Embed Size (px)
Citation preview
T
r
c
fl
n
m
b
s
a
r
e
m
(
l
m
a
r
i
p
n
Ttanrcaemaimt
lnoldgal
ictbhotft
S
PET and PET-CT for Evaluation of Colorectal Carcinoma
Dominique Delbeke and William H. Martin
i
h
f
m
r
p
b
c
d
a
l
p
m
i
f
i
a
c
©
he evaluation of patients with known or suspected
ecurrent colorectal carcinoma is now an accepted indi-
ation for positron emission tomography using 18F-
uorodeoxyglucose (FDG-PET) imaging. FDG-PET does
ot replace imaging modalities such as computed to-
ography (CT) for preoperative anatomic evaluation
ut is indicated as the initial test for diagnosis and
taging of recurrence and for preoperative staging (N
nd M) of known recurrence that is considered to be
esectable. FDG-PET imaging is valuable for the differ-
ntiation of posttreatment changes from recurrent tu-
or, differentiation of benign from malignant lesions
indeterminate lymph nodes, hepatic and pulmonary
esions), and the evaluation of patients with rising tu-
or markers in the absence of a known source. The
ddition of FDG-PET to the evaluation of these patients
educes overall treatment costs by accurately identify-
ng patients who will and will not benefit from surgical
rocedures. Although initial staging at the time of diag-
osis is often performed during colectomy, FDG-PET
he evaluation of patients with extracranial neoplasms,
mi
Ftlocbpisrectoa
sti
T
PmvN
eminars in Nuclear Medicine, Vol XXXIV, No 3 (July), 2004: pp 209-
maging is recommended for a subgroup of patients at
igh risk (with elevated CEA levels) and normal CT and
or whom surgery can be avoided if FDG-PET shows
etastases. Screening for recurrence in patients at high
isk has also been advocated. FDG-PET imaging seems
romising for monitoring patient response to therapy
ut larger studies are necessary. The diagnostic impli-
ations of integrated PET-CT imaging include improved
etection of lesions on both the CT and FDG-PET im-
ges, better differentiation of physiologic from patho-
ogic foci of metabolism, and better localization of the
athologic foci. This new powerful technology provides
ore accurate interpretation of both CT and FDG-PET
mages and therefore more optimal patient care. PET-CT
usion images affect the clinical management by guid-
ng further procedures (biopsy, surgery, radiation ther-
py), excluding the need for additional procedures, and
hanging both inter- and intramodality therapy.
2004 Elsevier Inc. All rights reserved.
HE RAPID ADVANCES in imaging technologiesare a challenge for physicians who must integrate
hese technologies for optimal patient care and outcomest minimal cost. Since the early 1990s, numerous tech-ological improvements have occurred in the field ofadiological imaging. These include 1) multislice spiralomputed tomography (CT), which permits the fastcquisition of CT angiographic images and multiphasenhancement techniques, and 2) positron emission to-ography (PET) using 18F-fluorodeoxyglucose (FDG)
s a radiopharmaceutical that provides the capability formaging glucose metabolism. Multiple indications forolecular imaging using FDG are now well accepted in
he fields of neurology, cardiology, and oncology.1
The goals of oncologic imaging are lesion detection,esion characterization, evaluation of the extent of theeoplasm, staging for malignant lesions, and assessmentf the therapeutic response. Staging includes lesionocalization, evaluation of proximity to vessels, andetection of nodal and distant metastases. Some of theseoals are better achieved with the high resolution ofnatomical imaging techniques and others with molecu-ar imaging using PET.
Molecular imaging using positron imaging is uniquen that positron emitters allow labeling of radiopharma-euticals that closely mimic endogenous molecules, andhere are continuing efforts to development of newiological tracers. FDG because of its relatively longalf life and its ability to assess cellular glucose metab-lism is the radiopharmaceutical most widely used withhe PET technology; it has been approved by the Centeror Medical Services for reimbursement by Medicare in
yocardial viability and in the presurgical assessment ofntractable epilepsy.
A wide variety of malignant tumors avidly accumulateDG. This is the result of increased numbers of glucose
ransporter proteins and increased intracellular enzymeevels of hexokinase and phosphofructokinase, amongthers, which promote glycolysis.2-5 FDG-PET imagingan be used to exploit the metabolic differences betweenenign and malignant cells for imaging pur-oses.6,7 The widespread oncologic applications includ-ng differentiation of benign from malignant lesions,taging malignant lesions, detection of malignant recur-ence, and monitoring therapy have contributed to thestablishment of the PET technology in many medicalenters in the United States, Europe, and progressivelyhroughout the world. Improvements in the distributionf FDG by commercial companies have now made FDGvailable to many medical centers as well.
Although numerous studies have shown that theensitivity and specificity of FDG imaging is superior tohat of CT in many clinical settings, the inability of FDGmaging to provide anatomical localization remains a
From the Vanderbilt University Medical Center, Nashville,N.Address reprint requests to Dominique Delbeke, MD, PhD,
rofessor and Director of Nuclear Medicine and PET, Depart-ent of Radiology and Radiological Sciences, Vanderbilt Uni-
ersity Medical Center, 21st Avenue South and Garland,ashville, TN 37232-2675.© 2004 Elsevier Inc. All rights reserved.0001-2998/04/3403-0005$30.00/0
doi:10.1053/j.semnuclmed.2004.03.006209223
sBdepaidvtoptme
kadrlp
CfpCgiicratdsPocta
ctephtbpaufs
N
tapoiudttulitiaau
erchtptto
icgtipaFpt(unFvlp
clmsie
210 DELBEKE AND MARTIN
ignificant impairment in maximizing its clinical value.ecause FDG is a tracer of glucose metabolism, itsistribution is not limited to malignant tissue. To avoidrrors, the interpreter must be familiar with the normalattern and physiologic variations of FDG distributionnd with clinical data relevant to the patients.8,9 It is alsomportant to standardize the environment of the patienturing the uptake period so as to limit physiologicariations of FDG uptake, (eg, in activated muscularissue). The problem of precise anatomical localizationf the foci of abnormal uptake and differentiation ofhysiologic from pathologic uptake is compounded byhe lower resolution and increased noise in the images ofany of the systems at the low end of the spectrum and
specially the hybrid gamma camera-based systems.Limitations of anatomical imaging with CT are well-
nown and are related to 1) size criteria for differenti-ting benign from malignant lymph nodes, 2) difficultyifferentiating posttherapy changes from tumor recur-ence, and 3) difficulty differentiating nonopacifiedoops of bowel from metastases in the abdomen andelvis.Close correlation of FDG studies with conventional
T scans helps to minimize these difficulties. In practiceor the past ten years, interpretation has been accom-lished by visually comparing corresponding FDG andT images. The interpreting physician visually inte-rates the two image sets to precisely locate a region ofncreased uptake on the CT scan. To aid in imagenterpretation, computer software has been developed tooregister the FDG-PET emission scans with the high-esolution anatomical maps provided by CT.10 Anotherpproach that has gained wider acceptance recently ishe hardware approach to image fusion using multimo-ality imaging with an integrated PET-CT imagingystem.11 The recent technical development of integratedET-CT systems provides CT and FDG-PET imagesbtained in a single imaging setting allowing optimaloregistration of images. The fusion images provided byhese systems allow accurate interpretation of both CTnd FDG-PET studies.
These advances in imaging technologies bring anotherhallenge to physicians at times when it is also importanto provide care at an acceptable cost. Increasing cost-ffectiveness and decreasing the number of invasiverocedures are currently two of the major trends inealth care. Pursuant to these goals, considerable atten-ion has recently been directed toward the use of meta-olic imaging using FDG-PET in the evaluation ofatients with cancer. Metabolic imaging, used in theppropriate setting, allows significant reduction in thetilization of more costly and invasive surgical methodsor diagnosing and staging disease in patients with
uspicious lesions. uormal Distribution of FDG
To interpret FDG images, one must be familiar withhe normal distribution of FDG, physiological variations,nd benign conditions that accumulate FDG.8,9,12 Somehysiological variations are important for interpretationf FDG in colorectal carcinoma. Uptake in the gastro-ntestinal tract is variable from patient to patient andptake along the esophagus is common, especially in theistal portion and at the gastroesophageal junction and inhe presence of esophagitis; the esophagus is best iden-ified on sagittal views. The wall of the stomach issually faintly seen and can be used as an anatomicalandmark, but occasionally the uptake can be relativelyntense. There is uptake in the cecum of many patientshat may be related to abundant lymphoid tissue in thentestinal wall, among other factors. When markedctivity is present in the bowel, evaluation for recurrencet the anastomotic site can be difficult. Mild-to-moderateptake is also usually seen at colostomy sites.Unlike glucose, FDG is filtered by the glomerulus and
xcreted into the urine. The accumulation of FDG in theenal collecting system may mask FDG uptake in adja-ent organs. Therefore, the patient should be kept wellydrated to promote diuresis. For optimal evaluation ofhe pelvis, the bladder should be empty. Therefore,atients are usually asked to void before acquisition ofhe images and images are acquired from the pelvis tohe cranium. The administration of furosemide canccasionally be useful to avoid focal ureteral activity.In the resting state, there is low accumulation of FDG
n the muscular system, but following exercise signifi-ant accumulation of FDG occurs in selected muscularroups, and may mislead the interpreter. Hyperventila-ion may induce uptake in the diaphragm and stress-nduced muscle tension is often seen in the trapezius andaraspinal muscles. Muscle relaxants such as benzodi-zepines (diazepam, 5-10 mg orally, 30-60 min beforeDG administration) may be helpful in these tenseatients. The PET-CT technology allowed characteriza-ion of FDG uptake in metabolically active fatty tissuebrown fat) that was previously believed to be muscleptake.13 In patients with lung tumors and laryngealerve palsy, PET-CT images helped to localize unilateralDG uptake at the base of the neck in the contralateralocal cord,14 allowing discrimination between physio-ogical laryngeal uptake from metastasis or a secondrimary neoplasm.Inflammation in general can result in FDG uptake that
an be severe enough to be confused with malignantesions, especially when there is granulomatous inflam-ation, including tuberculosis, sarcoidosis, histoplasmo-
is and aspergillosis among others.15 This is particularlymportant when evaluating patients posttreatment; forxample, sites of surgical intervention demonstrate FDG
ptake in the early healing phase due to inflammatory
ccwad
pfttFmsiupcbf
cptac5UActbscpc
pruotmsbssccKFpfio
oCwca
tdfTp1(tns
oppa
ctfiotprmtappthdecvmw
CS
or8s
211PET AND PET-CT FOR COLORECTAL CARCINOMA
hanges. Inflammatory changes after radiation therapyan make interpretation of FDG uptake challenging asell, although comparison with baseline FDG images
nd knowledge of the radiation port are helpful. Postra-iation therapy uptake may persist for several months.It is critical to standardize the environment of the
atient during the uptake period to examine the patientor postoperative sites, tube placement, stoma, etc., ando know the history and time of invasive procedure andherapeutic interventions to avoid misinterpretation ofDG images. In addition, a 4-h fasting period is recom-ended including no consumption of beverages with
ugar and no intravenous dextrose; a 12-h fasting periods better if the chest is evaluated to prevent myocardialptake. Drinking water should be encouraged to keep theatient hydrated and promote diuresis, which will de-rease activity in the renal collecting system and theladder. Patients are advised to avoid strenuous exerciseor the preceding 24 h.
DIAGNOSIS AND INITIAL STAGING OFCOLORECTAL CANCER
Colorectal cancer is the third most common cause ofancer in men and women and affects 5% of theopulation in the United States and most western coun-ries. The American Cancer Society estimates that therere approximately 135,000 new cases of colorectalancer per year in the United States and approximately7,000 patients per year die from this disease in thenited States, representing 10% of all cancer deaths.pproximately 70-80% of patients are treated with
urative intent and the overall survival at 5 years is lesshan 60%. The diagnosis of colorectal carcinoma isased on colonoscopy and biopsy. The preoperativetaging with imaging modalities is usually limited be-ause most patients will benefit from colectomy torevent intestinal obstruction. The extent of the diseasean be evaluated during surgery.
Three studies have been performed to evaluate theerformance of FDG-PET in the initial staging of colo-ectal cancer. Abdel-Nabi and coworkers16 evaluated thesefulness of FDG-PET for staging patients with knownr suspected primary colorectal carcinomas. In 48 pa-ients, FDG-PET imaging identified all primary carcino-
as. They found that FDG and CT were equally poorlyensitive for detecting local lymph node involvement,oth with a sensitivity of 29%. FDG-PET was, however,uperior to CT for detecting hepatic metastases, withensitivity and specificity of 88% and 100% respectivelyompared with 38% and 97% for CT. These data wereonfirmed in the studies of Mukai and coworkers17 andantorova and coworkers,18 which also reported thatDG-PET changed the treatment modality in 8% ofatients and the range of surgery in 13%. False-positivendings include abscesses, fistulas, diverticulitis and
ccasionally adenomas. Figure 1 illustrates the example wf a patient presenting with multiple hepatic lesions onT, a biopsy revealed adenocarcinoma and no primaryas found. FDG-PET-CT imaging identified the primary
olon carcinoma (the color version of this figure isvailable online).
In addition, a study of 110 patients has demonstratedhat these precancerous adenomatous polyps can beetected incidentally on whole body images performedor other indications with a sensitivity of 24% (24/59).he size of the lesions ranged from 5 to 30 mm. Theositivity rate increased to 90% for lesions greater than3 mm in size, and the false-positive rate was 5.5%6/10).19 Although PET is not recommended for detec-ion or screening for precancerous or malignant coloniceoplasms, the identification of focal colon uptakehould not be ignored.
Although the sensitivity of FDG-PET for the detectionf a primary colon carcinoma may be high, its role in thereoperative staging is still debated except in high-riskatients for whom surgery can be avoided if metastasesre identified.
DETECTION OF RECURRENT OR METASTATICCOLORECTAL CARCINOMA
Approximately 70% of the patients are resectable withurative intent but recurrence is noted in one third ofhese patients in the first 2 years after resection. Twenty-ve percent of these patients have recurrence limited tone site and are potentially curable by surgical resec-ion.20 For example, about 14,000 patients per yearresent with isolated liver metastases as their firstecurrence, and about 20% of these patients die withetastases exclusively to the liver.21 Hepatic resection is
he only curative therapy in these patients, but it isssociated with a mortality of 2 to 7% and has theotential for significant morbidity.22 Early detection andrompt treatment of recurrences may lead to a cure in upo 25% of patients. However, the size and number ofepatic metastases and the presence of extra-hepaticisease affect the prognosis. The poor prognosis ofxtra-hepatic metastases is believed to be a contraindi-ation to hepatic resection.23 Therefore, accurate nonin-asive detection of inoperable disease with imagingodalities plays a pivotal role in selecting patients whoould benefit from surgery.
onventional Modalities for Detecting andtaging Recurrence
The measurement of serum levels of carcinoembry-nic antigen may be used to monitor the detection ofecurrence with a sensitivity of 59% and specificity of4% but does not localize recurrent lesions.24 Bariumtudies have been used for detection of local recurrence
ith accuracy in the range of 80%. However, barium
s8
lf
pipm
w
(
i
l
u
c
212 DELBEKE AND MARTIN
tudies have been reported to be only 49% sensitive and5% specific for overall recurrence.25
CT has been the conventional imaging modality used toocalize recurrence with an accuracy of 25 to 73%, but it
Fig 1. A 45-year-old female with multiple hepatic lesions
orkup failed to demonstrate a primary tumor. Whole-bo
FDG-PET) imaging was performed using an integrated PET-co
mages, FDG-PET images, and fusion images. A, FDG-PET ma
esions in the liver that are FDG-avid and 2) a focus of uptake in
pper pelvis demonstrates that the focus of uptake correspo
arcinoma. A repeat colonoscopy revealed colon carcinoma.
ails to demonstrate hepatic metastases in up to 7% of c
atients and underestimates the number of lobes involvedn up to 33% of patients. In addition, metastases to theeritoneum, mesentery and lymph nodes are commonlyissed on CT, and the differentiation of postsurgical
ound to have adenocarcinoma at biopsy. The conventional
sitron emission tomography using 18F-fluorodeoxyglucose
tomography (CT) imaging system providing transmission CT
intensity projection (MIP) image demonstrates: 1) multiple
ht upper pelvis. B, A PET-CT transaxial view through the right
lesion in the wall of the cecum suggesting a primary colon
was f
dy po
mputed
ximum
the rig
nds to a
hanges from local tumor recurrence is often equivo-
cfaa(cp
rosa
DCI
Frart
bfFcctoiSa
dCFswertFarwtaw
Pstmw
vatCtem(dorlcafsP7
PCnCrcoPpA
sltmtapmne3fDlnpiwniei
dP
213PET AND PET-CT FOR COLORECTAL CARCINOMA
al.26-30 Among the patients with negative CT, 50% will beound to have nonresectable lesions at the time of explor-tory laparotomy. CT portography (superior mesentericrterial portography) is more sensitive (80 to 90%) than CT70 to 80%) for detection of hepatic metastases, but has aonsiderable rate of false-positive findings, lowering theositive predictive value.31-34
In patients undergoing exploration for recurrent colo-ectal cancer, the presence of adhesions or the limitationsf surgical exposure (transverse upper abdominal inci-ion for liver resection) often preclude a detailed oper-tive staging.
etection and Staging Recurrentolorectal Carcinoma with FDG-PET
maging
A number of studies have demonstrated the role ofDG-PET as a functional imaging modality for detectingecurrent or metastatic colorectal carcinoma.35-56 Over-ll, the sensitivity of FDG-PET imaging is in the 90%ange and the specificity greater than 70%, both superioro CT.
However, false-negative FDG-PET findings haveeen reported with mucinous adenocarcinoma. White-ord and coworkers57 reported that the sensitivity ofDG-PET imaging for detection of mucinous adenocar-inoma (n � 16) is significantly lower than the nonmu-inous adenocarcinoma (n � 93), 58% and 92%, respec-ively (P � 0.005). They suspect that the low sensitivityf FDG-PET for detection of mucinous adenocarcinomas due to the relative hypocellularity of these tumors.imilar findings (41% sensitivity) have been reported insubsequent series of 22 patients.58
Several studies have compared FDG-PET and CT forifferentiation of scar from local recurrence.36,37,40-42,46
T was equivocal in most cases and the accuracy ofDG-PET imaging was greater than 90%. In the largesttudy (76 patients),42 the accuracy of FDG-PET and CTere 95% and 65%, respectively. Figure 2 shows an
xample of a common clinical scenario: a patient iseferred with rising CEA levels and a negative conven-ional workup; local recurrence is demonstrated onDG-PET-CT images (the color version of this figure isvailable online). This case also illustrates that concur-ent PET-CT imaging permits a definite diagnosishereas identification of pathological FDG uptake along
he transverse colon would be equivocal on PET alonend a subtle soft tissue density at the anastomotic siteould be equivocal on CT alone.Other studies have compared the accuracy of FDG-
ET and CT for detection of hepatic metasta-es.42,43,45,46,48 Overall, FDG-PET was more accuratehan CT. However, most of these studies suffered from aajor limitation: PET was performed prospectively
hile CT was reviewed retrospectively and performed at aarious institutions, resulting in variable quality. Vitoland coworkers43 and Delbeke and coworkers45 reportedhe comparison of FDG with CT and CT portography.T portography, which is more invasive and more costly
han FDG-PET or CT alone, is regarded as the mostffective means of determining resectability of hepaticetastasis by imaging. FDG-PET had a higher accuracy
92%) than CT (78%) and CT portography (80%) foretection of hepatic metastases. Although the sensitivityf FDG-PET (91%) was lower than that of CT portog-aphy (97%), the specificity was much higher, particu-arly at postsurgical sites. A meta-analysis performed toompare noninvasive imaging methods (US, CT, MRI,nd FDG-PET) for the detection of hepatic metastasesrom colorectal, gastric and esophageal cancers demon-trated that at an equivalent specificity of 85%, FDG-ET had the highest sensitivity of 90% compared with6% for MRI, 72% for CT and 55% for US.59
Flanagan and coworkers47 reported the use of FDG-ET in 22 patients with unexplained elevation of serumEA level after resection of colorectal carcinoma, ando abnormal findings on conventional workup, includingT. Sensitivity and specificity of FDG-PET for tumor
ecurrence were 100% and 71% respectively. Valk andoworkers48 reported sensitivity of 93% and specificityf 92% in a similar group of 18 patients. In both studies,ET correctly demonstrated tumor in two-thirds ofatients with rising CEA levels and negative CT scans.n example is illustrated in Fig 2.Valk and coworkers48 compared the sensitivity and
pecificity of FDG-PET and CT for specific anatomicocations and found that FDG-PET was more sensitivehan CT in all locations except the lung, where the twoodalities were equivalent. The largest difference be-
ween PET and CT was found in the abdomen, pelvisnd retroperitoneum, where over one-third of PET-ositive lesions were negative by CT. PET was alsoore specific than CT at all sites except the retroperito-
eum, but the differences were smaller than the differ-nces in sensitivity. Lai and coworkers44 in their study of4 patients found that FDG-PET was especially usefulor detecting retroperitoneal and pulmonary metastases.elbeke and coworkers45 concluded that outside the
iver, FDG-PET was especially helpful in detectingodal involvement, differentiating local recurrence fromostsurgical changes, and evaluating the malignancy ofndeterminate pulmonary nodules—indications forhich CT has known limitations. In addition, by theature of being a whole-body technique, FDG-PETmaging allowed identification of distant metastatic dis-ase in the chest, abdomen, or pelvis, which might not ben the field of view of routine CT staging exams.
A meta-analysis of 11 clinical reports and 577 patientsetermined that the sensitivity and specificity of FDG-ET for detecting recurrent colorectal cancer were 97%
60
nd 76% respectively. A comprehensive review of the
f
p
i
t
P
o
P
b
214 DELBEKE AND MARTIN
Fig 2. A 63-year-old male with prior colectomy for carcinoma presented with rising serum CEA levels; conventional workup
ailed to reveal a recurrence. Whole-body positron emission tomography using 18F-fluorodeoxyglucose (FDG-PET) imaging was
erformed using an integrated PET-computed tomography (CT) imaging system providing transmission CT images, FDG-PET
mages, and fusion images. A, FDG-PET MIP image demonstrates: 1) A focus of uptake in the left upper abdomen projecting over
he hilum of the left kidney and 2) Mild FDG uptake along the laparotomy mid-line scar caused by inflammatory changes. B, A
ET-CT transaxial view through the right upper pelvis demonstrates that the focus of uptake seen on PET corresponds to the wall
f the transverse colon in the region of the anastomosis indicating local recurrence. This case also illustrates that concurrent
ET-CT imaging permits a definitive diagnosis whereas identification of pathological FDG uptake along the transverse colon would
e equivocal on PET alone and a subtle soft tissue density at the anastomotic site would be equivocal on CT alone.
Pwia
tpo(ncoabCivcdip1
cgiIlPdc
CpwomicCpeaIovuaHpimwodt
ucNucttbwiSd
otdse
IM
dfbipuoaswiIssccs6mt5cdbbmcTrr
t
215PET AND PET-CT FOR COLORECTAL CARCINOMA
ET literature (2244 patient’s studies) has reported aeighted average for FDG-PET sensitivity and specific-
ty of 94% and 87% respectively compared with 79%nd 73% for CT.61
Concurrent PET-CT imaging with an integrated sys-em may be especially important in the abdomen andelvis. PET images alone may be difficult to interpretwing to both the absence of anatomical landmarksother than the kidneys and bladder), the presence ofonspecific uptake in the stomach, small bowel andolon and urinary excretion of FDG. If possible, imagesf the abdomen and pelvis should be obtained with therms elevated to avoid artifacts due to motion and toeam hardening artifacts on the CT transmission images.oncurrent PET-CT imaging is helpful for differentiat-
ng focal retention of urine in the ureter for exampleersus an FDG-avid lymph node. The usefulness ofoncurrent PET-CT imaging providing fusion images forifferentiating physiologic from pathologic FDG uptaken the abdomen has been reported in a study of 28atients with abdominal tumors62 and in another study of0 patients with ovarian malignancies.63
A more recent study of 45 patients with colorectalancer referred for FDG-PET imaging using an inte-rated PET-CT system concluded that PET-CT imagingncreases the accuracy and certainty of locating lesions.n their study, the frequency of equivocal and probableesion characterization was reduced by 50% withET-CT compared with PET alone, the number ofefinite locations was increased by 25%, and the overallorrect staging increased from 78% to 89%.64
At the time of this writing, most institutions acquireT transmission images without intravenous contrast toermit optimal attenuation correction but CT imagesithout intravenous contrast do not allow visualizationf many hepatic metastases. Therefore, although hepaticetastases are commonly seen as FDG-avid on the PET
mages, no corresponding lesions are seen on the non-ontrasted CT transmission images. A standard of careT with intravenous and oral contrast need to beerformed if surgery is contemplated. Evaluation of theffects of intravenous and oral contrast agents on thettenuation correction of the PET images is ongoing.ntravenous contrast appears as regions of high contrastn CT images, especially during the arterial and arterio-enous phase of enhancement. If these CT images aresed for attenuation correction, overcorrection may creatertifacts of increased uptake on the FDG-PET images.65
igh-density oral contrast agents66,67 and metallic im-lants68 can create similar artifacts. However, the admin-stration of low-density oral contrast results in onlyinimal overcorrection and is not believed to interfereith accurate interpretation of the images.66,66 Reviewf the images without attenuation correction is helpful toiscriminate an overcorrection artifact from “true” up-
ake; and should be performed if there is abnormal vptake in a region of the body with accumulation ofontrast agents or in a region of metallic implants.akamoto and coworkers69 have compared standardptake value (SUV) measurements on PET imagesorrected for attenuation with transmission maps ob-ained using 68Germanium source and CT. They foundhat CT-based attenuation correction was overestimatedy 11% in the skeleton and 2% in soft tissue comparedith 68Germanium-based attenuation correction. It is
mportant to take these differences in consideration if theUV is used when comparing PET studies obtained withifferent protocols.Serosal metastases can usually be precisely localized
n the surface of the liver. As in the chest, the CTransmission images have to be carefully reviewed foretection of malignant lesions that may not be FDG-aviduch as mucinous tumors or renal cell carcinomas forxample.
mpact of FDG-PET Findings on Patient’sanagement
The greater sensitivity of PET compared with CT iniagnosis and staging of recurrent tumor results from twoactors: early detection of abnormal tumor metabolism,efore changes have become apparent by anatomic imag-ng, and the whole body nature of PET imaging, whichermits diagnosis of tumor when it occurs in unusual andnexpected sites. FDG-PET imaging allows the detectionf unsuspected metastases in 13-36% of patients and hasclinical impact in 14 to 65%.41,42,44-48,50,54-56,70,71 In the
tudy of Delbeke and coworkers,45 surgical managementas altered by PET in 28% of patients, in one-third by
nitiating surgery and in two-thirds by avoiding surgery.n a survey-based study of 60 referring oncologists,urgeons, and generalists, FDG-PET performed at initialtaging had a major impact on the management ofolorectal cancer patients and contributed to a change inlinical stage in 42% (80% upstaged and 20% down-taged) and a change in the clinical management in over0%. As a result of the PET findings, physicians avoidedajor surgery in 41% of patients for whom surgery was
he intended treatment.72 In a recent prospective study of1 patients evaluated for resection of hepatic metastases,linical management decisions based on conventionaliagnostic methods were changed in 20% of patientsased on the findings on FDG-PET imaging, especiallyy detecting unsuspected extrahepatic disease.71 In aeta-analysis of the literature, FDG-PET imaging
hanged the management in 29% (102/349) patients.60
he comprehensive review of the PET literature haseported a weighted average change of managementelated to FDG-PET findings in 32% of 915 patients.61
Although survival is not an endpoint for a diagnosticest, Strasberg and coworkers70 have estimated the sur-
ival of patients who underwent FDG-PET imaging in
tmactfifu
CI
atagiwactobufvffdaletfttn
gUolfawMcpPedta8m
wrtwbasrwVn
udiooue
C
retssr
tcgnar
dtrTaCPraaPie
iipc
216 DELBEKE AND MARTIN
heir preoperative evaluation for resection of hepaticetastases. The Kaplan-Meier test estimate of the over-
ll survival at three years was 77% and the loweronfidence limit was 60%. These percentages are higherhan those in previously published series that rangedrom 30% to 64%. In the patients undergoing FDG-PETmaging before hepatic resection, the three-year disease-ree survival rate was 40%, again higher than thatsually reported.
linical Impact of Concurrent PET-CTmaging
From the diagnostic point of view, the CT obtained forttenuation maps can also be used for precise localiza-ion of the foci of uptake with the help of the fusion ofnatomical and molecular images. Published data re-arding the incremental value of concurrent PET-CTmages obtained with an integrated system comparedith PET alone, or compared with PET correlated withCT obtained at a different time, are limited but
onclude the following: 1) Improvement of lesion detec-ion on both CT and FDG-PET images, 2) improvementf the localization of foci of FDG uptake resulting inetter differentiation of physiologic from pathologicptake, and 3) precise localization of the malignant foci,or example in the skeleton versus soft tissue, or liverersus adjacent bowel or node. Concurrent PET-CTusion images affect the clinical management by guidingurther procedures, excluding the need of further proce-ures, and changing both inter- and intramodality ther-py.73-77 For example, precise localization of metastaticymph nodes could result in a less invasive and morefficient surgical procedure or guide the biopsy of a masso FDG-avid regions of the tumor. Concurrent PET-CTusion images have the potential to provide better mapshan CT alone to modulate field and dose of radiationherapy including in patients with colorectal carci-oma.78,79
After performing 100 oncology studies using an inte-rated PET-CT system, the investigators at Pittsburghniversity concluded that combined PET-CT imagesffer significant advantages, including 1) more accurateocalization of foci of uptake, 2) distinction of pathologicrom physiologic uptake, and 3) improvement in guidingnd evaluating therapy.76,80 A study of 204 patients (34ith gastrointestinal tumors) performed at Rambamedical Center81 using an integrated PET-CT system
oncluded that the diagnostic accuracy of PET is im-roved in approximately 50% of patients. In that study,ET-CT fusion images improved characterization ofquivocal lesions as definitely benign in 10% of sites andefinitely malignant in 5% of sites. It precisely definedhe anatomic location of malignant FDG uptake in 6%nd led to retrospective lesion detection on PET or CT in%. The results of PET-CT images had an impact on the
anagement of 14% (28/204) of patients, 7/28 patients sith a change of management had colorectal cancerepresenting 20% (7/34) of patients with gastrointestinalumors. The changes in management in the 7 patientsith colorectal cancer included guiding colonoscopy andiopsy for a local recurrence (n � 2), guiding biopsy tometastatic supraclavicular lymph node (n � 1), guiding
urgery to localized metastatic lymph nodes (n � 3) andeferral to chemotherapy (n � 2). Similar conclusionsere found in a study of 173 patients performed atanderbilt University, 24 of which had colorectal carci-oma.82
It is also important to be aware of the potentiallyseful additional information provided by the indepen-ent interpretation by a radiologist experienced in bodymaging of the noncontrasted CT portion of the studybtained with integrated PET-CT systems. An analysisf 250 patients demonstrated that these findings arencommon (3% of patients) but could be importantnough to warrant alterations in clinical management.83
ost Analysis
Including FDG-PET in the evaluation of patients withecurrent colorectal carcinoma has been shown to be costffective in a study using clinical evaluation of effec-iveness with modeling of costs and studies using deci-ion tree sensitivity analysis.48,84,85 In both type oftudies, all costs calculations were based on Medicareeimbursement rates and a $1800 cost for a PET scan.
In a management algorithm where recurrence at morehan one site was treated as nonresectable, Valk andoworkers48 evaluated cost savings in 78 patients under-oing preoperative staging of recurrent colorectal carci-oma. This study was limited to preoperative patients,nd demonstrated potential savings of $3003/patientesulting from diagnosis of nonresectable tumor by PET.
In 1997, Gambhir and coworkers84 used a quantitativeecision tree model combined with sensitivity analysiso evaluate cost issues if all patients presenting withecurrent colorectal cancer undergo FDG-PET imaging.he conventional strategy for detection of recurrencend determination of resectability using CEA levels andT was compared with the conventional strategy plusET for all patients presenting with suspected recur-ence. The assumptions included prevalence of resect-ble disease of 3%, sensitivity and specificity of 65%nd 45% respectively for CT, and 90% and 85% forET. The conventional strategy plus PET showed an
ncremental saving of $220/patient without a loss of lifexpectancy.
Park and coworkers84 used the decision tree sensitiv-ty analysis to evaluate the cost of adding FDG-PETmaging in the evaluation of patients referred for sus-ected recurrence based on elevated CEA levels andandidates for hepatic resection. The CT plus PET
trategy was higher in mean cost by $429 per patient, but
rd
aw5hvtrCFfcmaptatc6rosdir
mptmsccthetmRaVFpgwlcsdl
sFsbipomsauoitnehl
lbiptovidnpt
mtssicimmc
tmpliatpc
217PET AND PET-CT FOR COLORECTAL CARCINOMA
esulted in an increase in the mean life expectancy of 9.5ays per patient.
FDG IMAGING TO MONITOR THERAPY OFCOLORECTAL CARCINOMA
FDG-PET is most helpful to monitor patients withdvanced-stage colorectal carcinoma that is associatedith a poor prognosis. Systemic chemotherapy with-fluorouracil often in combination with radiotherapyas demonstrated effective palliation and improved sur-ival.86 A preliminary study on 6 patients demonstratedhe FDG uptake decreased in the primary tumor duringadiation therapy whereas the size did not change onT.87 Another study of 44 patients demonstrated thatDG-PET imaging can differentiate local recurrencerom scarring after radiation therapy.88 However, in-reased FDG uptake immediately following radiationay be due to inflammatory changes and is not always
ssociated with residual tumor. The time course ofostirradiation FDG activity has not been studied sys-ematically; it is, however, generally accepted that FDGctivity present six months after completion of radiationherapy most likely represents tumor recurrence. Aase-controlled study of 60 FDG-PET studies performedmonths following external beam radiation therapy for
ectal cancer found a sensitivity of 84% and specificityf 88% for detection of local pelvic recurrence.89 A pilottudy of 15 patients with primary rectal carcinomaemonstrated that FDG-PET imaging adds incrementalnformation for assessing the response to preoperativeadiation and 5-fluorouracil-based chemotherapy.90
Hepatic metastases can be treated with systemic che-otherapy or regional therapy to the liver. A variety of
rocedures to administer regional therapy to hepatic metas-ases have been investigated including chemotherapy ad-inistered through the hepatic artery using infusion pumps,
elective chemoembolization, radiofrequency ablation,ryoablation, alcohol ablation and radiolabeled 90Y-mi-rospheres.91-94 There are preliminary reports suggestinghat the response to chemotherapy in patients withepatic metastases can be predicted with PET. Respond-rs may be discriminated from nonresponders after fouro five weeks of chemotherapy with fluorouracil by
easuring FDG uptake before and during therapy.95
egional therapy to the liver by chemoembolization canlso be monitored with FDG-PET imaging as shown byitola and coworkers96 and Torizuka and coworkers.97
DG uptake decreases in responding lesions and theresence of residual uptake in some lesions can help inuiding further regional therapy. Langenhoff and co-orkers98 have prospectively monitored 23 patients with
iver metastases following radiofrequency ablation andryoablation. Three weeks after therapy, 51/56 metasta-es became FDG negative, and there was no recurrenceuring 16 months follow-up; whereas among the 5/56
esions with persistent FDG uptake, 4/5 recurred. Data in amaller series of patients supports their findings.99,100
igure 3 illustrates residual/recurrent tumor adjacent to aite of radiofrequency ablation detected on FDG-PETut not on the CT images (the color version of this figures available online). Wong and coworkers101 have com-ared FDG-PET imaging, CT or MRI and serum levelsf CEA to monitor the therapeutic response of hepaticetastases to 90Y-glass microspheres. They found a
ignificant difference between the FDG-PET changesnd the changes on CT or MRI; the changes in FDGptake correlated better with the changes in serum levelsf CEA. Figure 4 illustrates the use of FDG-PETmaging to monitor the efficacy of regional therapy tohe liver with 90Y-microspheres. In summary, prelimi-ary data suggest that FDG-PET imaging may be able toffectively monitor the efficacy of regional therapy toepatic metastases but these data need to be confirmed inarger series of patients.
LIMITATIONS OF FDG IMAGING
Tumor detectability depends on both the size of theesion and the degree of uptake, as well as surroundingackground uptake and intrinsic resolution of the imag-ng system. False-negative lesions can be the result ofartial volume averaging, leading to underestimation ofhe uptake in small lesions (less than twice the resolutionf the imaging system) or in necrotic lesions with a thiniable rim, falsely classifying these lesions as benignnstead of malignant. The sensitivity of FDG-PET foretection of mucinous adenocarcinoma is lower than foronmucinous adenocarcinoma (41-58% versus 92%),robably because of the relative hypocellularity of theseumors.57,58
In view of the known high uptake of FDG by activatedacrophages, neutrophils, fibroblasts and granulation
issue, it is not surprising that inflamed tissue demon-trates FDG activity. Mild-to-moderate FDG activityeen early after radiation therapy, along recent incisions,nfected incisions, biopsy sites, drainage tubing andatheters, as well as colostomy sites can lead to errors innterpretation if the history is not known. Some inflam-
atory lesions, especially granulomatous ones, may bearkedly FDG-avid and can be mistaken for malignan-
ies; this includes inflammatory bowel disease.FDG uptake normally present in the gastrointestinal
ract can occasionally be difficult to differentiate from aalignant lesion. Incidental colonic FDG uptake in 27
atients without colorectal carcinoma has been corre-ated with colonoscopic and/or histolopathologic find-ngs.102 Diffuse uptake in 8 patients was normal andssociated with a normal colonoscopy. Segmental up-ake was due to colitis in 5/6 patients. Focal uptake in 7atients was associated with benign adenomas. Thelinical history, physical examination, pattern of uptake
nd correlation with anatomy as seen on the CT images
at
hapscriSMci
Pprahmttfco
r
i
w
c
n or.
218 DELBEKE AND MARTIN
re more helpful in avoiding false-positive interpreta-ions than semiquantitative evaluation by SUV.
COST AND REIMBURSEMENT ISSUES
Until recently, the implementation of clinical PET wasindered by the high cost of PET systems, the need forccess to a cyclotron and support laboratory for FDGroduction, high maintenance and operating expenses ofcanners and cyclotrons, and lack of reimbursement forlinical procedures by third-party payers. The third-partyeimbursement situation for oncologic PET has improvedn recent years. In July 2001, the Center for Medicalervices approved and implemented reimbursement byedicare for six types of malignant tumors including
olorectal carcinoma. This coverage is for diagnosis, stag-
Fig 3. A 46-year-old female with a history of colon cance
adiofrequency ablation. Contrast-enhanced CT and positro
maging with an integrated PET-computed tomography (CT)
ith contrast revealed necrosis corresponding to the pre
orresponding PET-CT transaxial view through the dome of
ecrosis observed on CT, indicating persistent/recurrent tum
ng and restaging, but not monitoring therapy. fl
POTENTIAL NEW PET TRACERS FORCLINICAL USE
Besides evaluation of glucose metabolism with FDG,ET can assess various other biologic parameter such aserfusion, metabolism of other compounds, hypoxia andeceptor expression. Some of these radiopharmaceuticalsre labeled with positrons emitters that have a shortalf-life, such as 15O (T1/2 � 2 min), 13N (T1/2 � 10in), and 11C (T1/2 � 20 min). The short half-life of
hese radioisotopes prevents any timely distribution ofhe radiopharmaceuticals labeled with them and there-ore, their use is restricted to institutions having ayclotron and associated laboratories and personneln-site. Some tracers labeled with 18F, such as 18F-
nted with a liver metastasis and underwent treatment with
sion tomography using 18F-fluorodeoxyglucose (FDG-PET)
system were performed 2 months after therapy. A, The CT
seen hepatic metastasis at the dome of the liver. B, A
er reveals a focus of FDG uptake adjacent to the region of
r prese
n emis
imaging
viously
the liv
uorothymidine (FLT), currently are investigated for
219PET AND PET-CT FOR COLORECTAL CARCINOMA
Fig 4. A 60-year-old male with prior colectomy for carcinoma
presented with multiple hepatic metastases. He underwent regional
therapy to the right lobe of the liver with 90Y-mcrospheres. Whole-
body positron emission tomography using 18F-fluorodeoxyglucose
(FDG-PET) imaging was performed using an integrated PET-CT imag-
ing system before therapy, 2 months, and 4 months after therapy. A,
FDG-PET maximum intensity projection (MIP) image before therapy
demonstrates multiple FDG-avid hepatic metastases in both the right
and left lobe of the liver. B, FDG-PET MIP image 2 months after regional
therapy to the right lobe of the liver with 90Y-mcrospheres demon-
strates some residual FDG uptake in the right lobe hepatic metastases
indicating a good response to therapy. However, there is persistent
FDG uptake in the untreated left lobe metastases with a new focus
indicating progressive left lobe disease. C, FDG-PET MIP image 4
months after regional therapy to the right lobe of the liver with90Y-mcrospheres demonstrates increased FDG uptake in both the right
and left lobe hepatic metastases indicating progression of disease. A
new FDG-avid focus is also seen at the base of the right lung indicating
a new pulmonary metastasis.
cp
T
a9
WoaAststsivIdlwdbIgocif
T
1
anvT
mtrhrlwwlaatfpctwFimTawe
L
fraib5tev
IY
n
gb
PoC
t3
n
220 DELBEKE AND MARTIN
linical use and may have applications for evaluation ofatients with colorectal carcinoma.
racer of Bone Metabolism18F-fluoride was first described as a skeletal imaging
gent in the 1960’s but then was replaced by the9mTc-labeled diphosphonate radiopharmaceuticals.103
ith the widespread applications of FDG-PET in oncol-gy, PET imaging systems are becoming more widelyvailable, and there is a renewed interest in18F-fluoride.lthough the mechanism of uptake for 18F-fluoride is
imilar to that for other bone-imaging radiopharmaceu-icals,104 the spatial resolution of the PET technology isuperior to that of both planar and SPECT imaging usinghe 99mTc-radiopharmaceuticals. Because of the betterpatial resolution and routine acquisition of tomographicmages, 18F-fluoride PET imaging offers potential ad-antages over bone scintigraphy in detecting metastases.n a study of 44 patients, Schirrmeister and coworkers105
emonstrated that twice as many benign and malignantesions were detected with 18F-fluoride PET comparedith planar scintigraphy. It was also possible to betterifferentiate benign from malignant lesions with PETecause of the better resolution, particularly in the spine.n a further study, the same authors demonstrated thereater accuracy of 18F-fluoride PET leading to a changef management in a group of patients with breastancer.106 Although skeletal metastases are not commonn colorectal cancer, 18F-fluoride may have a role in theuture if skeletal metastases are suspected clinically.
racers of DNA Synthesis
The rate of DNA synthesis can be assessed using1C-thymidine or FLT. Thymidine is a DNA precursor andllows direct assessment of tumor proliferation. In the earlyineties, Higashi and coworkers107 demonstrated that initro uptake correlates with the tumor proliferative rate.
hen, other investigators demonstrated in an animal tumor rREFERENC
ated glucose analog, 2-fluoro-2-deoxy-2-D-glucose [18F]:
N6
tffl
ldt3
Pt
A2
odel that uptake of 11C-thymidine correlated with viableumor cells better than FDG uptake cells after fractionatedadiotherapy.108 More recently, Shields and coworkers109
ave developed and evaluated 18FLT, a more promisingadiopharmaceutical for clinical use because of its 18Fabeling. Dittman and coworkers110 evaluated 16 patientsith thoracic tumors using both FLT and FDG. Comparedith FDG, FLT uptake was lower but with a significant
inear correlation. The authors concluded that FLT PETccurately visualizes thoracic tumors but that in the livernd bone marrow, high physiologic uptake prevents detec-ion of metastases. On the other hand, FLT may beavorable for imaging cerebral metastases owing to the lowhysiologic uptake. A report of 17 patients with colorectalancer comparing FDG and FLT demonstrated all primaryumors were visualized with both tracers but FDG uptakeas on average two-fold higher with FDG compared withLT.111 Pulmonary and peritoneal metastases were visual-
zed with both tracers, but the sensitivity of FLT for hepaticetastases was only 34% compared with 97% for FDG.his was due to the high physiologic hepatic backgroundctivity with FLT. Therefore, the authors concluded that itas unlikely that FLT would play an important role for
valuation of patients with colorectal carcinoma.
abeled Drugs
5-Fluorouracil is the mainstay chemotherapeutic agentor treatment of colorectal carcinoma and 18F-5-fluorou-acil is biochemically similar to 5-fluorouracil. Utilizingkinetic modeling approach with18F-fluorouracil, PET
maging has been used to study the influence of theiomodulator folinic acid on intracellular trapping of-fluorouracil within hepatic metastases with the expec-ation that this would correlate with the therapeuticffect.112 Trapping within hepatic metastases can beariable and 18F-5-fluorouracil PET can predict the
113,114
esponse to therapy and prognosis.ES
1. Delbeke D, Martin WH, Patton JA, et al. Practical FDGmaging: A teaching file. Springer Verlag, New York, Nework, 20022. Warburg O: Versuche und uberledbeudem carci-
omgewebe (methoden). Biochem Z 142:317-333, 19233. Flier JS, Mueckler MM, Usher P, et al: Elevated levels of
lucose transport and transporter messenger RNA are inducedy rats or src oncogenes. Science 235:1492-1495, 19874. Monakhov NK, Neistadt EI, Shaylovskii MM, et al:
hysiochemical properties and isoenzyme composition of hex-kinase from normal and malignant human tissues. J Natlancer Inst 61:27-34, 19785. Knox WE, Jamdar SC, Davis PA: Hexokinase, differen-
iation, and growth rates of transplanted tumors. Cancer Res0:2240-2244, 19706. Som P, Atkins HL, Bandoypadhayay D, et al: A fluori-
ontoxic tracer for rapid tumor detection. J Nucl Med 21:670-75, 19807. Gallagher BM, Fowler JS, Gutterson NI, et al: Metabolic
rapping as a principle of radiopharmaceutical design: Someactors responsible for the biodistribution of [18F]2-deoxy-2-uoro-D-glucose. J Nucl Med 19:1154-1161, 19788. Cook GJR, Fogelman I, Maisey MN: Normal physio-
ogical and benign pathological variants of 18-fluoro-2-eoxyglucose positron emission tomography scanning: Po-ential for error in interpretation. Semin Nucl Med 26:308-14, 19969. Engel H, Steinert H, Buck A, et al: Whole body PET:
hysiological and artifactual fluorodeoxyglucose accumula-ions. J Nucl Med 37:441-446, 1996
10. Hutton BF, Braun M: Software for image registration:lgorithms, accuracy, efficacy. Semin Nucl Med 33:180-192,
003
n3
o
sP
lw2
dam
ogC
eec
1pd
d
st
1
mp
ot1
em9
ct
pst
et6
Rt1
cr
Pa1
mU5
tC1
dC
pl
at
c1
a1
rF
sr
ec
od1
bC2
et
wo1
m1
to
221PET AND PET-CT FOR COLORECTAL CARCINOMA
11. Townsend DW, Beyer T, Bloggett TM: PET/CT scan-ers: A hardware approach to image fusion. Semin Nucl Med3:193-204, 200312. Bakheet SM, Powe J: Benign causes of 18-FDG uptake
n whole body imaging. Semin Nucl Med 28:352-358, 199813. Cohade C, Osman M, Pannu HK, et al: Uptake in
upraclavicular area fat (“USA-Fat”): description on 18F-FDGET/CT. J Nucl Med 44:170-176, 200314. Kamel EM, Goerres GW, Burger C, et al: Recurrent
aryngeal nerve palsy in patients with lung cancer: Detectionith PET-CT image fusion-Report od six cases. Radiology24:153-156, 200215. Kubota R, Yamada S, Kubota K, et al: Intratumoral
istribution of fluorine-18-fluorodeoxyglucose in vivo: highccumulation in macrophages and granulocytes studied byicroautoradiography. J Nucl Med 33:1972-1980, 199216. Abdel-Nabi H, Doerr RJ, Lamonica DM, et al: Staging
f primary colorectal carcinomas with fluorine-18 fluorodeoxy-lucose whole-body PET: Correlation with histopathologic andT findings. Radiology 206:755-760, 199817. Mukai M, Sadahiro S, Yasuda S, et al: Preoperative
valuation by whole-body 18F-fluorodeoxyglucose positronmission tomography in patients with primary colorectal can-er. Oncol Rep 7:85-87, 2000
18. Kantorova I, Lipska L, Belohlavek O, et al: Routine8F-FDG PET preoperative staging of colorectal cancer: Com-arison with conventional staging and its impact on treatmentecision making. J Nucl Med 44:1784-1788, 200319. Yasuda S, Fujii H, Nakahara T, et al: 18F-FDG PET
etection of colonic adenomas. J Nucl Med 42:989-992, 200120. August DA, Ottow RT, Sugarbaker PH: Clinical per-
pectives on human colorectal cancer metastases. Cancer Me-astases Rev 3:303-324, 1984
21. Foster JH, Lundy J: Liver metastases. Curr Probl Surg8:158-202, 198122. Holm A, Bradley E, Aldrete J: Hepatic resection ofetastases from colorectal carcinoma: morality, morbidity and
attern of recurrence. Ann Surg 209:428-434, 198923. Hughes KS, Simon R, Songhorabodi S, et al: Resection
f liver for colorectal carcinoma metastases: a multi-institu-ional study of indications for resection. Surgery 103:278-288,98824. Moertel CG, Fleming TR, McDonald JS, et al: An
valuation of the carcinoembryonic antigen (CEA) test foronitoring patients with resected colon cancer. JAMA 270:
43-947, 199325. Chen YM, Ott DJ, Wolfman NT, et al: Recurrent
olorectal carcinoma: Evaluation with barium enema examina-ion and CT. Radiology 163:307-310, 1987
26. Sugarbaker PH, Grianola FJ, Dwyer S, et al: A simplifiedlan for follow-up of patients with colon and rectal cancerupported by prospective studies of laboratory and radiologicest results. Surgery 102:79-87, 1987
27. Steele G Jr, Bleday R, Mayer R, et al: A prospectivevaluation of hepatic resection for colorectal carcinoma metas-ases to the liver: Gastrointestinal Tumor Study Group protocol584. J Clin Oncol 9:1105-1112, 199128. Granfield CAJ, Charnsangaveg C, Dubrow RA, et al:
egional lymph node metastases in carcinoma of the left side ofhe colon and rectum: CT demonstration. AJR 159:757-761,
992 A29. Charnsangavej C, Whitley NO: Metastases to the pan-reas and peripancreatic lymph nodes from carcinoma of theight colon: CT findings in 12 patients. AJR 160:49-52, 1993
30. McDaniel KP, Charnsangavej C, Dubrow R, et al:athways of nodal metastases in carcinoma of the cecum,scending colon and transverse colon: CT demonstration. AJR61:61-64, 199331. Soyer P, Levesque M, Elias D, et al: Detection of liveretastases from colorectal cancer: comparison of intraoperativeS and CT during arterial portography. Radiology 183:541-44, 199232. Nelson RC, Chezmar JL, Sugarbaker PH, et al: Hepatic
umors: comparison of CT during arterial portography, delayedT and MR imaging for preoperative evaluation. Radiology72:27-34, 198933. Small WC, Mehard WB, Langmo LS, et al: Preoperative
etermination of the resectability of hepatic tumors: efficacy ofT during arterial portography. AJR 161:319-322, 199334. Peterson MS, Baron RL, Dodd GD III, et al: Hepatic
arenchymal perfusion detected with CTPA: imaging-patho-ogic correlation. Radiology 183:149-155, 1992
35. Yonekura Y, Benua RS, Brill AB, et al: Increasedccumulation of 2-deoxy-2[18F]fluoro-D- glucose in liver me-astases from colon carcinoma. J Nucl Med 23:1133-1137, 1982
36. Strauss LG, Clorius JH, Schlag P, et al: Recurrence ofolorectal tumors: PET evaluation. Radiology 170:329-332,98937. Ito K, Kato T, Tadokoro M, et al: Recurrent rectal cancer
nd scar: Differentiation with PET and MR imaging. Radiology82:549-552, 199238. Kim EE, Chung SK, Haynie TP, et al: Differentiation of
esidual or recurrent tumors from post-treatment changes with-18 FDG-PET. Radiographics 12:269-279, 199239. Gupta NC, Falk PM, Frank AL, et al: Pre-operative
taging of colorectal carcinoma using positron emission tomog-aphy. Nebr Med J 78:30-35, 1993
40. Falk PM, Gupta NC, Thorson AG, et al: Positronmission tomography for preoperative staging of colorectalarcinoma. Dis Colon Rectum 37:153-156, 1994
41. Beets G, Penninckx F, Schiepers C, et al: Clinical valuef whole-body positron emission tomography with [18F]fluoro-eoxyglucose in recurrent colorectal cancer. Br J Surg 81:1666-670, 199442. Schiepers C, Penninckx F, De Vadder N, et al: Contri-
ution of PET in the diagnosis of recurrent colorectal cancer:omparison with conventional imaging. Eur J Surg Oncol1:517-522, 199543. Vitola JV, Delbeke D, Sandler MP, et al: Positron
mission tomography to stage metastatic colorectal carcinomao the liver. Am J Surg 171:21-26, 1996
44. Lai DT, Fulham M, Stephen MS, et al: The role ofhole-body positron emission tomography with [18F]fluorode-xyglucose in identifying operable colorectal cancer. Arch Surg31:703-707, 199645. Delbeke D, Vitola J, Sandler MP, et al: Staging recurrentetastatic colorectal carcinoma with PET. J Nucl Med 38:1196-
201, 199746. Ogunbiyi OA, Flanagan FL, Dehdashti F, et al: Detec-
ion of recurrent and metastatic colorectal cancer: comparisonf positron emission tomography and computed tomography.
nn Surg Oncol 4:613-620, 1997
Fi
Wm5
orC
vrc
i
st
Fc2
1c
flcA
st2
nsC
ew1
mnm
ac
r
FC
r4
FM
tm
mCJ
es
iE
ca4
pmA
piJ
Fr
P2
o(t
isdm
RUI
fr
im
tpO
is
pv
222 DELBEKE AND MARTIN
47. Flanagan FL, Dehdashti F, Ogunbiyi OA, et al: Utility ofDG PET for investigating unexplained plasma CEA elevation
n patients with colorectal cancer. Ann Surg 227:319-323, 199848. Valk PE, Abella-Columna E, Haseman MK, et al:hole-body PET imaging with F-18-fluorodeoxyglucose inanagement of recurrent colorectal cancer. Arch Surg 134:503-
11, 199949. Ruhlmann J, Schomburg A, Bender H, et al: Fluorode-
xyglucose whole-body positron emission tomography in colo-ectal cancer patients studied in routine daily practice. Disolon Rectum 40:1195-1204, 199750. Flamen P, Stroobants S, Van Cutsem E, et al: Additional
alue of whole-body positron emission tomography with fluo-ine-18-2-fluoro-2-deoxy-D-glucose in recurrent colorectal can-er. J Clin Oncol 17:894-901, 1999
51. Akhurst T, Larson SM: Positron emission tomographymaging of colorectal cancer. Semin Oncol 26:577-583, 1999
52. Vogel SB, Drane WE, Ros PR, et al: Prediction ofurgical resectability in patients with hepatic colorectal metas-ases. Ann Surg 219:508-516, 1994
53. Imbriaco M, Akhurst T, Hilton S, et al: Whole-BodyDG-PET in patients with recurrent colorectal carcinoma. Aomparative study with CT. Clin Positron Imaging 3:107-114,00054. Imdahl A, Reinhardt MJ, Nitzsche EU, et al: Impact of
8F-FDG-positron emission tomography for decision making inolorectal cancer recurrences. Arch Surg 385:129-134, 2000
55. Staib L, Schirrmeister H, Reske SN, et al: Is (18)F-uorodeoxyglucose positron emission tomography in recurrentolorectal cancer a contribution to surgical decision making?m J Surg 180:1-5, 200056. Kalff VV, Hicks R, Ware R: F-18 FDG PET for
uspected or confirmed recurrence of colon cancer. A prospec-ive study of impact and outcome. Clin Pos Imaging 3:183,00057. Whiteford MH, Whiteford HM, Yee LF, et al: Useful-
ess of FDG-PET scan in the assessment of suspected meta-tatic or recurrent adenocarcinoma of the colon and rectum. Disolon Rectum 43:759-767discussion 767-770, 200058. Berger KL, Nicholson SA, Dehadashti F, et al: FDG PET
valuation of mucinous neoplasms: correlation of FDG uptakeith histopathologic features. Am J Roentgenol 174:1005-008, 200059. Kinkel K, Lu Y, Both M, et al: Detection of hepaticetastases from cancers of the gastrointestinal tract by using
oninvasive imaging methods (US, CT, MR imaging, PET): Aeta-analysis. Radiology 224:748-756, 200260. Huebner RH, Park KC, Shepherd JE, et al: A meta-
nalysis of the literature for whole-body FDG PET detection ofolorectal cancer. J Nucl Med 41:1177-1189, 2000
61. Gambhir SS, Czernin J, Schimmer J, et al: A tabulatedeview of the literature. J Nucl Med 42:9S-12S, 2001 (suppl)
62. Meltzer CC, Martinelli MA, Beyer T, et al: Whole-bodyDG PET imaging in the abdomen: Value of combined PET-T. J Nucl Med 42:35P, 200163. Meltzer CC, Makhija S, Howden N, et al: PET/CT in
ecurrent ovarian and fallopian tube carcinoma. J Nucl Med2:284P, 200164. Cohade C, Osman M, Leal J, et al: Direct comparison of
DG PET and PET-CT imaging in colorectal carcinoma. J Nucl
ed 44:1797-1803, 2003 M65. Antoch G, Freudenberg LS, Egeldorf T, et al: Focalracer uptake: A potential artifact in contrasted-enhanced dual-odality PET/CT scans. J Nucl Med 43:1339-1342, 200266. Dizendorf E, Hany TF, Buck A, et al: Cause andagnitude of the error induced by oral CT contrast agent inT-based attenuation correction of PET emission studies.Nucl Med 44:732-738, 200367. Cohade C, Osman M, Nakamoto Y, et al: Initial experi-
nce with oral contrast in PET/CT: Phantom and clinicaltudies. J Nucl Med 44:412-416, 2003
68. Goerres GW, Hany TF, Kamel E, et al: Head and neckmaging with PET/CT: Artifacts from dental metallic implants.ur J Nucl Med Mol Imaging 29:367-370, 200269. Nakamoto Y, Osman M, Cohade C, et al: PET-CT:
omparison of quantitative tracer uptake between germaniumnd CT transmission attenuation-corrected images. J Nucl Med3:1137-1143, 200270. Strasberg SM, Dehdashti F, Siegel BA, et al: Survival of
atients evaluated by FDG PET before hepatic resection foretastatic colorectal carcinoma: A prospective database study.nn Surg 233:320-321, 200171. Ruers TJ, Langenhoff BS, Neeleman N, et al: Value of
ositron emission tomography with [F-18] fluorodeoxyglucosen patients with colorectal liver metastases: A prospective study.Clin Oncol 20:388-395, 200272. Meta J, Seltzer M, Schiepers C, et al: Impact of 18F-
DG PET on managing patients with colorectal cancer: Theeferring physicians’ perspective. J Nucl Med 42:586-590, 2001
73. Alyafei S, Inoue T, Zhang H, et al: Image fusion usingACS for MRI, CT, and PET images. Clin Pos Imaging:137-143, 199974. Israel O, Mor M, Guralnik L, et al: The new technology
f combined transmission and emission F-18 FDG tomographyFDG-TET) in the diagnosis and management of cancer pa-ients. Clin Pos Imaging 3:143, 2000
75. Delbeke D, Martin WH, Patton JA, et al: Value ofterative reconstruction, attenuation correction, and image fu-ion in the interpretation of FDG PET images with an integratedual-head coincidence camera and x-ray-based attenuationaps. Radiology 218:163-171, 200176. Martinelli M, Townsend D, Meltzer C, et al: Survey of
esults of whole body imaging using the PET/CT at theniversity of Pittsburgh Medical Center PET facility. Clin Pos
maging 3:161, 200077. Yeung HW, Schoder H, Larson SM: Utility of PET/CT
or assessing equivocal PET lesions in oncology—initial expe-ience. J Nucl Med 43:32P, 2002
78. Dizendorf E, Ciernik IF, Baumert B, et al: Impact ofntegrated PETCT scanning on external beam radiation treat-ent planning. J Nucl Med 43:33P, 200279. Ciernik IF, Dizendorf E, Baumert BG, et al: Radiation
reatment planning with integrated positron emission and com-uted tomography (PET/CT): A feasibility study. Int J Radiatncol Biol Phys 57:853-863, 200380. Charron M, Beyer T, Bohnen NN, et al: Image analysis
n patients with cancer studied with a combined PET and CTcanner. Clin Nucl Med 25:905-910, 2000
81. Bar-Shalom R, Yefremov N, Guralnik L, et al: Clinicalerformance of PET/CT in the evaluation of cancer: Additionalalue for diagnostic imaging and patient management. J Nucl
ed 44:1200-1209, 2003
ft
sP4
ao1
ar
l
c1
Pr3
cti2
mas
l9
a3
otc
An2
mptO
et2
fli
flt
m2
pt2
si5
9rC
cs2
b
rJ
is
ddt
2ac
orn
pt
PM
i2c
awm3
apt
am1
223PET AND PET-CT FOR COLORECTAL CARCINOMA
82. Roman CD, Delbeke D: Incremental diagnostic value ofusion imaging with integrated PET-CT in oncology comparedo PET alone. RSNA Scientific Program 487, 2003
83. Osman MM, Cohade C, Fishman E, et al: Clinicallyignificant incidental findings on non-contrast CT portion ofET-CT studies: Frequency in 250 patients. J Nucl Med3:307P, 200284. Gambhir SS, Valk P, Shepherd J, et al: Cost effective
nalysis modeling of the role of FDG-PET in the managementf patients with recurrent colorectal cancer. J Nucl Med 38:90P,99785. Park KC, Schwimmer J, Sheperd JE, et al: Decision
nalysis for the cost-effective management of recurrent colo-ectal cancer. Ann Surg 233:310-319, 2001
86. Bertino JR: Biomodulation of 5-fluorouracil with antifo-ates. Semin Oncol 24:S18-S56, 1997 (suppl)
87. Strauss LG, Clorius JH, Schlag P, et al: Recurrence ofolorectal tumors: PET evaluation. Radiology 170:329-332,98988. Haberkorn U, Strauss LG, Dimitrakopoulou A, et al:
ET studies of fluorodeoxyglucose metabolism in patients withecurrent colorectal tumors receiving radiotherapy. J Nucl Med1:1485-1490, 199189. Moore HG, Akhurst T, Larson SM, et al: A case
ontrolled study of 18-fluorodeoxyglucose positron emissionomography in the detection of pelvic recurrence in previouslyrradiated rectal cancer patients. J Am Coll Surg 197:22-28,00390. Guillem J, Calle J, Akhurst T, et al: Prospective assess-ent of primary rectal cancer response to preoperative radiation
nd chemotherapy using 18-Fluorodeoxyglucose positron emis-ion tomography. Dis Colon Rectum 43:18-24, 2000
91. Liu LX, Zhang WH, Jiang HC: Current treatment foriver metastases from colorectal cancer. World J Gastroenterol:193-200, 200392. Ruers T, Bleichrodt RP: Treatment of liver metastases,
n update on the possibilities and results. Eur J Cancer8:1023-1033, 200293. Gray B, Van Hazel G, Hope M, et al: Randomized trial
f Sir-spheres plus chemotherapy vs chemotherapy alone forreating patients with liver metastases from primary large bowelancer. Ann Oncol 12:1711-1720, 2001
94. Nijsen JF, van het Schip AD, Hennink WE, et al:dvances in nuclear oncology: microspheres for internal radio-uclide therapy of liver tumours. Curr Med Chem 9:73-82,00295. Findlay M, Young H, Cunningham D, et al: Noninvasiveonitoring of tumor metabolism using fluorodeoxyglucose and
ositron emission tomography in colorectal cancer liver metas-ases: correlation with tumor response to fluorouracil. J Clinncol 14:700-708, 199696. Vitola JV, Delbeke D, Meranze SG, et al: Positron
mission tomography with F-18-fluorodeoxyglucose to evaluatehe results of hepatic chemoembolization. Cancer 78:2216-222, 199697. Torizuka T, Tamaki N, Inokuma T, et al: Value of
uorine-18-FDG PET to monitor hepatocellular carcinoma afternterventional therapy.
98. Langenhoff BS, Oyen WJ, Jager GJ, et al: Efficacy ofuorine-18-deoxyglucose positron emission tomography in de-
ecting tumor recurrence after local ablative therapy for liver
etastases: A prospective study. J Clin Oncol 20:4453-4458,00299. Anderson GS, Brinkmann F, Soulen MC, et al: FDG
ositron emission tomography in the surveillance of hepaticumors treated with radiofrequency ablation. Clin Nucl Med8:192-197, 2003100. Ludwig V, Hopper OW, Martin WH, et al: FDG-PET
urveillance of hepatic metastases from prostate cancer follow-ng radiofrequency ablation—Case report. Am Surg 69:593-98, 2003101. Wong CY, Salem R, Raman S, et al: Evaluating
0Y-glass microsphere treatment response of unresectable colo-ectal liver metastases by [18F]FDG PET: a comparison withT or MRI. Eur J Nucl Med Mol Imag 29:815-820, 2002102. Tatlidil R, Jadvar H, Bading JR, et al: Incidental
olonic fluorodeoxyglucose uptake: correlation with colono-copic and histopathologic findings. Radiology 224:783-787,002103. Blau M, Nagler W, Bender MA: A new isotope for
one scanning. J Nucl Med 3:332-334, 1962104. Bang S, Baug CA: Topographical distribution of fluo-
ide in iliac bone of a fluoride-treated osteoporotic patient.Bone Miner Res 5:S87-S89, 1990105. Schirrmeister H, Gulhman A, Elsner K, et al: Sensitivity
n detecting osseous lesions depends on anatomic location: planarcintigraphy versus 18F PET. J Nucl Med 40:1623-1629, 1999
106. Schirrmeister H, Gulhman A, Kotzerke J, et al: Earlyetection and accurate description of extent of metastatic boneisease in breast cancer with fluoride ion and positron emissionomography. J Clin Oncol 17:2381-2389, 1999
107. Higashi K, Clavo AC, Wahl RL: In vitro assessment of-fluoro-2-deoxy-D-glucose, L-methionine, and thymidine asgents to monitor early response of a human adenocarcinomaell line to radiotherapy. J Nucl Med 34:773-780, 1993
108. Reinhardt MJ, Kubota K, Yamada S, et al: Assessmentf cancer recurrence in residual tumors after fractionatedadiotherapy: A comparison of fluorodeoxyglucose, L-methio-ine and thymidine. J Nucl Med 38:280-287, 1997109. Shields AF, Grierson JR, Dohmen BM, et al: Imaging
roliferation in vivo with [F-18]FLT and positron emissionomography. Nat Med 4:1334-1336, 1998
110. Dittman H, Dohman BM, Paulsen F, et al: [(18)F]FLTET for diagnosis and staging of thoracic tumors. Eur J Nucled Mol Imaging 30:1407-1412, 2003111. Francis DL, Visvikis D, Costa DC, et al: Potential
mpact of [(18)F]3�-deoxy-3�-fluorothymidine versus [(18)F]fluoro--deoxy-d-glucose in positron emission tomography for colorectalancer. Eur J Nucl Med Imaging 30:988-994, 2003
112. Kissel J, Brix G, Bellemann ME, et al: Pharmacokineticnalysis of 5-[18F]fluorouracil tissue concentrations measuredith positron emission tomography in patients with liveretastases from colorectal adenocarcinoma. Cancer Res 57:
415-3423, 1997113. Moelher M, Dimitrakopoulou-Strauss A, Gutzler F, et
l: 18F-fluorouracil positron emission tomography and therognosis of colorectal carcinoma patients with metastases tohe liver treated with 5-fluorouracil. Cancer 83:245-253, 1998
114. Dimitrakopoulou-Strauss A, Strauss LG, Schlag P, etl: Fluorine-18-fluorouracil to predict therapy response in liveretastases from colorectal carcinoma. J Nucl Med 39:1197-
202, 1998
