Translating Genomics in Cancer Care - CADTH.ca

Yvonne Bombard, Peter B. Bach and Kenneth Offit

Translating Genomics in Cancer Care

Harborside Press, 37 Main Street, Cold Spring Harbor, NY 11724 is published by JNCCN – The Journal of the National Comprehensive Cancer Network

Print ISSN: 1540-1405. Online ISSN: 1540-1413.

. All rights reserved. Copyright © 2013 by the National Comprehensive Cancer Network

2013;11:1343-1353J Natl Compr Canc Netw

Online article http://www.jnccn.org/content/11/11/1343.full

Subscriptions

http://www.jnccn.org/site/subscriptions/ is online at Comprehensive Cancer NetworkJNCCN – The Journal of the NationalInformation about subscribing to

Permissionshttp://www.NCCN.org/permissionsmaterial, please go online to

For information about photocopying, republishing, reprinting, or adapting

NCCN.org

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from

. A

ll rig

hts

rese

rved

.C

opyr

ight

© 2

013

by th

e N

atio

nal C

ompr

ehen

sive

Can

cer

Net

wor

kfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from

http://www.jnccn.org/content/11/11/1343.full

http://www.jnccn.org/content/11/11/1343.full

http://www.jnccn.org/site/subscriptions/

http://www.jnccn.org/site/subscriptions/

http://www.NCCN.org/permissions

http://www.jnccn.org/




Original Article

© JNCCN—Journal of the National Comprehensive Cancer Network | Volume 11 Number 11 | November 2013

1343

AbstractThere is increasing enthusiasm for genomics and its promise in advancing personalized medicine. Genomic information has been used to personalize health care for decades, spanning the fields of cardiovascular disease, infectious disease, endocrinology, meta-bolic medicine, and hematology. However, oncology has often been the first test bed for the clinical translation of genomics for diagnostic, prognostic, and therapeutic applications. Notable he-reditary cancer examples include testing for mutations in BRCA1 or BRCA2 in unaffected women to identify those at significantly el-evated risk for developing breast and ovarian cancers, and screen-ing patients with newly diagnosed colorectal cancer for mutations in 4 mismatch repair genes to reduce morbidity and mortality in their relatives. Somatic genomic testing is also increasingly used in oncology, with gene expression profiling of breast tumors and EGFR testing to predict treatment response representing com-monly used examples. Health technology assessment provides a rigorous means to inform clinical and policy decision-making through systematic assessment of the evidentiary base, along with precepts of clinical effectiveness, cost-effectiveness, and consid-eration of risks and benefits for health care delivery and society. Although this evaluation is a fundamental step in the translation of any new therapeutic, procedure, or diagnostic test into clinical care, emerging developments may threaten this standard. These include “direct to consumer” genomic risk assessment services and the challenges posed by incidental results generated from next-

From the aCenter for Health Policy and Outcomes, Department of Epidemiology and Biostatistics, and bClinical Genetics Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York; and cSchool of Public Health, Department of Epidemiology and Public Health, Yale University, New Haven, Connecticut. Submitted March 13, 2013; accepted for publication May 17, 2013.Dr. Bombard has disclosed that she was supported by a fellowship from the Canadian Institutes of Health Research. Drs. Bombard and Offit have disclosed that they have no financial interests, arrangements, affiliations, or commercial interests with the manufacturers of any products discussed in this article or their competitors. Dr. Bach has disclosed that he is a speaker for Genentech, Inc.Correspondence: Yvonne Bombard, PhD, University of Toronto, Institute of Health Policy, Management, and Evaluation, The Keenan Research Centre in the Li Ka Shing Knowledge Institute, St. Michael’s Hospital, 30 Bond Street, Toronto, Ontario M5B 1T8, Canada. E-mail: [email protected]


Yvonne Bombard, PhDa,b,c; Peter B. Bach, MD, MAPPa; and Kenneth Offit, MD, MPHb

generation sequencing (NGS) technologies. This article presents a review of the evidentiary standards and knowledge base sup-porting the translation of key cancer genomic technologies along the continuum of validity, utility, cost-effectiveness, health service impacts, and ethical and societal issues, and offers future research considerations to guide the responsible introduction of NGS tech-nologies into health care. It concludes that significant evidentiary gaps remain in translating genomic technologies into routine clini-cal practice, particularly in efficacy, health outcomes, cost-effec-tiveness, and health services research. These caveats are especially germane in the context of NGS, wherein efforts are underway to translate NGS results despite their limited accuracy, lack of proven efficacy, and significant computational and counseling challenges. Further research across these domains is critical to inform the ef-fective, efficient, and equitable translation of genomics into can-cer care. (JNCCN 2013;11:1343–1353)

Genomics in Cancer CareThere is increasing enthusiasm for genomics and its promise in advancing personalized medicine.1–3 How-ever, the application of genomics in health care is not a novel concept. Genetic testing and risk assessment based on family history has been an established part of standard medical care for decades. Genomics in health care builds on the principles established by the integra-tion of genetics into medical practice.4 What is novel, however, is the scale of genomic analysis with regard to the amount of information considered. Although “classic” genetics evaluates genes in isolation, genom-ics assesses multiple genes or individuals’ entire genetic sequences and the interactions between genes and epi-genetic interactions with the environment, to improve the prevention, diagnosis, and treatment of disease.

Many examples exist of the use of genomic in-formation to personalize health care; these span the fields of cardiovascular disease, infectious disease, en-docrinology, metabolic medicine, and hematology.5 However, oncology has often been the first test bed for

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article

Bombard et al


1344

the clinical translation of genomics for diagnostic, prognostic, and therapeutic applications (Table 1).1 Notable hereditary cancer examples include testing for mutations in BRCA1 or BRCA2 in unaffected women to identify those at significantly elevated risk for developing breast and ovarian cancer,6 and screening patients with newly diagnosed colorec-tal cancer for mutations in mismatch repair genes (including MLH1, MSH2, EPCAM, MSH6, and PMS2) to reduce morbidity and mortality in their relatives.7 These examples characterize high-pene-trant germline mutations that are associated with large increases in cancer risk (relative risks of 9–33 based on various ages6), for which detection facili-tates risk-reducing surgeries and/or use of surveil-lance technologies.

Germline genomic testing is distinct from so-matic genomic testing of tumors to predict progno-sis and treatment response. Somatic genomic test-ing is increasingly used in oncology.7 For example, gene expression profiling (GEP) of breast tumors

examines expression levels of a panel of prognos-tically relevant genes to establish the likelihood of benefit from chemotherapy and recurrence risk within 10 years among node-negative, estrogen receptor–positive patients.8,9 A second example is testing for somatic mutations in the EGFR gene to predict response to tyrosine kinase inhibitors, such as gefitinib in first-line therapy or erlotinib in the second- or third-line setting, in patients with ad-vanced non–small cell lung cancer (NSCLC).10,11 These applications represent commonly used ge-nomic technologies in cancer care.

Genomic Translation and Health Technology AssessmentThe translation of genomic technologies to health care is typically established through the develop-ment of clinical practice and policy guidance. Health technology assessment (HTA) provides a rigorous means to inform clinical and policy deci-

Table 1 Examples of Genomic Technologies Used Across the Continuum of Cancer CareClinical Stage Target Indication

Risk assessment BRCA1, BRCA2 Risk-reducing surgery for breast and ovarian cancers

MLH1, MSH2 (including EPCAM), MSH6, PMS2

Intensified colonoscopic screening for Lynch syndrome (hereditary nonpolyposis colorectal cancer)

CDH1 Preventive surgery for hereditary gastric cancer

Screening HPV genotypes Screening for cervical cancer

Diagnosis BCR-ABL, E2A-PBX1, TEL-AML1, MLL fusions and rearrangements

Individualizing Rx for leukemia

Gene expression profiles Individualizing Rx for subtypes of breast cancer and lymphomas

Prognosis Gene expression profiles Individualizing Rx for breast, colon, and prostate cancers and lymphomas

HER2/neu, ER, PR Individualizing Rx for breast cancer

Treatment EGFR mutations Targeting Rx with gefitinib, erlotinib treatment for lung cancer and glioblastoma

KIT, PDGFR mutations Targeting Rx for imatinib, nilotinib, sunitinib, sorafenib treatment for sarcomas gliomas, liver cancer, renal cancer, and melanoma

BRAF mutations Targeting Rx with RAF inhibitors in melanoma

BCR-ABL translocation Targeting Rx with imatinib for chronic myelogenous leukemia

KRAS mutations Precluding need for EGFR inhibition in colorectal cancer

BRCA mutation Targeting Rx with PARP inhibitors for breast, ovarian, prostate, and pancreatic cancers

HER2 Targeting Rx with Herceptin (trastuzumab) for HER2+ breast cancer

Abbreviations: ER, estrogen receptor; HPV, human papillomavirus; PARP, poly (ADP-ribose) polymerase; PR, progesterone receptor; Rx, treatment. Adapted from refs.4,5,7

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article



1345

sion-making through systematic assessment of the supporting evidentiary base, including consideration of clinical effectiveness (eg, validity, utility, safety), cost-effectiveness (eg, economic evaluation), and risks and benefits for health care delivery and society (eg, impact on health services, consistency with soci-etal and ethical values) (Table 2).12 The Evaluation of Genomic Applications in Practice and Prevention (EGAPP) was established as an advisory body to the Office of Public Health Genomics at the Centers for Disease Control and Prevention to provide evidence reviews for genomic technologies. This indepen-dent group adopted review methods similar to HTA frameworks,13 although HTA frameworks typically include broader considerations of health service delivery, economic analysis, and ethical or social is-sues. Although evaluation of the evidentiary base of a technology is a fundamental step in the transla-tion of any new therapeutic, procedure, or diagnostic test into clinical care, emerging developments may threaten this standard.

Emerging Issues in Genomics Translation: Cancer as HarbingerGenome-wide association studies have defined hun-dreds of single nucleotide polymorphisms (SNPs) as low-penetrance genetic markers of risk for cancer and common disease. Unlike high- and intermedi-ate-penetrance mutations, SNPs associated with dis-ease risk are common in the population but, in isola-tion, confer small increases in risk, which vary based on environmental and lifestyle factors.14 More than 100 SNPs are currently associated with cancer risk, but these tests have uncertain clinical validity, and are therefore not currently considered part of stan-dard oncology or preventive care.14 However, com-mercial laboratories currently offer these tests as part of their genomic risk assessment services, termed di-rect to consumer (DTC) testing.

A second emerging development is massively parallel sequencing technologies of exomes or whole genomes, collectively termed next genera-tion sequencing (NGS). To date, NGS has largely

Table 2 Health Technology Assessment Dimensions, Aims, and Methods in GenomicsHTA Precept Aims Methods

Validity

Analytic validity Accuracy and reliability of the test in detecting the genetic changes of interest

Analytic sensitivity and specificity, assay reproducibility, robustness, and laboratory proficiency testing

Clinical validity Accuracy and reliability of the test in identifying patients with the disorder of interest

Specificity, sensitivity, genotype/phenotype relationships, and predictive values

Utility

Clinical utility Likelihood that use of the test will significantly improve health-related outcomes

Morbidity, mortality, clinical outcomes, and quality-of-life data

Personal utility Benefits of information beyond clinical utility, such as richer self-knowledge, relieving uncertainty, enabling reproductive and life planning, or empowering risk-reducing behaviors

Metric does not exist though (actual or hypothetical) participation rates and satisfaction may be proxies

Economic evaluation Value for money: health gain (benefit), relative to resources consumed for a particular health technology (costs) compared with available alternatives

Cost/benefit analysis, cost-effectiveness analysis, and cost-utility analysis

Health services impact Feasibility of adopting a genomic technology or implementing a genomic screening program into the health system

Operational and economic feasibility, net budget impact

Ethical and social values Consistency of adopting a genomic technology with broadly shared moral and societal values

Primary research, secondary analyses, expert advice/report, and public engagement

Adapted from Johnson AP, Sikich NJ, Evans G, et al. Health technology assessment: a comprehensive framework for evidence-based recommendations in Ontario. Int J Technol Assess Health Care 2009;25:141–150; and Teutsch SM, Bradley LA, Palomaki GE, et al. The Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Initiative: methods of the EGAPP Working Group. Genet Med 2009;11:3–14.

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article

Bombard et al


1346

been directed toward the discovery of the genetic cause of specific disorders.15 NGS is also increas-ingly applied clinically to panels of genes for the si-multaneous analysis of multiple susceptibility genes (multiplex testing)16 and to identify mutations in tumors to guide therapy. In the process, compari-sons are typically made between the tumor genome and the germline genetic sequence to distinguish causative mutations.17 However, in the course of se-quencing, NGS reveals incidental germline infor-mation, including genetic variants associated with unrecognized current disease, drug response, future risk for common diseases, and variants of unknown significance, on the scale of thousands of variants per individual18 (although few would be considered medically actionable). Consequently, oncologists will likely be the first line of health providers to be confronted by dozens of incidental inherited susceptibilities for which their patients with cancer did not necessarily request testing. In a research set-ting, these incidental findings may be anticipated in discussions of informed consent. However, whether in a research or clinical context, a major challenge has emerged regarding whether incidental NGS results should be actively sought or analyzed, and whether and how these should be disclosed to in-dividuals whose genomes or exomes have been sequenced.19–25 Policy guidance has increasingly taken the position that investigators have a duty to disclose actionable incidental results.18 In fact, recent recommendations from the American Col-lege of Medical Genetics have delineated a mini-mum set of clinically actionable incidental results that laboratories are required to seek and report in patients they are clinically sequencing.26 These de-velopments will create significant ethical and clini-cal challenges for providers and institutions.19,27,28

Collectively, these emerging issues threaten to undermine the evidence-based translation of ge-nomics in health care because a potential surge of genomic information will be disseminated, which will not have established efficacy based on eviden-tiary standards required before translation. Further-more, translation of genomic information of limited clinical validity or utility threatens to create undue pressure on the health care system if DTC users or receivers of NGS incidental results return to their health care providers to request follow-up tests and procedures that may be unnecessary or even harm-

ful. Commentators have posited that this incidental genomic information may lead to a cascade effect, resulting in inappropriate follow-up that would con-stitute an unjustified use of health care resources.27,29

Thus, because of the advent of tumor sequencing to target therapies and resulting generation of inci-dental findings, cancer care will continue to chart the course for the translation of genomics in health care. This article presents a comprehensive review of the evidentiary standards and knowledge base supporting the translation of key cancer genomic technologies along the continuum of validity, util-ity, cost-effectiveness, health service impacts, and ethical and societal issues (Table 3), and offers future research considerations for the responsible introduc-tion of NGS technologies into health care.

Translating Genomics in Cancer Care: Evaluating the Evidence

Analytical ValidityAnalytic validity of a genomic test focuses on the accuracy and reliability of the test in detecting the genetic changes of interest (genotype or analyte).13 Measures typically include analytic sensitivity and specificity, assay reproducibility, robustness, and lab-oratory proficiency testing.Examples: Tumor-Based Screening for Lynch Syndrome: EGAPP found that the overall analytic validity of the preliminary and diagnostic tests for Lynch syndrome is high based on the analytic performance of microsat-ellite instability (MSI) testing and DNA sequencing, respectively.30 However, the analytic validity of the diagnostic sequencing and the Multiplex Ligation-Dependent Probe Amplification (MRC-Holland; Amsterdam, The Netherlands) was not assessed. EGFR Testing to Predict Response to Treatment for NSCLC: A review by Blue Cross Blue Shield Associa-tion (BCBSA) found sequencing to have sufficient evi-dence to support the predictive value of EGFR mutation testing for gefitinib therapy based on sensitivity, speci-ficity, positive predictive value, and negative predictive value of mutation presence compared with its absence.31 Future Research and Considerations: Although DNA (Sanger) sequencing remains the gold standard for diagnostic testing, external proficiency testing pro-grams for sequencing do not exist. Furthermore, there is often a lack of a gold standard in genomic technologies, which remains a significant limitation in evaluating an-

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article



1347

alytic validity. Heterogeneity both within and between tumors is a significant confounder in somatic genomic testing, including for patients undergoing EGFR test-ing.32 Importantly, neither SNP testing nor NGS have been fully validated. In the case of SNP risk assess-ment, different risk estimates have been obtained from the same individual’s sample, largely because of varia-tions in analytic algorithms and differences in SNPs analyzed.33 In the case of NGS, exome sequencing plat-forms vary in capture efficiency and may fail to detect structural genomic changes or distinguish mutations on the same or different chromosomes.34 No proficiency testing currently exists in the United States for NGS.33

Clinical ValidityClinical validity of a genomic test aims to ensure the ac-curacy and reliability of the test in identifying patients

with the disorder or outcome of interest.13 Key metrics include clinical specificity, clinical sensitivity, genotype/phenotype relationships, and predictive values.Examples: Tumor-Based Screening for Lynch Syndrome: EGAPP determined that adequate evidence of clinical validity exists for the preliminary and diagnostic tests to identify Lynch syndrome, wherein sensitivity of prelimi-nary immunohistochemical (IHC) testing was 83% and specificity was 89%. The sensitivity and specificity of preliminary MSI testing varied according to mismatch repair mutations present, and ranged from 55% to 81% and 89% to 90%, respectively.30 However, EGAPP did not find enough evidence to recommend one prelimi-nary screening strategy over another (eg, MSI vs IHC).GEP for Recurrence Risk in Early-Stage Breast Cancer: Clinical validity for GEP relies on the ability of the

Table 3 Selected Evidence for Genetic and Genomic Technologies, Across Health Technology Assessment Domain

ExamplesAnalytic Validity

Clinical Validity

Clinical Utility

Economic Evaluation

Health Service Delivery

Ethical and Social Issues

Hereditary Cancer Genomics

Tumor-based screening for Lynch syndrome

High (EGAPP)

Adequate (EGAPP)

Adequatea (EGAPP)

Cost-effective (ICER $25,000 per life year) (EGAPP)b

$41–$281 million (EGAPP)c

Informed consent, psychological distress, uptake (EGAPP)d

BRCA testing for breast and ovarian cancer susceptibility

None found

None found

Faire (USPSTF)

Cost-effectivef None found

Mixed psychosocial response, low risk of harms (USPSTF)

Somatic Cancer Genomics

EGFR testing to predict response to treatment for NSCLC

Sufficient (BCBS)

No conclusions reached (BCBS)

No conclusions reached (BCBS)

Cost-effective (ICER of $46,021 per life year) (OHTAC)g

$4.6–$8.1 million (OHTAC)h

Improved QoL (OHTAC)

GEP for recurrence risk in breast cancer

Inadequate evidence (EGAPP)

Adequatei (EGAPP)

Inadequate evidence (EGAPP)j

Inconclusive evidence (EGAPP)

$4.14 million (OHTAC)k

Inadequate evidence (EGAPP)

Abbreviations: BCBS, Blue Cross Blue Shield Association; EGAPP, Evaluation of Genomic Applications in Practice and Prevention; EGFR, epidermal growth factor receptor; GEP, gene expression profiling; ICER, incremental cost-effectiveness ratio; NSCLC, non–small cell lung cancer; OHTAC, Ontario Health Technology Advisory Committee; QALY, quality-adjusted life year; QoL, quality of life; USPSTF, US Preventative Services Task Force. aEvidence relates to reduced morbidity and mortality. bBased on a study commissioned by EGAPP.55

cTotal programmatic costs, depending on testing strategy adopted.60

dEGAPP review did not conduct a review but commented on these issues in its considerations of the contextual issues. eEvidence relates to prophylactic surgeries in reducing breast and ovarian cancer risk. fBased on one analysis on preventative surgeries for women with BRCA1/2, taking a societal perspective.70

gOr ICER of $81,071 per QALY. hBudget impact assuming yearly incidence of advanced NSCLC in Ontario from 2011 to 2015 would be the same as that in 2010, and EGFR gene mutation testing was performed once for any patients newly diagnosed with advanced NSCLC from 2011 to 2015.56

iEvidence relates to MammaPrint and Oncotype DX. jEvidence relates to Oncotype DX (no evidence found for MammaPrint and H:1 Ratio Tests). kBudget impact assuming an update rate of 25%.71

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article

Bombard et al


1348

tests to predict overall or recurrence-free survival 5 to 10 years after surgery versus avoidance of chemothera-py toxicity. Several GEP tests are available, which use real-time PCR or microarray analysis. EGAPP found adequate quality evidence to support the association between recurrence scores produced by Oncotype DX

and MammaPrint and rates of 5- or 10-year metasta-sis.35 EGAPP also found adequate evidence to support the association between recurrence scores produced by Oncotype DX and chemotherapy benefit.35

Future Research and Considerations: Given the relatively low prevalence of genetic risk factors in the population, highly specific screening tests are re-quired to avoid excessive false-positive results.36 This is a particularly challenging issue with NGS because of the high rate of false-positive and -negative re-sults,37 and indeterminate results from computational limitations in assessing novel variants for functional consequences.33 Family history, clinical information, and alignment of the sequence data with reference databases are commonly used to filter germline vari-ants. Although sequencing remains the gold stan-dard for minimizing false-positive results, its capacity is dwarfed by NGS, creating a need for alternative high-throughput genotyping approaches to validate NGS findings. Validating somatic data is more chal-lenging because of the genetic heterogeneity and pu-rity of the tumors.33 Comparison with germline tissue from the same patient can filter variants, but this can also result in the identification of incidental, inher-ited predispositions. Minimizing false-negatives is more complex, and as yet cannot be accomplished given the coverage and quality issues inherent in NGS of exomes or whole genomes. In contrast, NGS technologies applied to panels of genes (multiplex testing) currently have better depth of coverage and accuracy than NGS applied to exomes and whole genomes.16

Clinical UtilityClinical utility focuses on safety, effectiveness, bur-den of illness, and need.12 Typical assessments in-clude morbidity, mortality, clinical outcomes, and quality-of-life data. In the case of genomics, clinical utility is based on the likelihood that use of the test will significantly improve health-related outcomes.13 Examples: BRCA Testing for Breast and Ovarian Cancer Susceptibility: Despite the fair quality of sup-porting evidence, the US Preventive Services Task Force (USPSTF) concluded that BRCA testing has

clinical utility.38 This decision was based on evidence indicating that prophylactic bilateral mastectomy in high-risk women has resulted in an 85% to 100% reduction in risk for breast cancer.38 Similarly, oo-phorectomy was shown to reduce ovarian cancer risk by 85% to 100% and breast cancer risk by 53% to 68%.38 GEP for Recurrence Risk in Early-Stage Breast Cancer: EGAPP found no direct evidence to support GEP’s clinical utility.35 However, in assessing indirect, ret-rospective evidence for Oncotype DX, they noted a benefit associated with chemotherapy and a 27% re-duction in 10-year recurrence rates for women with a high recurrence score who had node-negative, receptor-positive disease.35 Future Research and Considerations: Supportive evidence for clinical utility in the hereditary disease context is generally limited by low incidence rates and, by extension, small sample sizes. Although can-cer is highly prevalent, assessment of genomic tech-nology often requires cases to be stratified by molecu-lar subtypes, also limiting sample size. Thus, efficacy even for common tumor types may be difficult to demonstrate. For new genotyping technologies, these research gaps will necessarily lead to delays in dem-onstrating clinical utility. Nonetheless, additional research is needed to assess how individuals use the results of genomic tests in medical decision-making and lifestyle choices, and their associated clinical outcomes (whether testing is provided clinically or DTC, or results are generated incidentally from NGS). Further research on how genomic testing could be incorporated into behavioral interventions (eg, diet, exercise, smoking cessation, cancer screen-ing) support would also be important for determining the clinical value of these technologies.39

Personal UtilityAlthough not a traditional precept of HTA, propo-nents of genomic technologies have advanced the concept of personal utility, arguing that clinical util-ity is too restrictive a metric to use in judging the possible benefits of genomic information. Personal utility is believed to offer richer self-knowledge and to empower risk-reducing behaviors, and has gained prominence as a justification for disclosing genomic information in the absence of an evidentiary base or proof of a direct influence on health outcomes.40 Examples: Although considered an emerging jus-tification by some, the notion of personal utility

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article



1349

is an established concept in genetics. Provision of genetic tests for untreatable, fatal diseases (eg, Li-Fraumeni syndrome, Huntington disease) and diseases with complex environmental interactions (eg, hereditary melanoma) was predicated on the fact that, despite medical intervention, there is personal utility in knowing one’s risk for a future illness to empower risk reducing behaviors, re-lieve uncertainty, and inform reproductive, career, or financial decisions.41–44 Although a metric for personal utility does not exist, emerging literature suggests the presence of perceived benefits from genetic testing for fatal and complex diseases,42–44 and uptake of risk-reducing strategies among indi-viduals with a genetic predisposition.44,45 Future Research and Considerations: Broader considerations are required regarding whether per-sonal utility is a valid justification for translating genomic technologies, and what weight should be given to this purported benefit in relation to other HTA precepts. This is especially important given the increasing public interest in learning genom-ic results of uncertain or low utility46–48 through commercial DTC companies, or through research efforts designed to return incidental research find-ings to participants. The desire to improve health has been cited as a common motivator for pur-suing DTC testing or participating in genome sequencing research,49–52 although research sup-porting these outcomes has been limited. Given the increasing popularity of DTC testing for low to moderate penetrance alleles and the magni-tude of incidental findings from NGS technolo-gies, justifications based on personal utility ought to be weighed against the potential harms of false alarm, false reassurance, and adverse psychologi-cal sequelae,4 in addition to the unjustified use of health care resources.29 Longitudinal, comparative effectiveness research on the benefits, harms, and health care use after the receipt of DTC or inci-dental results from NGS is needed, including com-parative observational and randomized trials; as-sessment of patient-reported outcomes and health service use; decision modeling; and economic analyses.53 These studies would build the evidence base supporting the translation of genomic tech-nologies, and could inform the fundamental ques-tion of whether these technologies result in addi-tional or inappropriate use of health care services.

Economic EvaluationEconomic evaluations consider value for money. Analyses typically measure the health gain (ben-efit) relative to resources consumed for a particular health technology (costs) compared with avail-able alternatives.12 Various approaches to economic evaluations exist, and these differ in the way they measure benefits (but not costs). Cost-effectiveness analysis (CAE) and cost-utility analysis (CUA) are commonly used in HTA. CAE measures health ef-fects, such as life years saved or improvements in functional status (eg, deaths prevented), using an in-cremental cost-effectiveness ratio (ICER) (eg, incre-mental costs per death prevented).54 CUA is similar to CAE, but CUA measures a technology’s health effects using a health index, such as quality-adjusted life years (QALYs). The incremental ratio used in CUA assesses incremental costs per QALY gained.54

Examples: Tumor-Based Screening for Lynch Syn-drome: A CAE commissioned by EGAPP found that tumor-based screening for Lynch syndrome is cost-effective and that an IHC-first approach is su-perior, with an ICER of less than $25,000 per life year saved relative to no testing.55 EGFR Testing to Predict Response to Treatment for NSCLC: The economic analysis conducted by the Ontario Health Technology Advisory Committee (OHTAC) concluded that EGFR testing to guide the use of gefitinib as first-line therapy for patients with advanced NSCLC is cost-effective, with an ICER of $81,071 per QALY or $46,021 per life year.56 Furthermore, they concluded that EGFR testing is also cost-effective to guide use of erlotinib after the failure of docetaxel or pemetrexed in pa-tients with known EGFR gene mutation status.56

Future Research and Considerations: One of the main benefits of genomic technologies is the value of information provided, yet economic analyses do not typically take a value-based perspective, espe-cially when commissioned by a payer for reimburse-ment purposes.36 Economic evaluations also typically base payer decisions on the health system perspec-tive, rather than that of society or the patient.36 The patient’s perspective may be particularly relevant in the case of hereditary disorders, in which the need for some intervention may be high but its avail-ability may be limited.57 An overriding assumption about genomic technologies that stratify treatment response is that they will provide cost savings. For

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article

Bombard et al


1350

example, GEP for breast cancer recurrence risk as-sessment is estimated to save $400 million a year in unnecessary adjuvant chemotherapy.58 However, ac-quired drug resistance to molecularly targeted cancer therapies and the poor specificity of NGS challenge these assumptive cost savings.59 How these costs sav-ings or expenditures should be taken into account in economic analyses is unclear, but critical in assessing the value for money of genomic technologies.

Health Service ImpactsConsiderations of health service impacts focus on the feasibility of adopting a genomic technology into the health system. Analyses focus on organizational feasibil-ity, such as barriers to uptake, capital, human resources, and regulatory aspects.12 A net budget impact is also conducted to assess economic feasibility of adoption. Examples: Tumor-Based Screening for Lynch Syndrome: Using a cost-consequence analysis, EGAPP estimated total programmatic costs to range from $41 to $281 million, depending on testing strategy adopted.60 EGFR Testing to Predict Response to Treatment for NSCLC: The budget analysis conducted by the OHTAC reported program costs of $4.6 to $8.1 million from 2010 to 2015.56

Future Research and Considerations: Service delivery represents a significant gap in translating genomics in health care, including oncology. In the case of Lynch screening, for example, clinical resource needs, clinical expertise gaps, and educational deficits have been iden-tified as potential barriers to implementation.61 More research is needed on the optimal models for provid-ing genetic services, provider training, and tools to assist providers, such as facilities for outreach and telemedi-cine, just-in-time management plans, and emergency management tools.62 Despite the declining costs of NGS technologies, the costs of integrating NGS tech-nologies in clinical care will be substantial, given the need to invest in large-scale resources and build capac-ity to analyze, validate, and interpret variants. Arguably, these costs may be dwarfed by the additional costs from follow-up tests and procedures caused by incidental findings and false-positive results from NGS technolo-gies over time.

Ethical and Social ValuesConsiderations of ethical and social values in HTA typically focus on the consistency of adopting a genom-ic technology with broadly shared moral and societal values.12 There are 3 general approaches to addressing

ethical and social values in HTA: conducting primary research; conducting secondary analyses of published literature on the perceptions, acceptability, quality of life, attitudes, or values of stakeholders; and commis-sioning expert review.63,64 A relatively new approach to addressing ethical and social issues in HTA is pub-lic engagement, which is increasingly being used—and even mandated by some jurisdictions (eg, National In-stitute for Health Care and Excellence [NICE] in the United Kingdom65)—in an effort to incorporate citizen values or patient perspectives into HTA and ensuing guidance.64

Examples: BRCA Testing for Breast and Ovar-ian Cancer Susceptibility: The report by the USP-STF considered the impact of genetic counsel-ing on psychosocial response to BRCA testing (eg, anxiety, depression, cancer risk perception, uptake) and found mixed evidence. The report also reviewed evidence of harms, such as over-diagnosis and overtreatment, and concluded that the magnitude of potential harms is small.38

EGFR Testing To Predict Response To Treatment for NSCLC: OHTAC concluded that there was moderate quality evidence of improved qual-ity of life in patients with an EGFR muta-tion who are being treated with gefitinib rela-tive to those treated with chemotherapy.56

Future Research and Considerations: Few HTA reports address ethical issues in their analyses.66 This may be because of the abundance of methods that exist and the limited capacity within HTA agencies to address them.63 However, substantial literature exists on the ethical and societal issues concerning genomic technologies, which generally focus on informed consent, genetic discrimination, family communication, and other psychosocial is-sues. Future research would benefit from informed considerations of the broader risks and benefits of translating genomic technologies using novel methodologies that trade-offs benefits relative to risks, such as discrete choice experiments or delib-erative, consensus-building exercises. Particularly important ethical and social issues brought about with NGS and molecularly driven treatment are access to technologies and alternative treatment,67 overdiagnosis,68 and the auxiliary health care costs of follow-up tests and procedures caused by inci-dental findings from NGS technologies.29

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article



1351

ConclusionsCancer care represents the leading edge of the im-pact of genomics on personalized medicine. Notable examples have demonstrated the efficacy of genomic markers for targeting and personalizing cancer pre-vention, detection, and treatment. Because of the advent of tumor sequencing for targeting therapies, oncologists will be among the frontline of those seek-ing to responsibly communicate incidental germline findings to patients and their families. Most genomic research focuses on gene discovery and the clinical validity or utility of emerging tests.69 However, sig-nificant challenges remain in translating genomic technologies and NGS into personalized preventive and medical care. Most importantly, significant gaps exist in the evidence base in genomics, including efficacy, health outcomes, cost-effectiveness, and health services research. The efficacy of genomic-based interventions may be difficult to demonstrate, especially for stratified technologies and in the he-reditary disease context, in which sample size is often an inherent limitation. Combined, these evidentiary gaps make costs and benefits difficult to estimate.

These caveats are especially germane to NGS, wherein efforts are underway to translate NGS re-sults from exomes or whole genomes, particularly in oncology, despite their unproven diagnostic accu-racy, requirement for validation, and computation-al and counseling challenges. Targeted funding is needed for centralized databases collating the patho-genicity of NGS sequence variants, as is support for outcome-based research using genomic markers and clinical and behavioral end points. Finally, contin-ued support for systematic evidence-based reviews of the emerging clinical genomics literature is critical to inform the effective, efficient, and equitable trans-lation of genomics into cancer care.

References1. Pasche B, Absher D. Whole-genome sequencing: a step closer to

personalized medicine. JAMA 2011;305:1596–1597.

2. Hamburg MA, Collins FS. The path to personalized medicine. N Engl J Med 2010;363:301–304.

3. Burke W, Psaty BM. Personalized medicine in the era of genomics. JAMA 2007;298:1682–1684.

4. Offit K. Personalized medicine: new genomics, old lessons. Hum Genet 2011;130:3–14.

5. Chan IS, Ginsburg GS. Personalized medicine: progress and promise. Annu Rev Genomics Hum Genet 2011;12:217–244.

6. Robson M, Offit K. Clinical practice. Management of an inherited predisposition to breast cancer. N Engl J Med 2007;357:154–162.

7. Weitzel JN, Blazer KR, Macdonald DJ, et al. Genetics, genomics, and cancer risk assessment: state of the art and future directions in the era of personalized medicine. CA Cancer J Clin, in press.

8. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med 2004;351:2817–2826.

9. 9. Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol 2006;24:3726–3734.

10. Kris MG, Natale RB, Herbst RS, et al. Efficacy of gefitinib, an inhibitor of the epidermal growth factor receptor tyrosine kinase, in symptomatic patients with non-small cell lung cancer: a randomized trial. JAMA 2003;290:2149–2158.

11. Perez-Soler R, Chachoua A, Hammond LA, et al. Determinants of tumor response and survival with erlotinib in patients with non–small-cell lung cancer. J Clin Oncol 2004;22:3238–3247.

12. Johnson AP, Sikich NJ, Evans G, et al. Health technology assessment: a comprehensive framework for evidence-based recommendations in Ontario. Int J Technol Assess Health Care 2009;25:141–150.

13. Teutsch SM, Bradley LA, Palomaki GE, et al. The Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Initiative: methods of the EGAPP Working Group. Genet Med 2009;11:3–14.

14. Stadler ZK, Thom P, Robson ME, et al. Genome-wide association studies of cancer. J Clin Oncol 2010;28:4255–4267.

15. Bamshad MJ, Ng SB, Bigham AW, et al. Exome sequencing as a tool for Mendelian disease gene discovery. Nat Rev Genet 2011;12:745–755.

16. Domchek SM, Bradbury A, Garber JE, et al. Multiplex genetic testing for cancer susceptibility: out on the high wire without a net? J Clin Oncol 2013;31:1267–1270.

17. Bombard Y, Robson ME, Offit K. Revealing the incidentalome when targeting the tumor genome. JAMA 2013;310:795–796.

18. Cassa CA, Savage SK, Taylor PL, et al. Disclosing pathogenic genetic variants to research participants: quantifying an emerging ethical responsibility. Genome Res 2012;22:421–428.

19. Bombard Y, Offit K, Robson ME. Risks to relatives in genomic research: a duty to warn? Am J Bioeth 2012;12:12–14.

20. Miller FA, Christensen R, Giacomini M, Robert JS. Duty to disclose what? Querying the putative obligation to return research results to participants. J Med Ethics 2008;34:210–213.

21. Bredenoord AL, Kroes HY, Cuppen E, et al. Disclosure of individual genetic data to research participants: the debate reconsidered. Trends Genet 2011;27:41–47.

22. Knoppers BM, Joly Y, Simard J, Durocher F. The emergence of an ethical duty to disclose genetic research results: international perspectives. Eur J Hum Genet 2006;14:1170–1178.

23. Berg JS, Khoury MJ, Evans JP. Deploying whole genome sequencing in clinical practice and public health: meeting the challenge one bin at a time. Genet Med 2011;13:499–504.

24. Biesecker LG. Incidental variants are critical for genomics. Am J Hum Genet 2013;92:648–651.

25. Goddard KA, Whitlock EP, Berg JS, et al. Description and pilot results from a novel method for evaluating return of incidental findings from next-generation sequencing technologies. Genet Med 2013;15:721–728.

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article

Bombard et al


1352

26. Green RC, Berg JS, Grody WW, et al. ACMG recommendations for reporting of incidental findings in clinical exome and genome sequencing. Genet Med 2013;15:565–574.

27. Kohane IS, Masys DR, Altman RB. The incidentalome: a threat to genomic medicine. JAMA 2006;296:212–215.

28. Biesecker LG, Burke W, Kohane I, et al. Next-generation sequencing in the clinic: are we ready? Nat Rev Genet 2012;13:818–824.

29. McGuire AL, Burke W. An unwelcome side effect of direct-to-consumer personal genome testing: raiding the medical commons. JAMA 2008;300:2669–2671.

30. Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group. Recommendations from the EGAPP Working Group: genetic testing strategies in newly diagnosed individuals with colorectal cancer aimed at reducing morbidity and mortality from Lynch syndrome in relatives. Genet Med 2009;11:35–41.

31. Blue Cross Blue Shield Association. Epidermal growth factor receptor mutations and tyrosine kinase inhibitor therapy in advanced non-small-cell lung cancer. Assessment Program 2011;25:1–29.

32. Nakano H, Soda H, Takasu M, et al. Heterogeneity of epidermal growth factor receptor mutations within a mixed adenocarcinoma lung nodule. Lung Cancer 2008;60:136–140.

33. Schrijver I, Aziz N, Farkas DH, et al. Opportunities and challenges associated with clinical diagnostic genome sequencing: a report of the association for molecular pathology. J Mol Diagn 2012;14:525–540.

34. Ormond KE, Wheeler MT, Hudgins L, et al. Challenges in the clinical application of whole-genome sequencing. Lancet 2011;375:1749–1751.

35. Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group. Recommendations from the EGAPP Working Group: can tumor gene expression profiling improve outcomes in patients with breast cancer? Genet Med 2009;11:66–73.

36. Conti R, Veenstra DL, Armstrong K, et al. Personalized medicine and genomics: challenges and opportunities in assessing effectiveness, cost-effectiveness, and future research priorities. Med Decis Making 2010;30:328–340.

37. Lam HY, Clark MJ, Chen R, et al. Performance comparison of whole-genome sequencing platforms. Nat Biotechnol 2011;30:78–82.

38. U.S. Preventive Services Task Force (USPSTF) genetic risk assessment and BRCA mutation testing for breast and ovarian cancer susceptibility: recommendation statement. Ann Intern Med 2005;143:355–361.

39. McBride CM, Bowen D, Brody LC, et al. Future health applications of genomics: priorities for communication, behavioral, and social sciences research. Am J Prev Med 2011;38:556–565.

40. Grosse SD, Khoury MJ. What is the clinical utility of genetic testing? Genet Med 2006;8:448–450.

41. Lammens CR, Aaronson NK, Wagner A, et al. Genetic testing in Li-Fraumeni syndrome: uptake and psychosocial consequences. J Clin Oncol 2010;28:3008–3014.

42. Hayden MR, Bombard Y. Psychosocial effects of predictive testing for Huntington’s disease. Adv Neurol 2005;96:226–239.

43. Wiggins S, Whyte P, Huggins M, et al. The psychological consequences of predictive testing for Huntington’s disease.

Canadian Collaborative Study of Predictive Testing. N Engl J Med 1992;327:1401–1405.

44. Lammens CR, Bleiker EM, Aaronson NK, et al. Regular surveillance for Li-Fraumeni syndrome: advice, adherence and perceived benefits. Fam Cancer 2010;9:647–654.

45. Aspinwall LG, Leaf SL, Dola ER, et al. CDKN2A/p16 genetic test reporting improves early detection intentions and practices in high-risk melanoma families. Cancer Epidemiol Biomarkers Prev 2008;17:1510–1519.

46. Partridge AH, Winer EP. Informing clinical trial participants about study results. JAMA 2002;288:363–365.

47. Shalowitz DI, Miller FG. Communicating the results of clinical research to participants: attitudes, practices, and future directions. PLoS Medicine 2008;5:e91.

48. Murphy J, Scott J, Kaufman D, et al. Public expectations for return of results from large-cohort genetic research. Am J Bioeth 2008;8:36–43.

49. Bloss CS, Ornowski L, Silver E, et al. Consumer perceptions of direct-to-consumer personalized genomic risk assessments. Genet Med 2010;12:556–566.

50. Bloss CS, Wineinger NE, Darst BF, et al. Impact of direct-to-consumer genomic testing at long term follow-up. J Med Genet 2013;50:393–400.

51. Facio FM, Brooks S, Loewenstein J, et al. Motivators for participation in a whole-genome sequencing study: implications for translational genomics research. Eur J Hum Genet 2011;19:1213–1217.

52. Gollust SE, Gordon ES, Zayac C, et al. Motivations and perceptions of early adopters of personalized genomics: perspectives from research participants. Public Health Genomics 2012;15:22–30.

53. Goddard KA, Knaus WA, Whitlock E, et al. Building the evidence base for decision making in cancer genomic medicine using comparative effectiveness research. Genet Med 2012;14:633–642.

54. Palmer S, Byford S, Raftery J. Economics notes: types of economic evaluation. BMJ 1999;318:1349.

55. Mvundura M, Grosse SD, Hampel H, Palomaki GE. The cost-effectiveness of genetic testing strategies for Lynch syndrome among newly diagnosed patients with colorectal cancer. Genet Med 2010;12:93–104.

56. Health Quality Ontario. Epidermal growth factor receptor mutation (EGFR) testing for prediction of response to EGFR-targeting tyrosine kinase inhibitor (TKI) drugs in patients with advanced non-small-cell lung cancer: an evidence-based analysis. Ont Health Technol Assess Ser 2010;10:1–48.

57. Culyer AJ, Bombard Y. An equity framework for health technology assessments. Med Decis Making 2012;32:428–441.

58. Lyman GH, Cosler LE, Kuderer NM, Hornberger J. Impact of a 21-gene RT-PCR assay on treatment decisions in early-stage breast cancer: an economic analysis based on prognostic and predictive validation studies. Cancer 2007;109:1011–1018.

59. Yauch RL, Settleman J. Recent advances in pathway-targeted cancer drug therapies emerging from cancer genome analysis. Curr Opin Genet Dev 2012;22:45–49.

60. Palomaki GE, McClain MR, Melillo S, et al. EGAPP supplementary evidence review: DNA testing strategies aimed at reducing morbidity and mortality from Lynch syndrome. Genet Med 2009;11:42–65.

61. Hall MJ. Counterpoint: implementing population genetic screening for Lynch syndrome among newly diagnosed colorectal

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Original Article



1353

cancer patients—will the ends justify the means? J Natl Compr Canc Netw 2010;8:606–611.

62. Kaye CI. Genetic service delivery: infrastructure, assessment and information. Public Health Genomics 2012;15:164–171.

63. Lehoux P, Williams-Jones B. Mapping the integration of social and ethical issues in health technology assessment. Int J Technol Assess Health Care 2007;23:9–16.

64. Bombard Y, Abelson J, Simeonov D, Gauvin FP. Eliciting social values and ethics in health technology assessment: a participatory approach. Soc Sci Med 2011;73:135–144.

65. Davies C, Wetherell M, Barnett E, Seymour-Smith S. Opening the Box: Evaluating the Citizens Council of NICE. London: School of Health & Social Welfare and Psychology Discipline, The Open University; 2005.

66. DeJean D, Giacomini M, Schwartz L, Miller FA. Ethics in Canadian health technology assessment: a descriptive review. Int J Technol Assess Health Care 2009;25:463–469.

67. Bombard Y, Abelson J, Simeonov D, Gauvin FP. Citizens’ perspectives on personalized medicine: a qualitative public deliberation study. Eur J Hum Genet 2013;21:1197–1201.

68. Moynihan R, Doust J, Henry D. Preventing overdiagnosis: how to stop harming the healthy. BMJ 2012;344:e3502.

69. Khoury MJ, Gwinn M, Yoon PW, et al. The continuum of translation research in genomic medicine: how can we accelerate the appropriate integration of human genome discoveries into health care and disease prevention? Genet Med 2007;9:665–674.

70. Anderson K, Jacobson JS, Heitjan DF, et al. Cost-effectiveness of preventive strategies for women with a BRCA1 or a BRCA2 mutation. Ann Intern Med 2006;144:397–406.

71. Health Quality Ontario. Gene expression profiling for guiding adjuvant chemotherapy decisions in women with early breast cancer: an evidence-based and economic analysis. Ont Health Technol Assess Ser 2010;10:1–57.

. All

right

s re

serv

ed.

Cop

yrig

ht ©

201

3 by

the

Nat

iona

l Com

preh

ensi

ve C

ance

r N

etw

orkfrom

00.

000.

000.

0 o

n N

ovem

ber

19, 2

013

by g

uest

jn

ccn.

org

Dow

nloa

ded

from



Documents

Translating Genomics in Cancer Care - CADTH.ca