Role of positron emission tomography scanning in evaluating gastrointestinal neoplasms

Role of Positron Emission Tomography Scanning in Evaluating Gastrointestinal Neoplasms N. Gupta and H. Bradfield Positron Emission Tomography (PET) is rapidly evolving into a useful imaging modality for early and accurate detection of malignant tumor sites. Several recent studies have documented improved efficacy of detecting recurrent colorectal and hepatic (primary and metastatic) tumor sites w i t h a sensitivity ranging from 92% to 100% and an accuracy of 90% to 96%. PET-FDG imaging using 2-[18F]-fluoro-2-deoxy-D-glucose has been found to be superior to computed tomography (CT) in detecting recurrent colorectal, hepatic, and abdominopelvic recurrent tumor sites from different

primary cancers. PET-FDG imaging can be a costeffective tool in the screening of patients with an elevated carcinoembryonic antigen and/or equivocal CT findings and suspected colorectal cancer. The role of PET scanning using FDG or C-11-5-HTP or C-11-LDOPA appears promising in pancreatic carcinoma and functional endocrine tumors. Further studies are being carried out t o assess the role of PET scanning in other gastrointestinal cancers. Copyright 9 1996 by W.B. Saunders Company

MPRESSIVE ADVANCES in morphological imaging have improved imaging of gastroIintestinal tumors using ultrasound, computed

PHYSIOLOGICAL BASIS OF PET IMAGING

oncological imaging has only recently come under investigation and, therefore, the list of PET agents applied to human cancer studies has been limited. However, there is likely to be increasing interest and enthusiasm for further investigation of PET oncology. A list of PET tracers currently applied in the study of cancer is given in Table 1. Oncology PET studies were first applied by Di Chiro et al 5 in the biochemical characterization of brain tumors. They showed that glucose use quantitatively measured by FDG-PET correlated with the histological grade of astrocytomas. They also showed that FDG-PET imaging was useful in differentiating areas of tumor recurrence from the areas of treatment-induced necrotic change? The biochemical basis of increased glycolysis in malignant cells that permits FDG-PET imaging of neoplasms is not fully understood. Investigators have shown that the neoplastic transformation of cell lines occurs in conjunction with increased membrane glucose transport. Additionally, investigators have found that the activities of rate-controlling enzymes for glycolysis, ie, hexokinase, phosphofructokinase, and pyruvate dehydrogenase, are increased in the tumor cells. 7 FDG uptake is a marker of the number of viable cancer ceils. While work continues in

PET imaging uses radiopharmaceuticals labeled with positron emitting isotopes. The radiopharmaceutical base determines the physiological or biochemical process to be quantitated. In the case of FDG, glycolysis is quantitated. 4 Several hundred positron emitting radiopharmaeeuticals have been developed. Unlike PET imaging of the brain or heart, the use of PET for

From the PET Center, West Virginia University, School of Medicine, Morgantown, Address reprint requests to N. Gupta, MD, West Virginia University Hospital, PO Box 8062, Medical Center Drive, Morgantown, WI/26506. Copyright 9 1996 by W.B. Saunders Company 0001-2998/96/2601-000855.00/0

tomography (CT), and magnetic resonance imaging (MRI). 1 However, the most intriguing development appears to be in the area of functional metabolic imaging as represented by positron emission tomography (PET). PET, which began primarily as a research tool for the biochemical and physiological study of the brain and heart. PET has been established as a clinical modality for selected disorders of the heart and brain. However, an emerging application of PET that is of potential significance for research, as well as, clinical practice, is in the field of oncology. Warberg 2 observed that the malignant transformation of cells is associated with a high glycolytic rate. PET-FDG uses a glucose analog 2-pSF]-fluoro-2-deoxy-D-glucose. This agent has been widely evaluated as an imaging agent for use as a quantitative tracer of tissue glucose metabolism and indeed is by far the most commonly used positron emitting radiopharmaceutical for cancer studies. The enhanced glycolysis in malignant cells allows for the detection of even small foci by PET-FDG. 3

Seminars in Nuc/ear Medicine, Vol XXVI, No 1 (January), 1996: pp 65-73

65

66

GUPTA AND BRADFIELD

Table 1. Radiopharmaceuticals Used for PET Oncology Studies Radiopharmaceutical

Process

Glucose metabolism (glycolysis) 13Nammonia (13NHa) 62Cu--PTSM Oxygen metabolism 1sO oxygen (1502) Amino acid uptake, protein syn11Cmethionine thesis 1~C leucine 13Nglutamate Nucleic acid metabolism (DNA rep- ~1Cthymidine lication) 11C putrescine Polyamine metabolism Organ-specific or specialized processes lSF fluoroestradiol Receptor specific ligands lSF fluoromisonidazole Hypoxic cell agents leF UdR Chemotherapeutic agents lSF tamoxifen leF-5-fluorouracil (lSF-FU) 2-[laF] fluoro-2-deoxy-D-glucose (FDG) Perfusion

these areas, the ability of PET to measure the kinetics of biochemical processes is fueling a revolution in the investigation of tumor biology. IMAGING TECHNIQUE FOR PET-FDG IMAGING

Patients or normal subjects are administered an intravenous bolus injection of 10.0 mCi of F-18 labeled FDG. It is recommended to prepare the subjects with at least 4 to 6 hours of fasting before injection to minimize the competitive effect of a high arterial glucose level and to suppress myocardial glucose use. A 60-minute wait is recommended after FDG injection to ensure maximal uptake in the tumor and to minimize liver uptake in the normal hepatocytes. Because of a relative higher phosphatase enzyme content in the liver, the time-activity curve usually shows washout of normal activity from the liver up to 45 to 60 minutes. PET scanners have evolved with improved resolution and sensitivity over the years. New modern scanners have more detector rings and larger fields-of-view. The whole-body or conventional tomographs have been developed by Siemens Gamma--Sonics or General Electric Co. One such typical system is the 931-08 Siemens/CTI scanner with eight rings. It produces 15 simultaneous slices at each field-ofview (in this case 12 cm). Usually two or three bed positions are required to include the entire abdomen and pelvis in the field-of-view. Initially, transmission scanning is performed be-

fore FDG injection using the external radioactive source for attenuation correction. If transmission scanning is omitted in whole-body imaging, the quantitative indices of FDG uptake cannot be computed. In this case the image analysis will be limited to visual analysis. In several instances it is possible to complete the transmission scan after the total body scan. Image reconstruction is performed using conventional filtered back projection. Varying filters with different cut-off values are usually available. The two-dimensional projection image slices are generated in the transaxial, sagittal, and coronal views. Often, dynamic transaxial imaging is performed for the first 60 minutes after FDG injection before emission imaging. This can generate temporal FDG kinetics in tumors. To minimize the bladder uptake in pelvic scanning, urinary catheterization is suggested. This is particularly important to enhance the detection of small pelvic abnormalities adjacent to the urinary bladder. The continuous bladder irrigation is preferred after placement of an indwelling catheter. Recent studies have developed external marker methods to result in image merging. This helps in topographic localization of hypermetabolic metastases, which makes them easier to find on surgery. PET EVALUATION OF COLORECTAL TUMORS

Detection of Tumor Recurrence Colorectal carcinoma is the second most common cause of cancer deaths in the United States. 8 There are approximately 110,000 new colon cancer cases, and 45,000 rectal cancer cases reported every year. About 70% of the patients are resectable with curative intent. However, only two-thirds of these patients are cured by resection. 9 Recurrence of the cancer is noted in the remaining one-third of these patients in the first 2 years after resection. Of all the patients with recurrent cancer, only about 25% are curable. It is estimated that of those patients currently being resected with apparent isolated regional disease, only 25% to 50% will be cured. CT is a bench-mark staging test with an accuracy of 25% to 73%. 1~The sensitivity of CT scanning may be satisfactory for hepatic metastases but is suboptimal for nodal involve-

PET iN GI NEOPLASMS

ment. Of those patients with negative CT, 50% will manifest nonresectable lesions at the time of exploratory laparotomy.

Comparison of PET with other methods Current methods for postsurgical follow-up after primary resection of colorectal cancers have proven inefficient for the early detection of recurrent tumor. Recent studies have shown that CEA monitoring is only 59% sensitive and 84% specific in detecting colorectal cancer recurrence, n Barium studies have also been used for detection of local recurrence with an accuracy estimated to be in the range of 80%. Barium enema studies are, however, only 49% sensitive and 85% specific for overall recurrence detection. Other techniques, such as radioimmunoscintigraphy, have been used in the localization of a questionable lesion but accuracy is variable, especially in hepatic or peri-hepatic lesions, in postsurgical evaluations, or after chemotherapy. 12 Strauss et al a3 evaluated the role of PET imaging in differentiating recurrent colorectal tumor from scar in 29 patients. Twenty-one patients with proven recurrent malignancy showed abnormalities on the PET-FDG scans. Similarly, all nonmalignant lesions showed very low or absent FDG accumulation. Strauss et al also reported comparable results with PET imaging using F-18 fluorodeoxygiucose and MRI in differentiating recurrent tumor from scar in 15 patients. In that group, PET imaging was

Fig 1. Transaxiel views from CT (A) and PET-FDG study (B: see color plate) showing a presacral soft tissue mass that proved to be recurrent cancer in s 60year-old man status post resection of rectal cancer 21h years ago.

67

superior to MRI and in one patient MRI failed to detect proven recurrence. We prospectively compared PET imaging using FDG with the conventional diagnostic standard of CT imaging in 24 suspected cases of colorectal cancer after prior surgical resection. a4PET imaging was performed 1 hour after injection of 10 mCi of F-18 FDG. PET and CT imaging results were reviewed in double-blind fashion by different reviewers and compared with the final diagnosis established by tissue biopsy in all patients. PET imaging accurately detected recurrence in all 17 patients proven to have colorectal recurrence (sensitivity 100%) compared with CT with a sensitivity of 70% (Fig 1). In four patients, CT findings that were interpreted as showing tumor recurrence were proven to be benign lesions on histological examination. These four cases with falsepositive CT findings were observed after an abdominal pelvic resection, 2 suspected softtissue mass in the presacral region, 1 and a suspected lesion of the liver.1 Thus, the specificity of CT was significantly lower (42.8%) as compared with the specificity of 85% with PET. The only false-positive PET finding was the increased FDG uptake noted in a patient with inflammatory disease. PET-FDG imaging showed significant higher predictive accuracy than CT scanning (94% vs 62%) (Fig 2). Similar results are also reported by Strauss et al. 15 In the multicenter retrospective study coordinated

68

GUPTA AND BRADFIELD

Fig 1B

Fig 5. PET-FDG study showing increased FDG uptake in a locally recurrent rectal carcinoma on this sagittal view.

Fig 2A

through the Institute for Clinical PET (ICP), the sensitivity and specificity of PET-FDG imaging for detecting recurrent colorectal tumor was 93% and 78% respectively.

Semiquantitative Indices One attractive way to increase specificity and accuracy of PET studies may be to quantitate

numerical values of glucose metabolism on PET-FDG study. 16 Semiquantitative differential uptake ratios (DUR) have an advantage because of their simplicity of methodology. D U R indices are calculated based on the net tumor uptake where FDG uptake is normalized to patients body weight and the dose of the tracer injected. However, for comparison sake

PET IN GI NEOPLASMS

69

may contribute to this uptake that is noted in a minority of patients. However, DUR of this bowel uptake usually tends to be less than 3 to 4. DUR values corrected for lean body mass or body surface area are more reliable than those computed using body weight. However, occasionally values as high as 5 or 10 have been reported in the stomach, liver, or gut. ]7

PET Versus CarcinoembroyonicAntigen Monitoring

Fig 2. (A: see color plate) PET-FDG study (transverse views) in this 75-year-old woman who had left hemicolectomy for slgmoid cancer 1 year ago showing recurrent tumor In the bowel and mesentery not clearly noted on (B) CT scan but confirmed on needle biopsy.

the images in different patients should be obtained at the same time after the injection. In PET studies on colorectal cancer, a DUR cutoffvalue of 3.0 may help to improve the specificity of detecting malignant disease. In our study we found the mean DUR indices to be significantly different in malignant and benign lesions 14(4.86 __. 1.82 vs 1.46 _ 0.83, respectively) (Fig 3). It is not uncommon to visualize FDG uptake in normal bowel. Occasionally, this can be especially prominent in the region of the cecum or stomach. It is uncertain what factors

DUR

We also found PET-FDG imaging to be a more sensitive method than the measurement of plasma carcinoembryonic antigen (CEA) levels for detecting tumor recurrence as also indicated by Haberkorn et al. ~8 In our study, six patients with proven colon recurrence failed to show increased CEA levels. Similarly, four patients with no histological evidence of colorectal cancer showed increased high CEA levels. In recent studies, the role of routine CEA monitoring in detecting recurrent colorectal cancer has been questioned because of its limited sensitivity (59%) and specificity (84%). H PET may be a more sensitive and specific technique for monitoring of patients with suspected colorectal cancer than CEA (Table 2).

Clinical Role of PET-FDG Imaging In a study comparing PET with CT for detecting abdominopelvic recurrent tumor sites, patients with primary malignancies of colon, ovary,

6 5 4

Fig 3. DUR indices in recurrent colorsctal cancer sites compared with DUR indices in patients with benign lesions (such as scar) with histological confirmarion in 24 patients with suspected cancer.

3

1 [

NG

Benign(N= 7)

Malignant (N = 17)

70

GUPTA AND BRADFIELD

uterus, and kidney were observed. 19 A total of 63 tumor sites were found on histological confir. mation. PET-FDG imaging accurately showed the increased FDG uptake in 44 of 47 positive sites (Fig 5) with a sensitivity of 94% and specificity Of 88%. However, CT missed 18 tumor sites and results were false positive at 7 sites (sensitivity of 62% and specifcity of 56%). These and similar results from the other investigators, confirm the superiority of PET-FDG imaging in the localization of recurrent abdominopelvic tumor foci regardless of the origin of the primary tumor. We have also studied the role of PET-FDG imaging in the preoperative staging of colorectal cancer. 2~ We found the sensitivity, specificity, and accuracy of PET scanning to be 90%, 66%, and 87% as compared with CT scanning (60%, 100%, and 65%, respectively). In this study of 16 patients. 20 tumor sites were found (lymph nodes, 5: colorectal, 13; liver. 2). The low specificity of PET-FDG imaging in this limited study was probably because of the very small number (two) of negative sites biopsied. In another comparative study using PETFDG imaging and CT scanning for the detection of hepatic metastatic disease, we found PET-FDG to be significantly more sensitive than CT (sensitivity of 92% vs 42% with CT). 21 PET-FDG imaging detected hepatic metastatic disease in 33% of these patients, not found on CT scanning. This is of significant clinical importance because detecting hepatic metastases in a patient with the localized recurrent disease may render the patient inoperable. Similarly a solitary hepatic lesion may be resectable in the absence of extrahepatic disease (Fig 4). However unsuspected extrahepatic metastases are often found on PET-FDG scanning obviating hepatic surgery. Thus. PET-FDG imaging can be a useful screening tool before curative resection. Table 2. Proven Clinical Applications of PET-FDG

Screening for recurrent disease 1' CEA, negative CT Mass on CT Confirming nonrecurrence? Staging of recurrent disease Hepatic, extrahepatic lesions Operability--extent Staging before initial surgery?

Fig 4. Recurrent hepatic metastatic tumor involving the right lobe of liver on PET-FDG study (A). CT scan (B) showed a focal area of translucency corresponding to tumor site.

Okazumi et a122 were able to differentiate benign from malignant tumors involving the liver in 35 patients including 23 with primary liver cancer and 10 with metastatic cancer (from the large bowel, esophagus, stomach, or pancreas) and 2 with hemangiomas. They also found that the degree of differentiation of hepatocellular carcinoma correlated with K3 and I~ rate constant values as indices of glucose metabolism. In the present day environment of managed care, PET-FDG scanning can be a cost-effective tool in selecting the most appropriate patients for expensive surgical or other invasive procedures such as biopsies, second look laparotomy, or exploratory laparotomy. To evaluate the patient for extensiVe metastatic disease, totalbody PET-FDG scanning has been successfully applied. Valk et al, 23 in their study of 20 patients, found 100% accuracy with PET as compared with 52% with CT in the detection of

PET IN GI NEOPLASMS

metastatic lesions. Sixty-six percent of proven metastatic sites were noted in the liver or lung whereas 30% of all sites occurred in the liver. Thus, it is postulated that whole-body PETFDG may alter therapy in 40% to 50% of patients. Patients with increased CEA but negative or equivocal CT findings are ideal candidates to benefit from whole-body PET scanning. Treatment Monitoring Haberkorn et al ~8 also found that PET with FDG can detect the changes in FDG uptake in colorectai tumors using radiation therapy. Comparing FDG uptake with CEA measurements, they found the measurement of FDG uptake to be more sensitive to assess tumor therapy response. In their study, CEA was in normal range in 14 of 41 examinations before and after radiotherapy. However, caution should be observed in very early measurement of change in FDG uptake after radiation therapy because of the inclusion of inflammatory cells. A widely used chemotherapeutic agent in the treatment of hepatic metastases is 5-Fluorouracil (5-FU). The metabolism of FU has been studied extensively using PET scanning. Tissue distribution studies analyzing the uptake of ~SF-labeled pyramidines was reported by Abe et al. 24 Shani and Wolf also reported that drug responsive tumors have higher concentration ratios of ~8F-labeled 5-FU than drug-resistant tumors. 25 Similar studies on pharmacokinetics of ~8F-labeled 5-FU have been reported recently. 26 The estimation of 5-FU chemotherapy outcome before therapy is one of the unsolved challenges. Strauss et al ~5examined FU metabolite concentration from PET images before chemotherapy and compared those with tumor growth rate during chemotherapy. They studied uptake in 18 metastases followed by intravenous administration of 5-FU (500 to 100 mg/m2/24 hr) for 5 days followed by a 3-week interval without chemotherapy. The volumes of metastases were measured from CT scan. In lesions with no or low 5-FU uptake in metastases before chemotherapy, there was significantly increased tumor volume but no response to therapy was noted. Only in patients with high F-18 concentration in tumor (SUV > 3.5), there was a high probability of response. There was

71

high correlation between 5-FU accumulation and tumor growth rate (r = 0.86). Further investigations of using PET with radiolabeled antibodies are being worked out using Iodine-124 and 18F positron emitters conjugated to the antibodies. The development of antibody-based cancer targeting agents for radioimmunotherapy has caused considerable interest. PET, in addition to radioimmuno-localization, may provide more accurate radiation dosimetry for radioimmunotherapy. 27 PET using C-11-5HTP has been shown to be an excellent tool for visualizing carcinoid tumors of midgut origins, even metastases in the liver, z8 It could also be used to monitor treatment of carcinoid tumors. PANCREATIC TUMORS

Pancreatic Carcinoma Pancreatic carcinoma is among the gastrointestinal tumors with the worst prognosis. Intended diagnostic procedures such as ultrasound, CT, endoscopic retrograde cholangiopancreatography (ERCP), and more recently MRI have been used for the diagnosis of the pancreatic cancer or ductal abnormalities suggesting pancreatic cancer. There is a wide variation in the sensitivity of CT for diagnosing pancreatic cancer reported in recent years (58% to 95%). z9 Similarly, the distinction between pancreatic carcinoma and benign pancreatic tumor or chronic pancreatitis is difficult to make using the current diagnostic modalities. 3~A new modality of endoscopic ultrasound is superior in the detection of tumors smaller than 5 mm in diameter. 31 However, endoscopic ultrasound does not seem to be well suited for differentiating inflammatory lesions from malignant tumors. Chronic Pancreatitis Versus Pancreatic Cancer However, several recent studies have highlighted the clinical potential and superiority of FDG-PET for the detection and differentiation of pancreatic carcinoma. Inokuma et al, 32 recently found PET-FDG to have high sensitivity (94%), specificity (82%), and accuracy (91%) as compared with CT scanning. There was an improved specificity and differentiation of benign pancreatic lesions from malignant disease using PET scanning. Although increased FDG

72

GUPTA AND BRADFIELD

uptake could be observed occasionally with inflammatory changes, f u r t h e r studies are warranted to confirm these results, However, these investigators reported significant differences in the SUV values in pancreatic cancer (4.05 _ 0.89), and chronic pancreatitis (1.47 --- 0.69). Pancreatic cancer, in a study by Stollfus et a133 showed discrete intense focus of F D G activity. In this study, SUV were greater than 3.0, In 60 of 73 patients with suspected cancer or pancreatitis, it was found on visual inspection of F D G - P E T images that carcinoma could be correctly identified in 95 % of patients; chronic pancreatitis could be identified in 90% of the patients. However, caution must be used in t h e generalization of the results because other investigators using CT alone have shown 90% accuracy in the diagnosis of pancreatic cancer. In this regard, P E T may be reserved for patients with inconclusive or equivocal findings on CT. An important limiting factor on PETF D G studies in patients with pancreatic disease occurs with glucose intolerance. This could provide a possible explanation for false-negative findings on P E T - F D G imaging in pancreatic cancer caused by the competitive inhibition of F D G transport with plasma glucose. There is also evidence that P E T - F D G may be useful in lymph node staging o f pancreatic cancer and, therefore, may enhance detection of lymph node disease as compared with conventional methods. CT scanning may not be as specific as PET, because enlarged lymph nodes can be caused by inflammation or metastases. Endocrine Pancreatic Tumors

Ahlstr6m et a134 reported P E T to be a valuable complementary technique in the detection of endocrine pancreatic tumors using

C-11-1abeled dihydroxyphenylalanine (C-11-LD O P A ) and C-11 labeled hydroxytryptophan (C-11-5-HTP).34 In Ahlstr6m et al's study of 22 patients with a wide variety of tumors such as gastrinoma and glucogonoma, P E T methods were more sensitive in the detection of functional endocrine pancreatic tumors and less sensitive in the detection of nonfunctional tumors. 5-HTP was better taken up by lesions than L-DOPA. Comparison between this technique and the P E T - F D G imaging is warranted in the future to evaluate the relative use of each technique. The study failed to show any general advantage of P E T compared with CT because all P E T positive lesions were also observed on CT. However, in case of misdiagnosis or equivocal cases, high uptake of the tracer noted on P E T may confirm the presence of a malignant lesion. Therefore. P E T can compliment CT in demonstration of hormona!ly functional tumors. in particular glucogonomas in selected patients. Eventually, tracers other than LD O P A or 5-HTP may prove more useful. CONCLUSION

Noninvasive PET scanning using various radioligands is a useful diagnostic tool to detect recurrent tumor sites of various gastrointestinal cancers. P E T scanning using F-18 F D G has proven to be useful in detecting recurrent colorectal and hepatic tumor sites in patients with suspected recurrence. P E T - F D G imaging can be useful in selecting appropriate patients with recurrent abdominopelvic tumor for surgical procedures with a curative intent. P E T - F D G may also be useful in differentiating chronic pancreatitis from pancreatic cancer. Further research using new P E T radiotracers may further expand the clinical role of P E T imaging.

REFERENCES

1. Trenkner SW, ThompsonWM: Imagingof gastrointestinal malignancies. Curr Opin Oncol 4:736-740,1992 2. Warburg O: On the origin of cancer cells. Science 123:309-314, 1956 3. Phelps ME, Mazziotta JC, Schelbert HR, et al: Positron emission tomography and autoradiography.Principles and Applications for the Brain and Heart. New York, NY, Raven, 1986 4. Som P, Atkins HL, Bandoypadhyay D, et al: A fluorinated glucose analog, 2-fiuoro-2-deoxy-D-glucose-

(18F): Nontoxic tracer for rapid tumor detection. J Nucl Med 21:670-675, 1980 5. Di Chiro G, De La Paz RL, Brooks RA, et al: Glucose utilization of cerebral gliomas measured by [18F]flUorodeoxy-glucose and positron emission tomography. Neurology32:1323-1329,1982 6. Di Chiro G: Positron emission tomography using [lSF]flu0ro-deoxyglucosein brain tumors. Invest Radiol 22:360-371, 1987 7. Weber G: Enzymologyof cancer cells. N Eng! J Med 296:486-493,541-551, 1977

PET IN GI NEOPLASMS

8. Silverberg E, Lubera J: Cancer statistics 1986. CA 36:9-25, 1986 9. August DA, Ottow RT, Sugarbaker PH: Clinical perspectives on human colorectal cancer metastases. Cancer Metastasis Rev 3:303-324, 1984 10. Sugarbaker PH, Grianola FJ, Dwyer S, et al: A simplified plan for follow up of patients with colon and rectal cancer supported by prospective studies of laboratory and radiologic test results. Surgery 102:79-87, 1987 11. Moertel CG, Fleming TR, MacDonald JS, et al: An evaluation of the carcinoembryonic antigen (CEA) test for monitoring patients with resected colon cancer. J A M A 270:943-947, 1993 12. Schlag P, Uhner B, Strauss LG, et al: Scar or recurrent rectal cancer: Positron emission tomography is more helpful for diagnosis than immunoscintigraphy. Arch Surg 324:197-200, 1989 13. Strauss LG, Clorius JH, Schlag P, et al: Recurrence of colorectal tumors--PET evaluation. Radiology 17a:329332, 1989 14. Gupta NC, Bowman BM, Frank AI, et al: PET-FDG imaging for follow-up evaluation of treated colorectal cancer. Radiology 199:181P, 1991 (abstr) 15. Strauss LG, Conti PS: The applications of PET in clinical ontology. J of Nucl Med 32:623-648, 1991 16. Hawkins RA: Pancreatic tumore: lmagingwith PET t. Radiology 195:320-322, 1995 17. Delaloye AB, Wahl RL: How high a level of FDG abdominal activity is considered normal? J of Nucl Med 36:106, 1995 (abstr) 18. Haberkorn U, Strauss LG, Dimitrakopoulou A, et al: PET studies of fluorodeoxyglucose metabolism in patients with recurrent colorectal tumors receiving radiotherapy. J Nucl Med 32:1485-1490, 1991 19. Gupta NC, Frank AI, Chandramouli B, et al: Clinical role of PET-FDG imaging in evaluation of patients with abdominopelvic malignancies. J of Nucl Med 34:55p, 1993 (abstr) 20. Gupta NC, Falk PM, Frank AL, et al: Pre-operative staging of colorectal carcinoma using positron emission tomography. Nebraska Medical Journal 26:30-35, 1993 21. Gupta N, Frank A, Mailiard J, et al: Accurate detection of liver metastases in patients (PTS) with primary malignancies using PET-FDG imaging. J Nucl Med 34:6P, 1993 22. Okazumi S, Isorio K, Enomoto K, et al: Evaluation of liver tumors using fluorine-18-fluorodeoxyglucose PET:

73

Characterization of tumor and assessment of effect of treatment. J Nucl Med 33:333-339, 1992 23. Pounds TR, Valk PE, Haseman MK, et al: Whole body PET-FDG imaging in diagnosis of recurrent colorectal cancer. J Nucl Med 36:57P, 1995 24. Abe Y, Matsuzawa T, Fujiwara T, et al: Assessment of radio-therapeutic effects on experimental tumors using 18F-2-fluoro-2-deoxy-D-glucose. Eur J Nucl Med 12:325328, 1986 25. Shani J, Wolf W: A model for prediction of chemotherapy response to 5-fluorouracii based on differential distribution of 5-[18F] fluorouracil in sensitive versus resistant lymphocytic leukemia in mice. Cancer Res 37:23062308, 1977 26. Wile AG, Stemmer EA, Andrews PA, et al: Pharmacokinetics of 5-fluoracil during hyperthermic pelvic isolation perfusion. J Clin Oncol 3:849-852, 1985 27. Westera G, Reist HW, Buchegger F, et al: Radioimmuno-positron emission tomography with monoclonal antibodies: A new approach to quantifying in vivo tumor concentration and biodistribution for radioimmunotherapy. Nucl Med Commun 12:429-437, 1991 28. Eriksson B, Lilja A, Ahlstrom H, et al: Positron emission tomography as a radiological technique in neuroendocrine gastrointestinal tumors. Ann NY Acad Sci 446-456, 1991 29. Zeiss J, Coombs RJ, Bielke D: CT presentation and staging accuracy of pancreatic adenocarcinoma. Int J Pancreatol 7:49-53, 1990 30. Mackie CR, Cooper MJ, Lewis MJ, et al: Nonoperative differentiation between pancreatic cancer and chronic pancreatitis. Ann Surg 189:480, 1979 31. Palazzo L, Roseau G, Gayet B, et al: Endoscopic ultrasonography in the diagnosis and staging of pancreatic adenocarcinoma. Endoscopy 25:143-150, 1993 32. lnokuma T, Tamaki N, Torizuka T, et al: Evaluation of pancreatic tumors with positron emission tomography and F-18 fluorodeoxyglucose: Comparison with CT and US. Radiology 195:345-352, 1995 33. Stollfus JC, Glatting G, Friess H, et al: 2-(fluorine-18)fluoro-2-deoxy-D-glucose PET in detection of pancreatic cancer: Value of quantitative image interpretation. Radiology 195:339-344, 1995 34. Alstr6m H, Eriksson B, Bergstr6m M, et al: Pancreatic neuroendocrine tumors: Diagnosis with PET. Radiology 195;2:333-337, 1995