglishClin Nucl Med

Exclude N/A N/A

Barkheet et al.,1999

Benign esophagealdisease

EnglishClin Nucl Med

Exclude N/A N/A

Stumpe et al.,2000

Soft-tissue and boneinfections

EnglishEur J Nucl Med

Include Patients with clinicalsymptoms of infection.Small sample (39 pa-tients); 40+ and 12- le-sions.

C



Table 43: Methodological Quality of Studies – Neurology

STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Reiman et al.,2001

Alzheimer’s disease English

Proc. Natl.Acad. Sci. USA

Exclude N/A N/A

Kawano et al. ,2001

Alzheimer’s disease English

Dementia andGeriatrix Cog-nitive Disorders

Exclude N/A N/A

Dupont et al.,2000

Epilepsy English

Arch Neurol

Include 30 consecutive patients.Comparison withpostsurgical outcome.

C

Muzik et al.,2000

Epilepsy English

Neurology

Include Very small sample(n=10).No mention of blindedreading.FDG-PET comparedwith FMZ-PET andwith intracranial EEG.

C

Ryvlin et al.,1998

Epilepsy English

Brain

Include 100 consecutive pa-tients, 12 controls.Blinded interpretation.FDG-PET, FMZ-PETand MRI.FDG-PET and FMZ-PET were performedon the same day.

B

Hwang et al.,2001

Epilepsy English

AJNR Am JNeurol

Include 117 patients.Blinded interpretationSPECT, MRI, pathol-ogy.Both tests were per-formed on the sameday.

B

Bader et al.,1999

Recurrent glioma EnglishEur J Nucl Med

Include Confirmation of recur-rence.30 patients.Interpretation of PETand SPECT blinded toclinical or histopa-thologic findings.Comparison withSPECT.

B

DeWitte et al.,2001

Glioma (high-gradeastrocytoma)

English

J Neurooncol

Include Tumor grading.91 patients.No comparison withconventional imaging.Blinded reading.No sensitivity data.

C

Derlon et al.,2000

Glioma (oligoden-drogliomas)

English

Eur J Nucl Med

Include FDG compared withMET.47 patients.Comparison with MRI+ CT.

C



STUDY SITE LANGUAGE

AND

SOURCE

INCLUDE

OR

EXCLUDE


QUALITY

Blinded reading.

Eary et al., 1999 Malignant brain tu-mors

English

Cancer Res

Include FDG compared withMET.13 patients.No mention of blindedreading.No sensitivity data.

C

Nuutinen et al.,2000

Glioma(astrocytoma)

English

Int J RadiatOncol BiolPhys

Include? 11C-MET PET only.14 patients.Comparison with con-ventional imaging.Blinded reading.

C

Sato et al., 1999 Glioma English (AB-STRACT)

Neuro SurgRev

Include? 11C-MET PET only.13 patients.

C?

Stokkel et al.,1999

Recurrent brain tu-mors

English (AB-STRACT)

Nucl MedComm

Include 16 patientsComparison withSPECT.No mention of blindedreading in abstract.

C?

Roelcke et al.,1999

Glioma (low-gradeastrocytoma)

English

J Neurol NeuroSurg Psychiatry

Include 30 patients.FDG compared withMET.No mention of blindedreading.No sensitivity data(tumor-to-contralateralbrain uptake ratios).

C


Appendix 10: PET cost components203

Appendix 10

PET Cost Components



APPENDIX 10: PET COST COMPONENTS

Table 44: Purchase prices (cyclotron + PET)(taxes not included)

[Source: AMSMNQ, 2000]Cyclotron: $2.5 to $3.8 millionPositron emission tomograph: $1.8 to $3.2 millionRadiochemistry/radiopharmacy facilities: $300,000Equipment and nonrecurrent expenses: $200,000Construction costs: Depend on the site.Total: $4.8 to $7.5 million

+ construction costs and taxes

Table 45: Purchase prices (PET without cyclotron)(taxes not included)

[Source: AMSMNQ, 2000]Positron emission tomograph: $1.8 to $3.2 millionEquipment and nonrecurrent expenses: $100,000Construction costs: $200,000 to $400,000

Total: $1.9 to 3.3 million+ construction costs and taxes

Remodelling (or construction) costs: varyconsiderably according to the site.

Example:

For a cyclotron, about $250,000 for self-shielded models and about $1 million if a vaultneeds to be built. Depending on the site, thecosts can therefore vary substantially.

For a scanner, about $100,000 if the space isalready available. More, depending on the re-modelling needs not associated with the scan-ner. Related equipment (monitors, scanners,shielding): about $250,000. Anticipate differentimpacts, depending on the type of equipment.Example 1: Models with BGO crystals need tobe cooled with water. The costs will be similarto those of a CT scanner (cold water circuitwith water from the hospital. They can easilybe higher if the circuit does not already exist.

Example 2: ADAC∗ scanners are installed in anair-conditioned room that requires specialcooling.

∗ ADAC Laboratories is part of Phillips Medical Sys-

tems, a worldwide supplier of diagnostic imagingequipment.



Table 46: Operating costs (cyclotron + PET)(This figure is calculated for the output of a single site, at the rate of 1,500 scans a year. The output of

other sites requires additional equipment and personnel expenses.)[Source: AMSMNQ, 2000; modified by Dr. F. Bénard, May 2001]

Salaries and employee benefitsTechnologists (2): $101,840Secretary’s office/reception: $32,382Cyclotron operator: $50,000Chemistry technician: $50,000Radiochemist: $80,000Radiopharmacist (14 hrs/wk): $25,775Nursing support (1 day/month): $2,665Total salaries: $342,662

Equipment and suppliesSupplies: $35,000Laboratory reagents: $100,000Film: $3,000Recording paper: $1,000Laundry: $1,000Parts, maintenance: $25,000Biomedical waste equipment: $2,400Office supplies: $2,400Total equipment and supplies: $169,800

Maintenance contracts/costsCyclotron: (service contract) $150,000PET: (service contract) $210,000Total maintenance contracts: $360,000

Grand total: $872,462



Table 47: Operating costs (PET without cyclotron)(This figure is for about 1,000 scans a year.)

[Source: AMSMNQ, 2000, modified by Dr. F. Bénard, May 2001]

Salaries and employee benefitsTechnologists (2): $101,840Secretary’s office/reception: $32,382Nursing support (1 day/period): $2,665Total salaries: $136,887

Equipment and suppliesRadiopharmaceuticals: $350,000Supplies: $35,000Film: $3,000Recording paper: $1,000Laundry: $1,000Parts, maintenance: $25,000Biomedical waste equipment: $1,200Office supplies: $2,400Total equipment and supplies: $418,600

Maintenance contracts/costsPET: $180,000Total maintenance contracts:

Grand total: $735,487


Appendix 11: Eeconomicmodels:Additional information209

Appendix 11

Economic Models: AdditionalInformation

APPENDIX 11: ECONOMIC MODELS: ADDITIONAL INFORMATION

Figure 4: Decision tree (lung cancer)

QAppendix 11: Economic models: Additional information

palliativetreatment

positive

death

distant metastases positive

distant metastases negativesurvival

surgerynegative

mediastinoscopy for confirmationof mediastinal metastases

CT positive formediastinal metastases

death

mediastinal metastasespositive

distant metastasespositivedistant metastasesnegative

mediastinal metastasesnegative

survival

surgeryCT negative formediastinal metastases

survivalCT

deathCT

CT

palliativetreatment

distant metastasespositive

mediastinoscopy for confirmof mediastinal metastases

distant metastasesnegative

biopsypositive

palliativetreatment

positive

surgerynegative

mediastinoscopy - confirmation ofmediastinal metastases

negative

PET - detection of distantmetastases

CT positive formediastinal metastases

biopsypositive

Mediastinoscopy - confirmamediastinal metastases

negative

PET - detection ofdistant metastases

positive

biopsypositive

surgerynegative

PET - detection ofdistant metastases

negative

PET – detection of mediastinalmetastases

CT negative formediastinal metastases

survival CT

death

CT+PET


Appendix 11: Economic models: Additional information214

Table 48: Variables, intervals and predefined distributions for the Monte Carlo analysis (lungcancer)

Variable Baseline valueLowerlimit

Upperlimit Distribution

Cost* of hospital stay for mediastinoscopy 5,054 4,549 5,559 Normal, 20% variance

Cost of surgery 8,424 7,582 9,266 Normal, 20% varianceCost of hospital stay for biopsy 6,130 5,517 6,743 Normal, 20% varianceCost of hospital stay for mediastinoscopy, biopsy andsurgery 9,163 8,247 10,079 Normal, 20% variance

Sensitivity of CT 0.75 0.6 0.9 Triangular

Specificity of CT 0.66 0.55 0.77 TriangularLife expectancy** of patients who receive palliativetreatment 1 0.1 2 UniformLife expectancy of patients who are treated surgically 7 1 15 Uniform

Surgical mortality rate 0.03 0.02 0.2 Uniform

Sensitivity of PET in detecting distant metastases 0.82 0.64 1 Triangular

Specificity of PET in detecting distant metastases 0.93 0.88 0.98 Triangular

Sensitivity of PET in detecting mediastinal metast. 0.91 0.81 1 Triangular

Specificity of PET in detecting mediastinal metast. 0.86 0.78 0.94 TriangularProbability of metastases detected by PET 0.07 0.05 0.11 Uniform

Prevalence 0.31 0.28 0.38 Uniform* In Canadian dollars** In years



Table 49: Univariate sensitivity analysis (lung cancer)Incremental cost-effectiveness ratio

Baseline scenario 4,689Surgical mortality rate

0.02 47650.04 48210.06 48780.08 49370.1 4997

Prevalence0.15 4925

0.2375 47880.325 4842

0.4125 50310.5 5358

Cost of stay for surgery alone7544 5376

8603.25 47429662.5 4109

10721.75 347511781 2841

Cost of stay for surgery and mediastinoscopy7609 4489

8743.5 46359878 4781

11012.5 492712147 5073

Surgical mortality rate0 4599

0.05 47510.1 4915

0.15 50890.2 5277

Life expectancy of patients who receive palliative treatment0.1 4118

0.575 44011.05 4725

1.525 51012 5542

Life expectancy of patients who are treated surgically1 724242

4.5 52678 2643

11.5 176415 1324

Specificity of PET in detecting mediastinal metastases0.65 70800.7 7099

0.75 71180.8 7137

0.85 7156

0.64 31450.73 31130.82 30820.91 3051

1 3019

0.73 32260.7975 31770.865 3128

0.9325 30801 3032

Probability of distant metastases detected by PET0.05 2660

0.065 29620.08 3359

0.095 39010.11 4689

Sensitivity of PET in detecting distant metastases

Specificity of PET in detecting distant metastases



Figure 5: Monte Carlo analysis: 1,000 simulations (lung cancer)

-0.20

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0 500 1000 1500 2000 2500 3000

Incremental cost

Incr

emen

tal e

ffic

acy

Each point on the graph represents the result of a simulation. The horizontal axis represents the incre-mental cost, the vertical axis the incremental efficacy.

Table 50: Monte Carlo analysis (lung cancer): mean, median and quartiles

Incremental cost Incremental efficacyIncremental cost-effectiveness

ratioMean 1,253 0.26 17,666Median 1,260 0.22 5,342Quartile 25 1,052 0.12 3,006

50 1,260 0.22 5,34275 1,442 0.39 10,618

100 2 598 0.99 4,058,550



Table 51: Distribution of incremental cost-effectiveness ratios for the 1,000 Monte Carlo simula-tions (lung cancer)

Interval (incre-mental cost-

effectiveness ratio)Percentage

0 to 4 999 465,000 to 9,999 26

10,000 to 14,999 915,000 to 19,999 420,000 to 24,999 325,000 to 29,999 230,000 to 34,999 235,000 to 39,999 140,000 to 44,999 145,000 to 49,999 1

> 50,000 5Total 100



Figure 6: Decision tree (myocardial viability)

survival

death

revascularizationpositive

survival

deathrevascularizationviable

survival

death

treatment

survival

deathtransplantation

nonviable

clinicaldecisionequivocal

thallium and clinicaldecision

survival

death

revascularizationpositive

survival

death

revascularizationviable

survival

deathtreatmentmedical

survival

deathtransplantation

nonviable

PETequivocal

thallium and PET



Table 52: List of variables and predefined distributions (myocardial viability)

Description DistributionCost of a PET scan 20% variance, normalCost of revascularization 20% variance, normalCost of a thallium scan 20% variance, normalCost of medical treatment 20% variance, normalCost of transplantation 20% variance, normal5-year probability of survival following revascularization Baseline value5-year probability of survival following medical treatment Baseline value5-year probability of survival following transplantation Baseline valueProbability of medical treatment UniformProbability of unequivocal thallium scan UniformProbability of detecting viable myocardium in the context of anequivocal thallium scan in the thallium-alone option

Uniform

Probability of detecting viable myocardium in the context of anequivocal thallium scan in the thallium + PET option

Baseline value

Figure 7: Monte Carlo analysis: 1,000 simulations, detection of myocardial viability

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

-8000 -7000 -6000 -5000 -4000 -3000 -2000 -1000 0 1000 2000

Incremental cost

Incr

emen

tal e

ffic

acy


Appendix 12: List of experts consulted for the purpose of creating the decision trees221

Appendix 12

List of Experts Consulted for the Purpose ofCreating the Decision Trees


Appendix 12: List of experts consulted for the purpose of creating the decision trees223

APPENDIX 12: LIST OF EXPERTS CONSULTED FOR THE PURPOSE OFCREATING THE DECISION TREES

Clinical experts at Hôpital Laval in QuébecCity were consulted for the purpose of creatingthe decision trees, determining the explicitprobabilities for the variables and determiningthe population for which PET technologywould be used first.

Experts consulted

Specialists in heart failure

� Dr. Marie-Hélène Leblanc� Dr. Onil Gleeton� Dr. Denis Coulombe

� Dr. Erik Augustin

Specialist in nuclear medicine

� Dr. Jean Guimont

Specialists in PET

� Dr. Heinrick Schelbert (UCLA at LosAngeles)

� Dr. Paolo Camici (Hammersmith Hos-pital, London, UK)


References225

References


References227

REFERENCES

REPORTS BY CURRENT OR POTENTIAL USERS

Association des médecins spécialistes en médecinenucléaire du Québec (AMSMNQ) L’utilisationclinique de la tomographie par émission depositrons au Québec. Québec: AMSMNQ, 2000.

Beanlands R, Chamberlain M, et al. Ontario Council ofMedical Imaging. Position paper on the future ofPositron emission tomography in Ontario. 1999.

Bourgouin P. Association des radiologistes du Québec.Tomographie par émission de positrons -Position de l'Association des radiologistes duQuébec. October 14, 2000.

Positron Technology Committee. Reports prepared forthe September 26, 2000 meeting (unpublished).

Benkelfat, C. La tomographie par émission depositrons (TEP) en psychiatrie. Comité de travailsur la technologie du positron. Montréal: Fédérationdes médecins spécialistes du Québec, September2000.

Bogaty P. "Position Paper": Le point de vue de lacardiologie dans le dossier du PET SCAN. Comitéde travail sur la technologie du positron. Montréal:Fédération des médecins spécialistes du Québec,September 2000.

Carmant L. Rôle de la tomographie par émission depositron dans les désordres neurologiques. Comitéde travail sur la technologie du positron. Montréal:Fédération des médecins spécialistes du Québec,September 2000.

Couillard P. La tomographie par émission depositrons (TEP) en pratique neurochirurgicale:indications et besoins. Comité de travail sur latechnologie du positron. Montréal: Fédération desmédecins spécialistes du Québec, September 2000.

Lacasse Y. Tomographie par émission de positronsen oncologie pulmonaire. Comité de travail sur latechnologie du positron. Montréal: Fédération desmédecins spécialistes du Québec, September 2000.

Laplante J. L'utilisation du Positron en oncologie.

Comité de travail sur la technologie du positron.Montréal: Fédération des médecins spécialistes duQuébec, September 2000.

Létourneau R. Évaluation du rôle du positron enregard des tumeurs solides (à l'exclusion destumeurs thoraciques, neurologiques et génito-urinaires). Comité de travail sur la technologie dupositron. Montréal: Fédération des médecinsspécialistes du Québec, September 2000.

ASSESSMENT REPORTS

Adams E. Evaluating diagnostic tests: A guide to theliterature. Boston, MA: Management Decisionand Research Center, Health Services Research& Development Service, Office of Research andDevelopment, Department of Veterans Affairs:1997. Technology Assessment Program ReportNo.4.

Adams E, Asua J, Conde OJ, Erlichman M, Flynn K,Hurtado-Saracho I, eds. On behalf of INAHTA.Positron emission tomography: Experience withPET and synthesis of the evidence. Stockholm:International Network of Agencies for HealthTechnology Assessment, 1999.

Adams E, Flynn K, eds. Positron Emission Tomogra-phy: descriptive analysis of experience with PETin VA. Boston, MA: Management Decision andResearch Center, Health Services Research andDevelopment Service, Office of Research andDevelopment, Department of Veterans Affairs,1998. Technology Assessment Program ReportNo.10.

Baffert S, Lajonchère JP, Féry-Lemonnier E, CÉDIT.Tomographie par émission de positrons: étatd'avancement du dossier. Paris: AssistancePublique - Hôpitaux de Paris, CÉDIT, 1999.

Blue Cross Blue Shield of Massachusetts. Policy 358:PET scans - positron emission tomography.2000 [consulted on November 29, 2000]. URL:http://www.bcbsma.com/hresource/358.htm.


References228

Catalan Agency for Health Technology Assessment(CAHTA). Assessments of clinical PET - vari-ous indications (internal document). Barcelona:CAHTA, 1993.

Cowley D, Corabian P, Hailey D. Functional diagnos-tic imaging in the assessment of myocardial vi-ability. Part 1 - Overview; Part 2 - Details ofstudies. Edmonton: Alberta Heritage Foundationfor Medical Research, 1999.

Medicare Services Advisory Committee. Report of thereview of positron emission tomography. Com-monwealth of Australia, Department of Healthand Aged Care (Diagnostics and TechnologyBranch): August 2000. URL: http://www.health.gov.au/haf/msac.

Flynn K, Adams E, eds. Positron emission tomogra-phy: Descriptive analysis of experience withPET in VA and systematic reviews: FDG-PETas a diagnostic test for cancer and Alzheimer’sdisease. Boston, MA: Management Decision andResearch Center, Health Services Research andDevelopment Service, Office of Research andDevelopment, Department of Veterans Affairs,1996.

Ghersi D, Howard K, Irwig L, Salkeld G, Simes J, eds.Commonwealth review of positron emission to-mography - final assessment report. Common-wealth of Australia. Medicare Services AdvisoryCommittee - Department of Health and AgedCare, March 2000.

HAYES Inc. Positron emission tomography (PET) forother malignancies: colorectal cancer, malignantlymphoma, malignant melanoma, breast cancer -Draft update - Executive summary - Internaldocument. Lansdale, PA: Winifred S. Hayes,Inc., 1999. Report number: none.

HAYES Inc. Positron emission tomography (PET) forcardiac applications - Draft update - Executivesummary - Internal document. Lansdale, PA:Winifred S. Hayes, Inc., 1999. Report number:none.

Lozano JP, de la Blanca E, Briones P, eds. Tomografia

de emision de positrones (PET): sintesis deinvestigacion sobre efectividad en diferentesindicaciones clinicas. Seville: Agencia deEvaluacion de Tecnologias Sanitarias deAndalucia, 2000.

Minnesota Health Technology Advisory Committee.Positron emission tomography (PET) for on-cologic applications [Web page]. October 2000.URL: http://www.health.state.mn.us/htac/pet.htm.

Robert G, Milne R, eds. Health Technology Assess-ment, NHS R&D HTA Programme. Positronemission tomography: establishing priorities forhealth technology assessment. Health TechnolAssessment 1999;3(16).

Tunis S, Stojak M, Richardson S et al., eds. FDGpositron emission tomography - CAG-00065 -Decision memorandum. Baltimore, MD: U.S.Department of Health and Human Services,Health Care Financing Administration, coverageand Analysis Group, 2000.

INDIVIDUAL REFERENCES

Acland KM, O’Doherty MJ, Russell-Jones R. Thevalue of positron emission tomography scanningin the detection of subclinical metastatic mela-noma. J Am Acad Dermatol 2000;42(4):606-11.

Akinboboye OO, Idris O, Cannon PJ, Bergmann SR.Usefulness of positron emission tomography indefining myocardial viability in patients referredfor cardiac transplantation. Am J Cardiol 1999Apr;83(8):1271-4,A9.

Allen-Auerbach M, Schoder H, Johnson J, et al. Rela-tionship between coronary function by positronemission tomography and temporal changes inmorphology by intravascular ultrasound (IVUS)in transplant recipients. J Heart Lung Transplant1999;18(3):211-9.

Allman K, Wieland D, Muzik O, et al. Carbon-11 hy-droxyephedrine with positron emission tomo-graphy for serial assessment of cardiac adrener-gic neuronal function after acute myocardial in-


References229

farction in humans. J Am Coll Cardiol1993;22:368-75.

Almkvist O, Winblad B. Early diagnosis of Alzheimerdementia based on clinical and biological fac-tors. Eur Arch Psychiatry Clin Neurosci1999;249 Suppl 3:3-9.

Alzheimer Society of Canada. La maladie d'Alzheimer- Statistiques. [Web page], 2001. URL:http://www.alzheimer.ca/french/disease/ stats-people.htm. (consulted in April 2001).

Anzai Y, Carroll WR, Quint DJ et al. Recurrence ofhead and neck cancer after surgery or irradia-tion: prospective comparison of 2-deoxy-2-[F-18]fluoro-D-glucose PET and MR imaging di-agnosis. Radiology 1996;200:135-141.

Arrighi JA, Ng CK, Dey HM, Wackers FJ, Soufer R.Effect of left ventricular function on the assess-ment of myocardial viability by technectium-99m sestamibi and correlation with positronemission tomography in patients with healedmyocardial infarcts or stable angina pectoris, orboth. Am J Cardiol 1997;80(8):1007-13.

Avril N, Scheidhauer K, Kuhn W. Breast cancer. In:Wieler HJ, Coleman RE, eds. PET in clinicaloncology. Damstadt: Steinkopff Verlag, 2000a,pp. 355-72.

Avril N, Weber W, Breul J. Prostate cancer. In: WielerHJ, Coleman RE, eds. PET in clinical oncology.Damstadt: Steinkopff Verlag, 2000b, pp. 287-98.

Bader JB, Samnick S, Moringlane JR, Feiden W,Schaefer A, Kremp S, et al. Evaluation of l-3-[123I]iodo-alpha-methyltyrosine SPET andfluorodeoxyglucose PET in the detection andgrading of recurrences in patients pretreated forgliomas at follow-up: a comparative study withstereotactic biopsy. Eur J Nucl Med. 1999Feb;26(2):144-51.

Baer FM, Voth E, Deutsch HJ, Schneider CA, HorstM, de Vivie ER, et al. Predictive value of lowdose dobutamine transesophageal echocardi-ography and fluorine-18 fluorodeoxyglucose

positron emission tomography for recovery ofregional left ventricular function after successfulrevascularization. J Am Coll Cardiol 1996Jul;28(1):60-9.

Baller D, Notohamiprodjo G, Gleichmann U,Holzinger J, Weise R, Lehmann J. Improvementin coronary flow reserve determined by positronemission tomography after 6 months of choles-terol-lowering therapy in patients with earlystages of coronary atherosclerosis. Circulation1999;99(22):2871-5.

Bangerter M, Kotzerke J, Griesshammer M, Elsner K,Reske SN, Bergmann L. Positron emission to-mography with 18-fluorodeoxyglucose in thestaging and follow-up of lymphoma in the chest.Acta Oncol 1999;38(6):799-804.

Bangerter M, Moog F, Buchmann I, Kotzerke J,Griesshammer M, Hafner M, Elsner K, Frick-hofen N, Reske SN, Bergmann L. Whole-body2-[18F]-fluoro-2-deoxy-D-glucose positronemission tomography (FDG-PET) for accuratestaging of Hodgkin's disease. Ann Oncol. 1998Oct;9(10):1117-22.

Banta HD, McNeil BJ. Evaluation of the CAT scannerand other diagnostic technologies. Health CareManage Rev 1978 Winter;3(1):7-19.

Bares R, Galonska P, Dempke W, Handt S, Bull U,Osieka R. Somatostatin receptor scintigraphy inmalignant lymphoma: first results and compari-son with glucose metabolism measured by posi-tron-emission tomography. Horm Metab ResSuppl. 1993;27:56-8.

Barkheet SM, Amin T, Alia A-G, Kuzo R, Powe J. F-18 FDG uptake in benign esophageal disease.Clin Nucl Med 1999;24(12):995-7.

Barkheet SM, Powe J. Benign causes of 18-FDG up-take on the whole body imaging. Sem Nucl Med1998;XXVIII(4):352-58.

Barkheet SMB, Powe J, Kandil A, Ezzat A, Rostom A,Amartey J. F-18 FDG uptake in breast infectionand inflammation. Clin Nucl Med 2000;25:100-3.


References230

Baum RA, Rutter CM, Sunshine JH, Blebea JS, BlebeaJ, Carpenter JP, et al. Multicenter trial to evalu-ate vascular magnetic resonance angiography ofthe lower extremity. American College of Radi-ology Rapid Technology Assessment Group.JAMA 1995 Sep;274(11):875-80.

Bax JJ, Cornel JH, Visser FC, Fioretti PM, van LingenA, Reijs AE, et al. Prediction of recovery ofmyocardial dysfunction after revascularization.Comparison of fluorine-18 fluorodeoxyglu-cose/thallium-201 SPECT, thallium-201 stress-reinjection SPECT and dobutamine echocardi-ography. J Am Coll Cardiol 1996Sep;28(3):558-64.

Bax JJ, Poldermans D, Elhendy A, Boersema E, Ra-himtoola SH. Sensitivity, specificity and predic-tive accuracies of various noninvasive tech-niques for detecting hibernating myocardium.Curr Probl Cardiol 2001 Feb;26(2):141-86.

Bax JJ, van Eck-Smit BL, van der Wall EE. Assess-ment of tissue viability: clinical demand andproblems. Eur Heart J 1998 Jun;19(6):847-58.

Beanlands RS, deKemp RA, Smith S, Johansen H,Ruddy TD. F-18-fluorodeoxyglucose PET im-aging alters clinical decision making in patientswith impaired ventricular function. Am J Car-diol 1997 Apr;79(8):1092-5.

Beanlands RS. Outcome and cost-effectiveness ofFDG-PET in LV dysfunction (PARR 2). Uni-versity of Ottawa Heart Institute, 2000-2001.

Beck JR, Kassirer JP, Pauker SG. A convenient ap-proximation of life expectancy (the "DEALE").I. Validation of the method. Am J Med1982;73(6):883-8.

Begg CB, Greenes RA, Iglewicz B. The influence ofuninterpretability on the assessment of diagnos-tic tests. J Chronic Dis 1986;39(8):575-84.

Begg CB, Greenes RA. Assessment of diagnostic testswhen disease verification is subject to selectionbias. Biometrics 1983 Mar;39(1):207-15.

Begg CB, McNeil BJ. Assessment of radiologic tests:control of bias and other design considerations.Radiology 1988 May;167(2):565-9.

Begg CB. Biases in the assessment of diagnostic tests.Stat Med 1987 Jun;6(4):411-23.

Begg CB. Experimental design of medical imaging tri-als. Issues and options. Invest Radiol 1989Nov;24(11):934-6.

Begg CB. Methodologic standards for diagnostic testassessment studies. J Gen Intern Med 1988 Sep-Oct;3(5):518-20.

Beller GA. Noninvasive assessment of myocardial vi-ability. N Engl J Med 2000;343(20):1488-90.

Bénard F, Smith RJ, Hustinx R, Karp JS, Alavi A.Clinical evaluation of processing techniques forattenuation correction with 137Cs in whole-bodyPET imaging. J Nucl Med 1999a;40(8):1257-63.

Bénard F, Sterman D, Smith RJ, Kaiser LR, AlbeldaSM, Alavi A. Prognostic value of FDG PET im-aging in malignant pleural mesothelioma. J NuclMed 1999b;40(8):1241-5.

Bender H, Straehler-Pohl HJ. PET in head and necktumors. In: Wieler HJ, Coleman RE, eds. PET inclinical oncology. Damstadt: Steinkopff Verlag;2000. Pp. 155-68.

Bengel FM, Schwaiger M. Nuclear medicine studies ofthe heart. Eur Radiol 1998;8(9):1698-706.

Bengel FM, Ucberfuhr P, Ziegler S, Nekolla S,Reichart B, Schwaiger M. Serial assessment ofsympathetic reinnervation after orthotopic hearttransplantation. A longitudinal study using PETand C-11 Hydroxyephedrine. Circulation1999;99:1866-71.

Bennett DA. Part I. Epidemiology and public healthimpact of Alzheimer's disease. Dis Mon2000a;46(10):657-65.

Bennett DA. Part II. Clinical diagnosis and course ofAlzheimer's disease. Dis Mon 2000b;46(10):666-86.


References231

Berlangieri SU, Scott AM, Knight SR, et al. F-18fluorodeoxyglucose positron emission tomogra-phy in the non-invasive staging of non-small celllung cancer. Eur J Cardio-Thoracic Surg1999;16(suppl. 1):S25-S30.

Blessing C, Feine U, Geiger L, Carl M, Rassner G,Fierlbeck G. Positron emission tomography andultrasonography. A comparative retrospectivestudy assessing the diagnostic validity in lymphnode metastases of malignant melanoma. ArchDermatol 1995;131:1394-8.

Bohuslavizki KH, Klutmann S, Kröger S, SonnemannU, Buchert R, Werner JA, et al. FDG PET de-tection of unknown primary tumors. J Nucl Med2000;41(5):816-22.

Bonow RO. Identification of viable myocardium.Circulation. 1996 Dec;94(11):2674-80.

Braunwald E, Rutherford JD. Reversible ischemic leftventricular dysfunction: evidence for the "hiber-nating myocardium". J Am Coll Cardiol 1986Dec;8(6):1467-70.

Brundage BH, Massie BM, Botvinick EH. Improvedregional ventricular function after successfulsurgical revascularization. J Am Coll Cardiol1984 Apr;3(4):902-8.

Buchmann I, Reinhardt M, Elsner K, Bunjes D, Alte-hoefer C, Finke J, et al. 2-(fluorine-18)fluoro-2-deoxy-D-glucose positron emission tomographyin the detection and staging of malignant lym-phoma. A bicenter trial. Cancer 2001 Mar1;91(5):889-99.

Bury T, Barreto A, Daenen F, Barthelemy N, Ghaye B,Rigo P. Fluorine-18 deoxyglucose positronemission tomography for the detection of bonemetastases in patients with non small cell lungcancer. Eur J Nucl Med 1998;25(9):1244-7.

Calkins H, Allman K, Bolling S, et al. Correlationbetween scintigraphic evidence of regional sym-pathetic neuronal dysfunction and ventricular re-fractoriness in the human heart. Circulation1993a;88:172-9.

Calkins H, Lehmann MH, Allman K, et al. Scinti-graphic pattern of regional cardiac sympathicinnervation in patients with familial long QTsyndrome using positron emission tomography.Circulation 1993b;87:1616-21.

Camici PG. Positron emission tomography and myo-cardial imaging. Heart 2000 Apr;83(4):475-80.

Campisi R, Czernin J, Schoder H, et al. L-argininenormalizes coronary vasomotion in long-termsmokers. Circulation 1999b;99:491-497.

Campisi R, Nathan L, Pampaloni MH, et al. Effect ofchronic hormone replacement therapy on coro-nary vasomotion in post-menopausal women.Circulation 1999a;100(suppl I):I-221.

Carr R, Barrington SF, Madan B, O'Doherty MJ,Saunders CA, van der Walt J, Timothy AR. De-tection of lymphoma in bone marrow by whole-body positron emission tomography. Blood.1998 May 1;91(9):3340-6.

Carrel T, Jenni R, Haubold-Reuter S, von SchulthessG, Pasic M, Turina M. Improvement of severelyreduced left ventricular function after surgicalrevascularization in patients with preoperativemyocardial infarction. Eur J Cardiothorac Surg1992;6(9):479-84.

Chalmers TC. PET scans and technology assessment.JAMA. 1988 Nov 11;260(18):2713-5.

Chatterjee K, Swan HJ, Parmley WW, Sustaita H,Marcus HS, Matloff J. Influence of direct myo-cardial revascularization on left ventricularasynergy and function in patients with coronaryheart disease. With and without previous myo-cardial infarction. Circulation 1973 Feb;47(2):276-86.

Chee MW, Morris HH 3rd, Antar MA, et al. Presurgi-cal evaluation of temporal lobe epilepsy usinginterictal temporal spikes and positron emissiontomography. Arch Neurol 1993;50(1):45-8.

Choi JY, Lee KH, Shim YM, Lee KS, Kim JJ, KimSE, et al. Improved detection of individual nodalinvolvement in squamous cell carcinoma of the


References232

esophagus by FDG PET. J Nucl Med 2000;41(5):808-15.

Choudhury L, Elliott P, Rimoldi O, et al. Transmuralmyocardial blood flow distribution in hypertro-phic cardiomyopathy and effect of treatment.Basic Res Cardiol 1999;94(1):49-59.

Choudhury L, Guzzetti S, Lefroy DC, et al. Myocar-dial beta adrenoceptors and left ventricularfunction in hypertrophic cardiomyopathy. Heart1996;75:50-4.

Cook GJ, Fogelman I. Skeletal metastases from breastcancer: imaging with nuclear medicine. SeminNucl Med 1999;29(1):69-79.

Cooper LS, Chalmers TC, McCally M, Berrier J, SacksHS. The poor quality of early evaluations ofmagnetic resonance imaging. JAMA. 1988 Jun10;259(22):3277-80.

Cowley D, Corabian P, Hailey D. Functional diagnos-tic imaging in the assessment of myocardial vi-ability. Part 1 - Overview; Part 2 - Details ofstudies. Edmonton: Alberta Heritage Foundationfor Medical Research, 1999.

Crippa F, Agresti R, Donne VD, Pascali C, Bogni A,Chiesa C, et al. The contribution of positronemission tomography (PET) with 18F- fluoro-deoxyglucose (FDG) in the preoperative detec-tion of axillary metastases of breast cancer: theexperience of the National Cancer Institute ofMilan. Tumori 1997;83(2):542-3.

Crippa F, Agresti R, Seregni E, Greco M, Pascali C,Bogni A, et al. Prospective evaluation of fluo-rine-18-FDG PET in presurgical staging of theaxilla in breast cancer. J Nucl Med 1998;39(1):4-8.

Crippa F, Leutner M, Belli F, Gallino F, Greco M, Pi-lotti S, et al. Which kinds of lymph node metas-tases can FDG PET detect? A clinical study inmelanoma. J Nucl Med 2000;41(9):1491-4.

Cummings SR, Lillington GA, Richard RJ. Estimatingthe probability of malignancy in solitary pulmo-nary nodules. A Bayesian approach. Am Rev

Respir Dis 1986;134(3):449-52.

Czernin J, Barnard J, Sun K, et al. Effect of short-termcardiovascular conditioning and low fat diet onmyocardial blood flow and flow reserve.Circulation 1995;92:197-204.

de Wit M, Bumann D, Beyer W, Herbst K, Clausen M,Hossfeld DK. Whole-body positron emissiontomography (PET) for diagnosis of residualmass in patients with lymphoma. Ann Oncol.1997; 8 Suppl 1:57-60.

De Witte O, Lefranc F, Levivier M, Salmon I, BrotchiJ, Goldman S. FDG-PET as a prognostic factorin high-grade astrocytoma. J Neurooncol. 2001Sep;49(2):157-63.

Delbeke D, Lawrence SK, Abou-Khalil BW,Blumenkopf B, Kessler RM. Postsurgical out-come of patients with uncontrolled complexpartial seizures and temporal lobe hypometabo-lism on 18FDG-positron emission tomography.Invest Radiol 1996;31(5):261-6.

Delbeke D,Vitola JV, Sandler MP, Arildsen RC, Pow-ers TA, Wright JK Jr, et al. Staging recurrentmetastatic colorectal carcinoma with PET. JNucl Med 1997;38:1196-1201.

Derlon JM, Chapon F, Noel MH, Khouri S, Benali K,Petit-Taboue MC, Houtteville JP, Chajari MH,Bouvard G. Non-invasive grading of oligoden-drogliomas: correlation between in vivo meta-bolic pattern and histopathology. Eur J NuclMed. 2000 Jul;27(7):778-87.

Deshmukh A, Scott JA, Palmer EL, Hochberg FH,Gruber M et Fischman AJ. Impact of fluorode-oxyglucose positron emission tomography onthe clinical management of patients with glioma.Clin Nucl Med 1996;21:720-5.

Devous MD Sr, Thisted RA, Morgan GF, Leroy RF,Rowe CC. SPECT brain imaging in epilepsy: ameta-analysis. J Nucl Med 1998;39(2):285-93.

Dewan NA, Reeb SD, Gupta NC, Gobar LS, Scott WJ.PET-FDG imaging and transthoracic needlelung aspiration biopsy in evaluation of pulmo-


References233

nary lesions. A comparative risk-benefit analy-sis. Chest 1995;108(2):441-6.

Dewan NA, Shehan CJ, Reeb SD, Gobar LS, Scott WJ,Ryschon K. Likelihood of malignancy in a soli-tary pulmonary nodule: comparison of Bayesiananalysis and results of FDG-PET scan. Chest1997;112(2):416-22.

Dhital K, Saunders CAB, Seed PT, O'Doherty MJ,Dussek J. [F-18]fluorodeoxyglucose positronemission tomography and its prognostic value inlung cancer. Eur J Cardio Thorac Surg 2000;18(4):425-8.

Di Carli MF, Asgarzadie F, Schelbert HR, BrunkenRC, Laks H, Phelps ME, Maddahi J. Quantita-tive relation between myocardial viability andimprovement in heart failure symptoms after re-vascularization in patients with ischemic car-diomyopathy. Circulation 1995 Dec;92(12):3436-44.

Di Carli MF, Davidson M, Little R, Khanna S, ModyFV, Brunken RC, et al. Value of metabolic im-aging with positron emission tomography forevaluating prognosis in patients with coronaryartery disease and left ventricular dysfunction.Am J Cardiol 1994 Mar;73(8):527-33.

Di Carli MF, Maddahi J, Rokhsar S, Schelbert HR, Bi-anco-Batlles D, Brunken RC, Fromm B. Long-term survival of patients with coronary arterydisease and left ventricular dysfunction: impli-cations for the role of myocardial viability as-sessment in management decisions. J ThoracCardiovasc Surg 1998 Dec;116(6):997-1004.

Di Carli MF, Tobes MC, Manger T, et al. Effects ofcardiac sympathetic innervation on coronaryblood flow. N Engl J Med 1997;336:1208-15.

Dietlein M, Krug B, Groth W, Smolarz K, ScheidhauerK, Psaras T, et al. Positron emission tomographyusing 18F-fluorodeoxyglucose in advancedstages of malignant melanoma: a comparison ofultrasonographic and radiological methods ofdiagnosis. Nucl Med Commun 1999;20(3):255-61.

Dietlein M, Weber K, Gandjour A, Moka D, TheissenP, Lauterbach KW, Schicha H. Cost-effectiveness of FDG-PET for the managementof potentially operable non small cell lung can-cer: priority for a PET-based strategy afternodal-negative CT results. Eur J Nucl Med.2000b Nov;27(11):1598-609.

Dietlein M, Weber K, Gandjour A, Moka D, TheissenP, Lauterbach KW, Schicha H. Cost-effectiveness of FDG-PET for the managementof solitary pulmonary nodules: a decision analy-sis based on cost reimbursement in Germany.Eur J Nucl Med. 2000a Oct;27(10):1441-56.

Dillemans B, Deneffe G, Verschakelen J, Decramer M.Value of computed tomography and mediastino-scopy in preoperative evaluation of mediastinalnodes in non small cell lung cancer. A study of569 patients. Eur J Cardiothorac Surg 1994;8(1):37-42.

Di Martino E, Nowak B, Hassan HA, et al. Diagnosisand staging of head and neck cancer - A com-parison of modern imaging modalities (positronemission tomography, computed tomography,color-coded duplex sonography) with panendo-scopic and histopathologic findings. Arch OtolarHead Neck Surgery 2000;126(12):1457-61.

Dimitrakopoulou-Strauss A. Colorectal carcinomas. In:Wieler HJ, Coleman RE, eds. PET in clinicaloncology. Damstadt: Steinkopff Verlag, 2000,pp. 235-254.

Duong TH, Hendi P, Fonarow G, Asgarzadie F,Stevenson L, Di Carli M, et al. Rôle of positronemission tomography assessment of myocardialviability in the management of patients who arereferred for cardiac transplantation. Circulation1995 October 15;92(8):I-123.

Dreyfus GD, Duboc D, Blasco A, Vigoni F, Dubois C,Brodaty D, et al. Myocardial viability assess-ment in ischemic cardiomyopathy: benefits ofcoronary revascularization. Ann Thorac Surg1994 Jun;57(6):1402-7;discussion 1407-8.

Dupont S, Semah F, Clemenceau S, Adam C, BaulacM, Samson Y. Accurate prediction of postop-


References234

erative outcome in mesial temporal lobe epi-lepsy: a study using positron emission tomogra-phy with 18fluorodeoxyglucose. Arch Neurol2000 Sep;57(9):1331-6.

Dwamena BA, Sonnad SS, Angobaldo JO, Wahl RL.Metastases from non small cell lung cancer: me-diastinal staging in the 1990s--meta-analyticcomparison of PET and CT. Radiology 1999;213(2):530-6.

Eary JF, Mankoff DA, Spence AM, Berger MS, Ol-shen A, Link JM, O'Sullivan F, Krohn KA. 2-[C-11]thymidine imaging of malignant braintumors. Cancer Res. 1999 Feb 1;59(3):615-21.

Eigtved A, Andersson AP, Dahlstrom K, Rabol A, Jen-sen M, Holm S, et al. Use of fluorine-18 fluoro-deoxyglucose positron emission tomography inthe detection of silent metastases from malignantmelanoma. Eur J Nucl Med 2000;27(1):70-5.

Eitzman D, al-Aouar Z, Kanter HL, vom Dahl J, KirshM, Deeb GM, Schwaiger M. Clinical outcomeof patients with advanced coronary artery dis-ease after viability studies with positron emis-sion tomography. J Am Coll Cardiol 1992Sep;20(3):559-65.

Elefteriades JA, Tolis G, Levi E, Mills LK, Zaret BL.Coronary artery bypass grafting in severe leftventricular dysfunction: excellent survival withimproved ejection fraction and functional state. JAm Coll Cardiol 1993 Nov;22(5):1411-7.

Engel J Jr, Henry TR, Risinger MW, et al. Presurgicalevaluation for partial epilepsy: relative contri-but ions of chronic depth-electroderecordings versus FDG-PET and scalp-sphenoidal ictal EEG. Neurology 1990 40:1670-7.

Epilepsy Canada. Épidémiologie [Web web]. 1998;URL:http://www.epilepsy.ca/fran/epidemioFR.html.(consulted in April 2001).

Erasmus JJ, McAdams HP, Rossi SE, Goodman PC,Coleman RE, Patz EF. FDG PET of pleural ef-fusions in patients with non small cell lung can-

cer. Am J Roentgenol 2000;175(1):245-9.

Evans EW. Resection for bronchial carcinoma in theelderly. Thorax 1973 Jan;28(1):86-8.

Farber LA, Benard F, Machtay M, Smith RJ, WeberRS, Weinstein GS, et al. Detection of recurrenthead and neck squamous cell carcinomas afterradiation therapy with 2-18F-fluoro-2-deoxy-D-glucose positron emission tomography. Laryn-goscope 1999;109(6):970-5.

Fath-Ordoubadi F, Pagano D, Marinho NV, KeoghBE, Bonser RS, Camici PG. Coronary revascu-larization in the treatment of moderate and se-vere postischemic left ventricular dysfunction.Am J Cardiol 1998;82(1):26-31.

Feinstein AR. Meta-analysis and meta-analytic moni-toring of clinical trials. Stat Med1996;15(12):1273-80;discussion 1281-3.

Feinstein AR. Problems, pitfalls, and opportunities inl o n g - t e r m r a n d o m i z e d t r i a l s .Arzneimittelforschung 1989;39(8A):980-3;discussion 984-5.

Ferrie CD, Maisey M, Cox T, et al. Focal abnormali-ties detected by 18FDG PET in epileptic en-cephalopathies. Arch Dis Child 1996;75(2):102-7.

Flamen P, Lerut A, Van Cutsem E, De Wever W,Peeters M, Stroobants S, et al. Utility of positronemission tomography for the staging of patientswith potentially operable esophageal carcinoma.J Clin Oncol 2000;18(18):3202-10.

Flamen P, Stroobants S, Van Cutsem E, Dupont P,Bormans G, De Vadder N, et al. Additionalvalue of whole-body positron emission tomo-graphy with fluorine-18-2-fluoro-2-deoxy-D-glucose in recurrent colorectal cancer. J ClinOncol 1999;17:894-901.

Flamen P, VanCutsem E, Mortelmans L. A new im-aging technique for colorectal cancer: positronemission tomography. Semin Oncol 2000;27(5):22-9.


References235

Fryback DG, Thornbury JR. The efficacy of diagnosticimaging. Med Decis Making 1991;11(2):88-94.

Gambhir SS, Barrio JR, Phelps ME, et al. Imagingadenoviral-directed reporter gene expression inliving animals with positron emission tomogra-phy. Proc Natl Acad Sci U S A 1999;96:2333-8

Gambhir SS, Czernin J, Schwimmer J, SilvermanDHS, Coleman E, Phelps ME. A tabulatedsummary of the FDG PET literature. J Nucl Med2001;42(5, suppl.):1S-93S.

Gambhir SS, Hoh CK, Phelps ME, Madar I, MaddahiJ. Decision tree sensitivity analysis for cost-effectiveness of FDG-PET in the staging andmanagement of non small-cell lung carcinoma. JNucl Med 1996;37(9):1428-36.

Gambhir SS, Shepherd JE, Shah BD, et al. Analyticaldecision model for the cost-effective manage-ment of solitary pulmonary nodules. J Clin On-col 1998;16(6):2113-25.

Garber AM, Solomon NA. Cost-effectiveness of alter-native test strategies for the diagnosis of coro-nary artery disease. Ann Intern Med1999;130(9): 719-28.

Gatsonis CA. Random-effects models for diagnosticaccuracy data. Acad Radiol 1995 Mar;2(Suppl1):S14-21;discussion S57-67, S61-4.

Gerber BL, Vanoverschelde JL, Bol A, Michel C,Labar D, Wijns W, Melin JA. Myocardial bloodflow, glucose uptake, and recruitment ofinotropic reserve in chronic left ventricularischemic dysfunction. Implications for thepathophysiology of chronic myocardial hiberna-tion. Circulation 1996 Aug;94(4):651-9.

Ginsberg RJ, Hill LD, Eagan RT, et al. Modern thirty-day operative mortality for surgical resections inlung cancer. J Thorac Cardiovasc Surg 1983;86(5):654-8.

Goldstein DS, Chang PC, Eisenhofer G, et al. Positronemission tomographic imaging of cardiac sym-pathetic innervation and function. Circulation1990;81:1606-21.

Gould K, Martucci J, Goldberg D, et al. Short-termcholesterol lowering decreases size and severityof perfusion abnormalities by positron emissiontomography after dipyridamole in patients withcoronary artery disease: a potential non invasivemarker of healing coronary endothelium. Circu-lation 1994;89:1530-38.

Gould K, Ornish D, Scherwitz L, et al. Changes inmyocardial perfusion abnormalities by positronemission tomography after long-term, intenserisk factor modification. JAMA 1995;274:894-901.

Gould MK, Maclean CC, Kuschner WG, Rydzak CE,Owens DK. Accuracy of positron emission to-mography for diagnosis of pulmonary nodulesand mass lesions: a meta-analysis. JAMA 2001;285(7):914-24.

Grandin C, Wijns W, Melin JA, Bol A, Robert AR,Heyndrickx GR, et al. Delineation of myocardialviability with PET. J Nucl Med 1995 Sep;36(9):1543-52.

Greven KM, Williams DW 3rd, Keyes JW Jr, McGuirtWF, Watson NE Jr, Case LD. Can positronemission tomography distinguish tumor recur-rence from irradiation sequelae in patientstreated for larynx cancer? Cancer J Sci Am.1997 Nov-1997 Dec 31;3(6):353-7.

Gropler RJ, Geltman EM, Sampathkumaran K, PerezJE, Moerlein SM, Sobel BE, et al. Functional re-covery after coronary revascularization forchronic coronary artery disease is dependent onmaintenance of oxidative metabolism. J AmColl Cardiol 1992 Sep;20(3):569-77.

Gropler RJ, Geltman EM, Sampathkumaran K, PerezJE, Schechtman KB, Conversano A, et al. Com-parison of carbon-11-acetate with fluorine-18-fluorodeoxyglucose for delineating viable myo-cardium by positron emission tomography. J AmColl Cardiol 1993 Nov;22(6):1587-97.

Gross BH, Glazer GM, Orringer MB, Spizarny DL,Flint A. Bronchogenic carcinoma metastatic tonormal-sized lymph nodes: frequency and sig-nificance. Radiology 1988;166(1 Pt 1):71-4.


References236

Gross MW, Weber WA, Feldmann HJ, Bartenstein P,Schwaiger M, Molls M. The value of F-18-fluorodeoxyglucose PET for the 3-D radiationtreatment planning of malignant gliomas. Int JRadiat Oncol Biol Phys 1998;41(5):989-95.

Guethlin M, Kasel AM, Coppenrath K, et al. Delayedresponse of myocardial flow reserve to lipid-lowering therapy with fluvastatin. Circulation1999;99:475-81.

Gupta NC, Graeber GM, Bishop HA. Comparative ef-ficacy of positron emission tomography withfluorodeoxyglucose in evaluation of small (<1cm), intermediate (1 to 3 cm), and large (>3 cm)lymph node lesions. Chest 2000;117(3):773-8.

Gupta NC, Graeber GM, Rogers JS 2nd, Bishop HA.Comparative efficacy of positron emission to-mography with FDG and computed tomographicscanning in preoperative staging of non smallcell lung cancer. Ann Surg 1999;229(2):286-91.

Haas F, Haehnel CJ, Picker W, Nekolla S, Martinoff S,Meisner H, Schwaiger M. Preoperative positronemission tomographic viability assessment andperioperative and postoperative risk in patientswith advanced ischemic heart disease. J Am CollCardiol 1997 Dec;30(7):1693-700.

Hanasono MM, Kunda LD, Segall GM, et al. Uses andlimitations of FDG positron emission tomogra-phy in patients with head and neck cancer. La-ryngoscope 1999;109:880-5.

Hara T, Inagaki K, Kosaka N, Morita T. Sensitive de-tection of mediastinal lymph node metastasis oflung cancer with C-11-choline PET. J Nucl Med2000;41(9):1507-13.

Hartman GW, Hattery RR, Written D, Williamson BJr. Mortality during excretory urography. MayoClinic experience. Am J Roentgenol 1982Nov;139(5):919-22

Hellman C, Schmidt DH, Kamath ML, Anholm J, BlauF, Johnson WD. Bypass graft surgery in severeleft ventricular dysfunction. Circulation 1980Aug;62(2 Pt 2):I103-10.

Higashi K, Ueda Y, Ayabe K, Sakurai A, Seki H,Nambu Y, et al. FDG PET in the evaluation ofthe aggressiveness of pulmonary adenocarci-noma: correlation with histopathological fea-tures. Nucl Med Commun 2000;21(8):707-14.

Hillman BJ, Gatsonis C, Sullivan DC. American Col-lege of Radiology Imaging Network: new na-tional cooperative group for conducting clinicaltrials of medical imaging technologies. Radiol-ogy 1999 Dec;213(3):641-5.

Hochberg MS,Parsonnet V, Gielchinsky I, HussainSM. Coronary artery bypass grafting in patientswith ejection fractions below forty percent.Early and late results in 466 patients. J ThoracCardiovasc Surg 1983 Oct;86(4):519-27.

Hoh CK, Glaspy J, Rosen P, Dahlbom M, Lee SJ,Kunkel L, Hawkin RA, Maddahi J, Phelps ME.Whole-body FDG-PET imaging for staging ofHodgkin's disease and lymphoma. J Nucl Med1997 Mar;38(3):343-8.

Holder WD Jr, White RL Jr, Zuger JH, Easton EJ Jr,Greene FL. Effectiveness of positron emissiontomography for the detection of melanoma me-tastases. Ann Surg 1998;227(5):764-9;discussion 769-71.

Hueltenschmidt B, Sautter-Bihl ML, Lang O, MaulFD, Fischer J, Mergenthaler HG, et al. Wholebody positron emission tomography in thetreatment of Hodgkin disease. Cancer 2001 Jan15;91(2):302-10.

Hwang SI, Kim JH, Park SW, Han MH, Yu IK, LeeSH, et al. Comparative analysis of MR imaging,positron emission tomography, and ictal single-photon emission CT in patients with neocorticalepilepsy. AJNR Am J Neuroradiol 2001May;22(5):937-46.

Imdahl A, Reinhardt MJ, Nitzsche EU, Mix M,Dingeldey A, Einert A, et al. Impact of 18F-FDG-positron emission tomography for decisionmaking in colorectal cancer recurrences.Langenbecks Arch Surg 2000;385(2):129-34.


References237

Irwig L, Macaskill P, Glasziou P, Fahey M. Meta-analytic methods for diagnostic test accuracy. JClin Epidemiol 1995 Jan;48(1):119-30;discussion 131-2.

Jadvar H, Johnson DL, Segall GM. The effect of fluo-rine-18 fluorodeoxyglucose positron emissiontomography on the management of cutaneousmalignant melanoma. Clin Nucl Med2000;25(1):48-51.

Jaeschke R, Guyatt G et Sackett DL. User’s guides tothe medical literature III. How to use an articleabout a diagnostic test. A. Are the results of thestudy valid? Evidence-Based Medicine WorkingGroup. J Am Med Assoc 1994;271:389-91.

Jaeschke RZ, Meade MO, Guyatt GH, Keenan SP,Cook DJ. How to use diagnostic test articles inthe intensive care unit: diagnosing weanabilityusing f/Vt. Crit Care Med 1997 Sep;25(9):1514-21.

Jerusalem G, Beguin Y, Fassotte MF, Najjar F, PaulusP, Rigo P, Fillet G. Whole-body positron emis-sion tomography using 18F-fluorodeoxyglucosefor posttreatment evaluation in Hodgkin's dis-ease and non Hodgkin's lymphoma has higherdiagnostic and prognostic value than classicalcomputed tomography scan imaging. Blood1999;94(2):429-33.

Jerusalem G, Beguin Y, Fassotte MF, Najjar F, PaulusP, Rigo P et al. Persistent tumor 18F-FDG uptakeafter a few cycles of polychemotherapy is predic-tive of treatment failure in non-Hodgkin's lym-phoma. Haematologica 2000 Jun;85(6):613-8.

Jungehulsing M, Scheidhauer K, Damm M, PietrzykU, Eckel H, Schicha H, Stennert E. 2[F]-fluoro-2-deoxy-D-glucose positron emission tomogra-phy is a sensitive tool for the detection of occultprimary cancer (carcinoma of unknown primarysyndrome) with head and neck lymph nodemanifestation. Otolaryngol Head Neck Surg2000 Sep;123(3):294-301.

Kao CH, ChangLai SP, Chieng PU, et al. detection ofrecurrent or persistent nasopharyngeal carcino-mas after radiotherapy with 18-fluoro-2-

deoxyglucose positron emission tomography andcomparison with computed tomography. J ClinOncol 1998;16:3550-5.

Kawano M, Ichimiya A, Ogomori K, Kuwabara Y, Sa-saki M, Yoshida T, Tashiro N. Relationshipbetween both IQ and Mini-Mental State Exami-nation and the regional cerebral glucose me-tabolism in clinically diagnosed Alzheimer'sdisease: a PET study. Dement Geriatr Cogn Dis-ord 2001 Mar-Apr 30;12(2):171-6.

Kennedy JW, Kaiser GC, Fisher LD, Fritz JK, MyersW, Mudd JG, Ryan TJ. Clinical and angi-ographic predictors of operative mortality fromthe collaborative study in coronary artery sur-gery (CASS). Circulation 1981 Apr;63(4):793-802.

Kent DL et Larson EB. Diagnostic technology assess-ments: problems and prospects. Ann Intern Med1988 May;108(5):759-61.

Kiffer JD, Berlangieri SU, Scott AM, Quong G, FeigenM, Schumer W, et al. The contribution of 18F-fluoro-2-deoxy-glucose positron emission tomo-graphic imaging to radiotherapy planning inlung cancer. Lung Cancer 1998 Mar;19(3):167-77.

Kim RJ, Wu E, Rafael A, Chen EL, Parker MA, Si-monetti O, et al. The use of contrast-enhancedmagnetic resonance imaging to identify reversi-ble myocardial dysfunction. N Engl J Med 2000Nov;343(20):1445-53.

Kitsiou AN, Bacharach SL, Bartlett ML, Srinivasan G,Summers RM, Quyyumi AA, Dilsizian V. 13N-ammonia myocardial blood flow and uptake:relation to functional outcome of asynergic re-gions after revascularization. J Am Coll Cardiol1999 Mar;33(3):678-86.

Klemenz B, Taaleb KM. Lung Cancer. In: Wieler HJ,Coleman RE, eds. PET in clinical oncology.Damstadt: Steinkopff Verlag, 2000, pp. 193-210.

Knuuti MJ, Saraste M, Nuutila P, Harkonen R,Wegelius U, Haapanen A, et al. Myocardial vi-


References238

ability: fluorine-18-deoxyglucose positron emis-sion tomography in prediction of wall motionrecovery after revascularization. Am Heart J1994 Apr; 127(4 Pt 1):785-96.

Krug B, Dietlein M, Groth W, Stützer H, Psaras T,Gossmann A, et al. Fluor-18-fluorodeoxyglu-cose positron emission tomography (FDG-PET)in malignant melanoma - Diagnostic comparisonwith conventional imaging methods. Acta Radi-ologica 2000;41(5):446-52.

Kushwaha SS, Narula J, Narula N, Zervos G, Semi-gran MJ, Fischman AJ, Alpert NA, Dec GW,Gewirtz H. Pattern of changes over time inmyocardial blood flow and microvascular dilatorcapacity in patients with normally functioningcardiac allografts. Am J Cardiol 1998;82(11):1377-81.

Kuwert T, Delbeke D. Brain tumours. In: Wieler HJ,Coleman RE, eds. PET in clinical oncology.Damstadt: Steinkopff Verlag; 2000, p.125-38.

Landoni C, Lucignani G, Paolini G, Zuccari M, GalliL, Di Credico G, et al. Assessment of CABG-related risk in patients with CAD and LVD.Contribution of PET with [18F]FDG to the as-sessment of myocardial viability. J CardiovasSurg (Torino) 1999 Jun;40(3):363-72.

Lassen U, Daugaard G, Eigtved A, Damgaard K,Friberg L. 18F-FDG whole body positron emis-sion tomography (PET) in patients with un-known primary tumours (UPT). Eur J Cancer1999; 35(7):1076-82.

Law MP, Osman S, Pike VW, et al. Evaluation of [11C]GB67, a novel radioligand for imaging myocar-dial ∂1-adrenoceptors with positron emissiontomography. Eur J Nucl Med 2000;27:7-17.

Lee KS, Marwick TH, Cook SA, Go RT, Fix JS, JamesKB, et al. Prognosis of patients with left ven-tricular dysfunction, with and without viablemyocardium after myocardial infarction. Rela-tive efficacy of medical therapy and revasculari-zation. Circulation 1994 Dec;90(6):2687-94.

Lefroy DC, de Silva R, Choudhury L, et al. Diffuse re-duction of myocardial beta-adrenoceptors in hy-pertrophic cardiomyopathy: A study with posi-tron emission tomography. J Am Coll Cardiol1993;22:1653-60.

Leskinen-Kallio S, Ruotsalainen U, Nagren K, TerasM, Joensuu H. Uptake of carbon-11-methionineand fluorodeoxyglucose in non-Hodgkin's lym-phoma: a PET study. J Nucl Med 1991 Jun;32(6):1211-8.

Leung GK, Fan YW, Fong KY. Temporal lobe resec-tion for intractable epilepsy: review of 11 cases.Hong Kong Med J 1999;5(4):329-36.

Lips P, Comans EFI, Hoekstra OS, Clement EV, van-Mourik JC, Teule GJJ. Positron emission tomo-graphy for the detection of metastases of differ-entiated thyroid carcinoma. Neth J Med 2000;57(4):150-6.

Lonneux M, Lawson G, Ide C, Bausart R, Remacle M,Pauwels S. Positron emission tomography withfluorodeoxyglucose for suspected head and necktumor recurrence in the symptomatic patient.Laryngoscope 2000;110(9):1493-7.

Lowe VJ, Boyd JH, Dunphy FR, Kim H, Dunleavy T,Collins BT, et al. Surveillance for recurrent headand neck cancer using positron emission tomo-graphy. J Clin Oncol 2000;18(3):651-8.

Lowe VJ, Fletcher JW, Gobar L, et al. Prospective in-vestigation of positron emission tomography inlung nodules. J Clin Oncol 1998;16:1075-84.

Lucignani G, Paolini G, Landoni C, Zuccari M,Paganelli G, Galli L, et al. Presurgical identifi-cation of hibernating myocardium by combineduse of technetium-99m hexakis 2-methoxyisobutyli-sonitrile single photon emis-sion tomography and fluorine-18 fluoro-2-deoxy-D-glucose positron emission tomographyin patients with coronary artery disease. Eur JNucl Med 1992;19(10):874-81.

Maes A, Flameng W, Nuyts J, Borgers M, Shivalkar B,Ausma J, et al. Histological alterations inchronically hypoperfused myocardium. Correla-


References239

tion with PET findings. Circulation 1994 Aug;90(2):735-45.

Maes AF, Borgers M, Flameng W, Nuyts JL, van deWerf F, Ausma JJ, et al. Assessment of myocar-dial viability in chronic coronary artery diseaseusing technetium-99m sestamibi SPECT. Cor-relation with histologic and positron emissiontomographic studies and functional follow-up. JAm Coll Cardiol 1997 Jan;29(1):62-8.

Magnani P, Carretta A, Rizzo G, et al. FDG/PET andspiral CT image fusion for mediastinal lymphnode assessment of non-small cell lung cancerpatients. J Cardiovasc Surg (Torino) 1999;40(5):741-8.

Manolidis S, Donald PJ, Volk P, et al. The use ofpositron emission tomography scanning in oc-cult and recurrent head and neck cancer. ActaOtolaryngol Suppl (Stockh) 1998:5341.

Markand ON, Salanova V, Worth R, Park HM, Well-man HN. Comparative study of interictal PETand ictal SPECT in complex partial seizures.Acta Neurol Scand 1997;95(3):129-36.

Marom EM, McAdams HP, Erasmus JJ, Goodman PC,Culhane DK, Coleman RE, et al. Staging nonsmall cell lung cancer with whole-body PET.Radiology 1999;212(3):803-9.

Marwick TH, Shan K, Go RT, MacIntyre WJ, LauerMS. Use of positron emission tomography forprediction of perioperative and late cardiacevents before vascular surgery. Am Heart J 1995Dec;130(6):1196-202.

Marwick TH, Zuchowski C, Lauer MS, Secknus MA,Williams J, Lytle BW. Functional status andquality of life in patients with heart failure un-dergoing coronary bypass surgery after assess-ment of myocardial viability. J Am Coll Cardiol1999 Mar;33(3):750-8.

McFalls EO, Baldwin D, Kuskowski M, Liow J,Chesler E, Ward HB. Utility of positron emis-sion tomography in predicting improved leftventricular ejection fraction after coronary arterybypass grafting among patients with ischemic

cardiomyopathy. Cardiology 2000;93(1-2):105-12.

McLoud TC, Bourgouin PM, Greenberg RW, et al.Bronchogenic carcinoma: analysis of staging inthe mediastinum with CT by correlative lymphnode mapping and sampling. Radiology1992;182(2):319-23.

McNeil BJ. Prospects for outcomes research in radiol-ogy and health policy. Acad Radiol 1996Apr;3(Suppl 1):S19-20

McNeil BJ. Technology assessment. Acad Radiol.1995 Sep;2(Suppl 2):S100-2.

Meltzer CC, Luketich JD, Friedman D, Charron M,Strollo D, Meehan M, et al. Whole-body FDGpositron emission tomographic imaging forstaging esophageal cancer - Comparison withcomputed tomography. Clin Nucl Med2000;25(11):882-7.

Merlet P, Delforge J, Syrota A, et al. Positron emissiontomography with 11C CGP-12177 to assess beta-adrenergic receptor concentration in idiopathicdilated cardiomyopathy. Circulation 1993;87:1169-78.

Mijnhout GS, Hoekstra OS, van Tulder MW, TeuleGJ, Deville WL. Systematic review of the diag-nostic accuracy of (18)F-fluorodeoxyglucosepositron emission tomography in melanoma pa-tients. Cancer 2001;91(8):1530-42.

Moog F. Hodgkin and Non Hodgkin lymphoma. In:Wieler HJ, Coleman RE, eds. PET in clinicaloncology. Damstadt: Steinkopff Verlag, 2000,pp.255-72.

Moog F, Kotzerke J, Reske SN. FDG PET can replacebone scintigraphy in primary staging of malig-nant lymphoma. J Nucl Med. 1999Sep;40(9):1407-13.

Muzik O, da Silva EA, Juhasz C, Chugani DC, Shah J,Nagy F, et al. Intracranial EEG versus fluma-zenil and glucose PET in children with ex-tratemporal lobe epilepsy. Neurology 2000 Jan11; 54(1):171-9.


References240

Nestle U, Walter K, Schmidt S, Licht N, Nieder C,Motaref B, et al. 18F-deoxyglucose positronemission tomography (FDG-PET) for the plan-ning of radiotherapy in lung cancer: high impactin patients with atelectasis. Int J Radiat OncolBiol Phys 1999 Jun 1;44(3):593-7.

Noh DY, Yun IJ, Kang HS, Kim YC, Kim JS, ChungJK, Lee DS, Lee MC, Moon WK, Youn YK, OhSK, Choe KJ. Detection of cancer in augmentedbreasts by positron emission tomography. Eur JSurg 1999 Sep;165(9):847-51.

Nowak B, Di Martino E, Janicke S, et al. Diagnosticevaluation of malignant head and neck cancer byF-18-FDG PET compared to CT/MRI. Nuk-learmedizin 1999;38:312-8.

Nuutinen J, Sonninen P, Lehikoinen P, Sutinen E,Valavaara R, Eronen E, Norrgard S, Kulmala J,Teras M, Minn H. Radiotherapy treatment plan-ning and long-term follow-up with methioninePET in patients with low-grade astrocytoma. IntJ Radiat Oncol Biol Phys 2000 Aug 1;48(1):43-52.

Ogunbiyi OA, Flanagan FL, Dehdashti F, Siegel BA,Trask DD, Birnbaum EH, Fleshman JW, ReadTE, Philpott GW, Kodner IJ. Detection of recur-rent and metastatic colorectal cancer: compari-son of positron emission tomography and com-puted tomography. Ann Surg Oncol 1997;4:613-20.

Okada J, Oonishi H, Yoshikawa K, Itami J, Uno K,Imaseki K, Arimizu N. FDG-PET for predictingthe prognosis of malignant lymphoma. AnnNucl Med 1994 Aug; 8(3):187-91.

Pagano D, Townend JN, Littler WA, Horton R, CamiciPG, Bonser RS. Coronary artery bypass surgeryas treatment for ischemic heart failure: the pre-dictive value of viability assessment with quan-titative positron emission tomography forsymptomatic and functional outcome. J ThoracCardiovasc Surg 1998a Apr;115(4):791-9.

Pagano D, Bonser RS, Townend JN, Ordoubadi F,Lorenzoni R, Camici PG. Predictive value ofdobutamine echocardiography and positron

emission tomography in identifying hibernatingmyocardium in patients with postischaemicheart failure. Heart 1998b Mar;79(3):281-8.

Pagley PR, Beller GA, Watson DD, Gimple LW, Ra-gosta M. Improved outcome after coronary by-pass surgery in patients with ischemic cardio-myopathy and residual myocardial viability.Circulation 1997 Aug;96(3):793-800.

Paquet P, Henry F, Belhocine T, Hustinx R, Najjar F,Pierard-Franchimont C, et al. An appraisal of18-fluorodeoxyglucose positron emission tomo-graphy for melanoma staging. Dermatology2000;200(2):167-9.

Park KC, Schwimmer J, Shepherd JE, et al. Decisionanalysis for the cost-effective management ofrecurrent colorectal cancer. Ann Surg2001;233(3):310-9.

Pasquet A, Lauer MS, Williams MJ, Secknus MA,Lytle B, Marwick TH. Prediction of global leftventricular function after bypass surgery in pa-tients with severe left ventricular dysfunction.Impact of pre-operative myocardial function,perfusion, and metabolism. Eur Heart J 2000Jan;21(2):125-36.

Pasquet A, Robert A, D'Hondt AM, Dion R, Melin JA,Vanoverschelde JL. Prognostic value of myo-cardial ischemia and viability in patients withchronic left ventricular ischemic dysfunction.Circulation 1999 Jul;100(2):141-8.

Patterson RE, Eisner RL, Horowitz SF. Comparison ofcost-effectiveness and utility of exercise ECG,s ingle photon emiss ion computedtomography, positron emission tomography, andcoronary angiography for diagnosis of coronaryartery disease. Circulation 1995 Jan 1;91(1):54-65.

Patz EF Jr, Lowe VJ, Goodman PC, Herndon J. Tho-racic nodal staging with PET imaging with18FDG in patients with bronchogenic carci-noma. Chest 1995;108(6):1617-21.

Paulus P, Sambon A, Vivegnis D, et al. 18FDG PETfor the assessment of primary head and neck tu-


References241

mors: clinical, computed tomography and histo-pathological correlation in 38 patients.Laryngoscope 1998;108:1578-83.

Perie S, Talbot JN, Monceaux G, et al. Use of a coin-cidence gamma camera to detect primary tumorwith 18fluoro- 2-deoxy-glucose in cervicallymph node metastases from an unknown origin.Ann Otol Rhinol Laryngol 2000;109(8 Pt1):755-60.

Phelps ME, Hoffman EJ, Coleman RE, et al. Tomo-graphic images of blood pool and perfusion inbrain and heart. J Nucl Med 1976;17(7):603-12.

Phelps ME. PET: The merging of biology and imaginginto molecular imaging. J Nucl Med 2000;41(4):661-81.

Pieterman RM, van Putten JWG, Meuzelaar JJ, JohnWG, Meuzelaar JJ, et al. Preoperative staging ofnon small-cell lung cancer with positron-emission tomography. New Engl J Med2000;343(4):254-61.

Pirich C, Schwaiger M. The clinical role of positronemission tomography in management of the car-diac patient. Rev Port Cardiol 2000;19 Suppl1:I89-100.

Prauer HW, Weber WA, Romer W, Treumann T,Ziegler SI and Schwaiger M. Controlled pro-spective study of positron emission tomographyu s i n g t h e g l u c o s e a n a l o g u e[18f]fluorodeoxyglucose in the evaluation ofpulmonary nodules. Br J Surg 1998;85:1506-11.

Primack SL, Hartman TE, Lee KS, Muller NL. Pulmo-nary nodules and the CT halo sign. Radiology1994;190(2):513-5.

Primack SL, Lee KS, Logan PM, Miller RR, MullerNL. Bronchogenic carcinoma: utility of CT inthe evaluation of patients with suspected lesions.Radiology 1994;193(3):795-800.

Rahimtoola SH. The hibernating myocardium. AmHeart J 1989 Jan;117(1):211-21.

Ransohoff DF, Feinstein AR. Problems of spectrum

and bias in evaluating the efficacy of diagnostictests. N Engl J Med 1978;299(17):926-30.

Raylman RR, Majewski S, Wojcik R, WeisenbergerAG, Kross B, Popov V, Bishop HA. The poten-tial role of positron emission mammography fordetection of breast cancer. A phantom study.Med Phys 2000;27(8):1943-54.

Rees G, Bristow JD, Kremkau EL, Green GS, HerrRH, Griswold HE, Starr A. Influence of aorto-coronary bypass surgery on left ventricular per-formance. New Engl J Med 1971 May;284(20):1116-20.

Rege S, Maass A, Chaiken L, et al. Use of positronemission tomography with fluorodeoxyglucosein patients with extracranial head and neck can-cers. Cancer 1994;73:3047-3058.

Reiman EM, Caselli RJ, Chen K, Alexander GE,Bandy D, Frost J. Declining brain activity incognitively normal apolipoprotein E varepsilon4 heterozygotes: a foundation for using positronemission tomography to efficiently test treat-ments to prevent Alzheimer's disease. Proc NatlAcad Sci U S A. 2001 Mar 13;98(6):3334-9.

Rinne D, Baum RP, Hor G, Kaufmann R. Primarystaging and follow-up of high risk melanomapat i e n t s w i t h w h o l e - b o d y 1 8 F -fluorodeoxyglucose positron emission tomogra-phy: results of a prospective study of 100 pa-tients. Cancer 1998;82:1664-71.

Roberts PF, Follette DM, vonHaag D, Park JA, ValkPE, Pounds TR, Hopkins DM. Factors associ-ated with false-positive staging of lung cancerby positron emission tomography. Ann ThoracSurg 2000;70(4):1154-9.

Rodriguez M, Ahlstrom H, Sundin A, Rehn S, Sund-strom C, Hagberg H, Glimelius B. [18F] FDGPET in gastric non-Hodgkin's lymphoma. ActaOncol 1997;36(6):577-84.

Rodriguez M, Rehn S, Ahlstrom H, Sundstrom C,Glimelius B. Predicting malignancy grade withPET in non-Hodgkin's lymphoma. J Nucl Med.1995 Oct; 36(10):1790-6.


References242

Roelcke U, von Ammon K, Hausmann O, Kaech DL,Vanloffeld W, Landolt H, et al. Operated lowgrade astrocytomas: a long term PET study onthe effect of radiotherapy. J Neurol NeurosurgPsychiatry 1999 May; 66(5):644-7

Rossetti C, Landoni C, Lucignani G, Garraffa G,Ippolito M, Pepi M, et al. Lack of prognosticvalue of 99mTc MIBI rest, 18F FDG and 201TLrest/redistribution on PTCA outcome of patientswith moderate left ventricle dysfunction.Panminerva 1999 Sep;41(3):199-206.

Russell LB, Gold MR, Siegel JE, Daniels N, WeinsteinMC. The role of cost-effectiveness analysis inhealth and medicine. Panel on Cost-Effectiveness in Health and Medicine. JAMA1996;276(14):1172-7.

Ryvlin P, Bouvard S, Le Bars D, De Lamerie G,Gregoire MC et al. Clinical utility of flumazenil-PET versus [18F]fluorodeoxyglucose-PET andMRI in refractory partial epilepsy. A prospectivestudy in 100 patients. Brain 1998 Nov;121(Pt11):2067-81.

Sackett DL. Clinical epidemiology: a basic science forclinical medicine. Boston: Little, Brown, 1991.

Sato N, Suzuki M, Kuwata N, Kuroda K, Wada T,Beppu T, Sera K, Sasaki T, Ogawa A. Evalua-tion of the malignancy of glioma using 11C-methionine positron emission tomography andproliferating cell nuclear antigen staining. Neu-rosurg Rev 1999 Dec;22(4):210-4.

Saunders CAB, Dussek JE, O'Doherty MJ, MaiseyMN. Evaluation of fluorine-18-fluorodeoxyglu-cose whole body positron emission tomographyimaging in the staging of lung cancer. Ann Tho-rac Surg 1999;67(3):790-7.

Schafers M, Dutka D, Rhodes CG, et al. Myocardialpresynaptic and postsynaptic autonomic dys-function in hypertrophic cardiomyopathy. CircRes 1998a;82:57-62.

Schafers M, Wichter T, Lerch H, et al. Cardiac sym-pathetic innervation in patients with idiopathic

right ventricular outflow tract tachycardia. J AmColl Cardiol 1998b;32:181-6.

Scheidhauer K, Avril N, Bonkowsky VM. Carcinomaof unknown origin. In: Wieler HM et ColemanRE. PET in clinical oncology. Damstadt:Steinkopff Verlag, 2000, pp. 169-77.

Schelbert HR. The usefulness of positron emission to-mography. Curr Probl Cardiol 1998 Feb;23(2):69-120.

Schulte M, Brecht-Krauss D, Heymer B, Guhlmann A,Hartwig E, Sarkar MR, et al. Grading of tumorsand tumorlike lesions of bone: evaluation byFDG PET. J Nucl Med 2000;41(10):1695-701.

Schwaiger M, Kalff V, Rosenpire K, et al. Noninva-sive evaluation of the sympathetic nervous sys-tem in the human heart by PET. Circulation1990a;8:457-64.

Schwaiger M, Hutchins G, Rosenspire K, Haka M,Wieland DM. Quantitative evaluation of thesympathetic nervous system by PET in patientwith cardiomyopathy. J Nucl Med 1990b;31:792

Scott WJ, Shepherd J, Gambhir SS. Cost-effectivenessof FDG-PET for staging non small cell lungcancer: a decision analysis. Ann Thorac Surg1998;66(6):1876-83;discussion 1883-5.

Seltzer MA, Barbaric Z, Belldegrun A, Naitoh J,Dorey F, Phelps ME, et al. Comparison of heli-cal computerized tomography, positron emissiontomography and monoclonal antibody scans forevaluation of lymph node metastases in patientswith prostate specific antigen relapse aftertreatment for localized prostate cancer. J Urol1999;162(4):1322-8.

Siebelink HM, Blanksma PK, Crijns HJ, Bax JJ, vanBoven AJ, Kingma T, et al. No difference incardiac event-free survival between positronemission tomography-guided and single-photonemission computed tomography-guided patientmanagement: a prospective, randomized com-parison of patients with suspicion of jeopardizedmyocardium. J Am Coll Cardiol 2001 Jan;37(1):81-8.


References243

Silberstein EB. Prevalence of adverse reactions topositron emitting radiopharmaceuticals in nu-clear medicine. Pharmacopeia Committee of theSociety of Nuclear Medicine. J Nucl Med 1998Dec;39(12):2190-2.

Spaepen K, Stroobants S, Dupont P, Van SteenweghenS, Thomas J, Vandenberghe P, et al. Prognosticvalue of positron emission tomography (PET)with f luorine-18 f luorodeoxyglucose([18F]FDG) after first-line chemotherapy in non-Hodgkin's lymphoma: is FDG-PET a valid al-ternative to conventional diagnostic methods?. JClin Oncol. 2001 Jan 15;19(2):414-9.

Staib L, Schirrmeister H, Reske SN, Beger HG. Is F-18-fluorodeoxyglucose positron emission tomo-graphy in recurrent colorectal cancer a contribu-tion to surgical decision making? Amer J Surg2000;180(1):1-5.

STARD Group. Towards complete and accurate re-porting of studies on diagnostic accuracy: theSTARD statement. November 17, 2000 (unpub-lished).

Steinert HC, Huch BR, Buck A, et al. Malignant mela-noma: staging with whole-body positron emis-sion tomography and 2-[F-18]-fluoro-2-deoxy-D-glucose. Radiology 1995;195:705-9.

Stevens MJ, Dayanakli F, Raffel DM, et al. Scinti-graphic assessment of regionalized defects inmyocardial sympathic innervation and bloodflow regulation in diabetic patients with auto-matic neuropathy. J Am Coll Cardiol1998a;31:1575-84.

Stevens MJ, Raffel DM, Allman DM, et al. Cardiacsympathetic dysinnervation in diabetes: impli-cations for enhanced cardiovascular risk.Circulation 1998b;98:961-8.

Stevens MJ, Raffel DM, Allman KC, Schwaiger M,Wieland DM. Regression and progression ofcardiac sympathetic dysinnervation complicat-ing diabetes: an assessment by C-11 hy-droxyephedrine and positron emission tomogra-phy. Metabolism 1999;48:92-101.

Stokkel MPM, Hoekstra A, van Rijk PP. The detectionof small carcinoma with 18F-FDG using a dualhead coincidence camera. Eur J Radiology1999;32(3):160-2.

Strasberg SM, Dehdashti F, Siegel BA, Drebin JA,Linehan D. Survival of patients evaluated byFDG-PET before hepatic resection for metas-tatic colorectal carcinoma: a prospective data-base study. Ann Surg 2001;233(3):293-9.

Stumpe KD, Dazzi H, Schaffner A, von SchulthessGK. Infection imaging using whole-body FDG-PET. Eur J Nucl Med 2000;27(7):822-32.

Stumpe KD, Urbinelli M, Steinert HC, Glanzmann C,Buck A, von Schulthess GK. Whole-body posi-tron emission tomography using fluorodeoxy-glucose for staging of lymphoma: effectivenessand comparison with computed tomography. EurJ Nucl Med. 1998 Jul;25(7):721-8.

Swartz BE, Tomiyasu U, Delgado-Escueta AV, Man-delkern M, Khonsari A. Neuroimaging in tem-poral lobe epilepsy: test sensitivity and relation-ships to pathology and postoperative outcome.Epilepsia 1992;33(4):624-34.

Swets JA, Pickett, RM. Evaluation of diagnostic sys-tems: methods from signal detection theory.New York: Academic Press, 1982.

Tamaki N, Kawamoto M, Tadamura E, Magata Y, Yo-nekura Y, Nohara R, et al. Prediction of reversi-ble ischemia after revascularization. Perfusionand metabolic studies with positron emissiontomography. Circulation 1995 Mar;91(6):1697-705.

Tamaki N, Kawamoto M, Takahashi N, Yonekura Y,Magata Y, Nohara R, et al. Prognostic value ofan increase in fluorine-18 deoxyglucose uptakein patients with myocardial infarction: compari-son with stress thallium imaging. J Am CollCardiol 1993 Nov;22(6):1621-7.

Tamaki N, Ohtani H, Yamashita K, Magata Y, Yo-nekura Y, Nohara R, et al. Metabolic activity inthe areas of new fill-in after thallium-201 rein-jection: comparison with positron emission to-


References244

mography using fluorine-18-deoxyglucose. JNucl Med 1991 Apr;32(4):673-8.

Tamaki N, Yonekura Y, Yamashita K, Saji H, MagataY, Senda M, et al. Positron emission tomogra-phy using fluorine-18 deoxyglucose in evalua-tion of coronary artery bypass grafting. Am JCardiol 1989 Oct;64(14):860-5.

Tamaki N. Imaging of cardiac neuronal and receptorfunction. J Nucl Cardiol 1997;4:553-4.

Tashiro M, Kubota K, Itoh M, Moser E. Trends in useof positron emission tomography. Lancet 2001Mar 17;357(9259):886.

Tatlidil R, Luther S, West A, Jadvar H, Kingman T.Comparison of fluorine-18 deoxyglucose and O-15 water PET in temporal lobe epilepsy. ActaNeurol Belg 2000 Dec;100(4):214-20.

Thelen MH, Bares R. Malignant melanoma. In: WielerHJ, Coleman RE, eds. PET in clinical oncology.Damstadt: Steinkopff Verlag, 2000, pp. 299-308.

Thompson TP, Lunsford LD, Kondziolka D. Distin-guishing recurrent tumor and radiation necrosiswith positron emission tomography versusstereotactic biopsy. Stereotact Funct Neurosurg.1999; 73(1-4):9-14.

Tiepolt C, Beuthien-Baumann B, Hliscs R, Bredow J,Kühne A, Kropp J, et al. F-18-FDG for thestaging of patients with differentiated thyroidcancer: Comparison of a dual-head coincidencegamma camera with dedicated PET. Ann NuclMed 2000;14(5):339-45.

Tillisch J, Brunken R, Marshall R, Schwaiger M,Mandelkern M, Phelps M, Schelbert H. Reversi-bility of cardiac wall-motion abnormalities pre-dicted by positron tomography. New Engl J Med1986 Apr;314(14):884-8.

Tyler DS, Onaitis M, Kherani A, et al. Positron emis-sion tomography scanning in malignant mela-noma. Cancer 2000;89(5):1019-25.

Uberfuhr P, Frey AW, Ziegler S, Reichart B, Schwai-ger M. Sympathetic reinnervation of sinus nodeand left ventricle after heart transplantation inhumans: regional differences assessed by heartrate variability and positron emission tomogra-phy. J Heart Lung Transplant 2000aApr;19(4):317-23.

Uberfuhr P, Ziegler S, Schwaiblmair M, Reichart B,Schwaiger M. Incomplete sympathetic reinner-vation of the orthotopically transplanted humanheart: observation up to 13 years after hearttransplantation. Eur J Cardiothorac Surg 2000b;17(2):161-8.

Valk PE, Abella-Columna E, Haseman MK, PoundsTR, Tesar RD, Myers RW, et al. Whole-bodyPET imaging with 18F-fluorodeoxyglucose inmanagement of recurrent colorectal cancer. ArchSurg 1999;134(5):503-11;discussion 511-3.

Valk PE, Pounds TR, Hopkins DM, et al. Staging nonsmall cell lung cancer by whole-body positronemission tomographic imaging. Ann ThoracSurg 1995;60(6):1573-81;discussion 1581-2.

Valk PE, Pounds TR, Hopkins DM, Haseman MK.Cost-effectiveness of PET imaging in clinicaloncology. Nucl Med and Biol 1996;23:737-43.

van den Berg EK, Popma JJ, Dehmer GJ, Snow FR,Lewis SA, Vetrovec GW, Nixon JV. Reversiblesegmental left ventricular dysfunction aftercoronary angioplasty. Circulation 1990Apr;81(4):1210-6.

Vanoverschelde JL, Wijns W, Borgers M, HeyndrickxG, Depre C, Flameng W, Melin JA. Chronicmyocardial hibernation in humans. From bed-side to bench. Circulation 1997 Apr;95(7):1961-71.

Vanoverschelde JL, Wijns W, Depre C, Essamri B,Heyndrickx GR, Borgers M, et al. Mechanismsof chronic regional postischemic dysfunction inhumans. New insights from the study of nonin-farcted collateral-dependent myocardium.Circulation 1993 May;87(5):1513-23.

Vansteenkiste J, Stroobants S, De Leyn P,


References245

Verschakelen J, Nackaerts K, Deneffe G, et al.Mediastinal lymph node staging with FDG-PETscan in patients with potentially operable nonsmall cell lung cancer: a prospective analysis of50 cases. Leuven Lung Cancer Group. Chest1997 Dec; 112(6):1480-6.

Vansteenkiste JF, Stroobants SG, Dupont PJ, et al.FDG-PET scan in potentially operable non-small cell lung cancer: do anatometabolic PET-CT fusion images improve the localisation ofregional lymph node metastases? The LeuvenLung Cancer Group. Eur J Nucl Med1998;25(11):1495-501.

van Tinteren, H, Hoekstra OS, Smit EF, van den BerghJH, Schreurs AJ, Stallaert RA, et al. Random-ized controlled trial (RCT) to evaluate the cost-effectiveness of positron emission tomography(PET) added to conventional diagnosis in non-small cell lung cancer (NSCLC). Abstract 1885.2000 Program/Proceedings Abstracts. AmericanSociety of Clinical Oncology. URL: http:www.asco.org/prf/me/html/00abstracts/m_toc.htm

Vanuytsel LJ, Vansteenkiste JF, Stroobants SG, DeLeyn PR, De Wever W, Verbeken EK, et al. Theimpact of (18)F-fluoro-2-deoxy-D-glucosepositron emission tomography (FDG-PET)lymph node staging on the radiation treatmentvolumes in patients with non small cell lungcancer. Radiother Oncol 2000;55(3):317-24.

Vesselle H, Schmidt RA, Pugsley JM, et al. Lung can-cer proliferation correlates with [F-18]fluorodeoxyglucose uptake by positron emis-s ion tomography. Cl in Can Res2000;6(10):3837-44.

vom Dahl J, Altehoefer C, Sheehan FH, Buechin P,Schulz G, Schwarz ER, et al. Effect of myocar-dial viability assessed by technetium-99m-sestamibi SPECT and fluorine-18-FDG PET onclinical outcome in coronary artery disease. JNucl Med 1997 May;38(5):742-8.

vom Dahl J, Altehoefer C, Sheehan FH, Buechin P,Uebis R, Messmer BJ, et al. Recovery of re-gional left ventricular dysfunction after coronary

revascularization. Impact of myocardial viabilityassessed by nuclear imaging and vessel patencyat follow-up angiography. J Am Coll Cardiol1996 Oct;28(4):948-58.

vom Dahl J, Eitzman DT, al-Aouar ZR, Kanter HL,Hicks RJ, Deeb GM, et al. Relation of regionalfunction, perfusion, and metabolism in patientswith advanced coronary artery disease under-going surgical revascularization. Circulation1994 Nov;90(5):2356-66.

von Schulthess GK, Steinert HC, Dummer R, WederW. Cost-effectiveness of whole-body PET im-aging in non small cell lung cancer and malig-nant melanoma. Acad Radiol 1998;5 Suppl2:S300-2.

Wagner JD, Schauwecker D, Davidson D, Coleman JJ3rd, Saxman S, Hutchins G, et al. Prospectivestudy of fluorodeoxyglucose positron emissiontomography imaging of lymph node basins inmelanoma patients undergoing sentinel node bi-opsy. J Surg Oncol 1999;17:1508-15.

Walter SD, Irwig L, Glasziou PP. Meta-analysis of di-agnostic tests with imperfect reference stan-dards. J Clin Epidemiol 1999 Oct;52(10):943-51.

Weber WA, Neverve J, Sklarek J, et al. Imaging oflung cancer with fluorine-18 fluorodeoxyglu-cose: comparison of a dual-head gamma camerain coincidence mode with a full-ring positronemission tomography system. Eur J Nucl Med1999;26(4):388-95.

White PG, Adams H, Crane MD, Butchart EG. Preop-erative staging of carcinoma of the bronchus:can computed tomographic scanning reliablyidentify stage III tumours? Thorax 1994Oct;49(10):941-3.

Wieland D, Rosenpire K, Hutchins G, et al. Neuronalmapping of the hear t wi th 6-[18F]fluorometaraminol. J Med Chem1990;33:956-64.

Wiggers H, Nielsen TT, Bottcher M, Egeblad H, Bot-ker HE. Positron emission tomography and low-


References246

dose dobutamine echocardiography in the pre-diction of postrevascularization improvement inleft ventricular function and exercise parameters.Am Heart J 2000 Dec;140(6):928-36.

Wijns W, Vatner SF, Camici PG. Hibernating myocar-dium. N Engl J Med 1998;339:173-81.

Williams DE, Pairolero PC, Davis CS, et al. Survivalof patients surgically treated for stage I lungcancer. J Thorac Cardiovasc Surg1981;82(1):70-6.

Willkomm P, Bender H, Bangard M, Decker P,Grünwald F, Biersack HJ. FDG PET and immu-noscintigraphy with Tc-99m-labeled antibodyfragments for detection of the recurrence of co-lorecta l ca rc inoma. J Nuc l Med2000;41(10):1657-63.

Wolpers HG, Burchert W, van den Hoff J, WeinhardtR, Meyer GJ, Lichtlen PR. Assessment of myo-cardial viability by use of 11C-acetate and posi-tron emission tomography. Threshold criteria ofreversible dysfunction. Circulation 1997 Mar18;95(6):1417-24.

Wong WL, Chebretton EB, McGurk M, et al. A pro-spective study of PET-FDG imaging for the as-sessment of head and neck squamous cell carci-noma. Clin Otolaryngol 1997;22:209-14.

Wong WL, Hussain K, Chebretton E, et al. Validationand clinical application of computer combinedcomputed tomography and positron emissiontomography with 2-[18F]-fluoro-2-deoxy-D-glucose head and neck images. Am J Surg1996;172:628-32.

Wyllie E. Surgical treatment of epilepsy in children.Pediatr Neurol 1998;19(3):179-88.

Yokoyama I, Momomura S, Ohtake T, et al. Improve-ment of impaired myocardial vasodilatation dueto diffuse coronary atherosclerosis in hypercho-lesterolemics after lipid-lowering therapy. Cir-culation 1999;100:117-122.

Yoshida K, Gould KL. Quantitative relation of myo-cardial infarct size and myocardial viability by

positron emission tomography to left ventricularejection fraction and 3-year mortality with andwithout revascularization. J Am Coll Cardiol1993 Oct;22(4):984-97.

Young TH, Tang KP, Tang HS, et al. Primary malig-nant lymphoma of the spine diagnosed by Tc-99m MDP and Ga-67 imaging. Clin Nucl Med1997;22(4):274-6.

Yusuf S, Zucker D, Peduzzi P, Fisher LD, Takaro T,Kennedy JW, et al. Effect of coronary artery by-pass graft surgery on survival: overview of 10-year results from randomised trials by the Coro-nary Artery Bypass Graft Surgery Trialists Col-laboration. Lancet 1994 Aug 27;344(8922):563-70.

Zhang X, Liu X, Shi R, Wu Q, Gao R, Liu Y, et al.Evaluation of the clinical value of combinationof 99mTc-MIBI myocardial SPECT and 18F-FDG PET in assessing myocardial viability. Ra-diat Med 1999 May-Jun;17(3):205-10.

Zhuang H, Sinha P, Pourdehnad M, Duarte PS, Ya-mamoto AJ, Alavi A. The role of positron emis-sion tomography with fluorine-18-deoxyglucosein identifying colorectal cancer metastases toliver. Nucl Med Commun 2000;21(9):793-8.

Zinzani PL, Magagnoli M, Chierichetti F, ZompatoriM, Garraffa G, Bendandi M, et al. The role ofpositron emission tomography (PET) in themanagement of lymphoma patients. Ann Oncol.1999 Oct; 10(10):1181-4.

Zubiate P, Kay JH, Mendez AM. Myocardial revascu-larization for the patient with drastic impairmentof function of the left ventricle. J Thorac Car-diovasc Surg 1977 Jan;73(1):84-6.

Documents

POSITRON EMISSION TOMOGRAPHY IN QUÉBEC...PET, to the exclusion of its research applica-tions. POSITRON EMISSION TOMOGRAPHY Introduced as a research tool in the mid-70s, PET differs