The Visualization S.W.J.

of Gastroenteropancreatic Lamberts,

Endocrine

Tumors

J.-A. Chayvialle, and E.P. Krenning

In this review, we evaluate radiological techniques currently used to localize gastroenteropancreatic (GEP) endocrine tumors. We also describe the visualization, using intravenous (IV) administration of two isotope-labeled somatostatin analogues (‘231-Ty& octreotide and “‘In-DTPA-octreotide) of islet-cell tumors in 25 patients and carcinoids in 39 patients. The primary tumor and previously unrecognized distant metastases were visualized in 20 of the 25 patients (80%) and in 37 of the 39 patients (95%). Parallel in vitro detection of somatostatin receptors on those tumors also visualized in vivo showed that ligand binding to the tumor in viva represents binding to specific somatostatin receptors. The detection of somatostatin receptors on tumors in vivo predicted a good suppressive effect of octreotide on hormonal hypersecretion by these tumors. It is an easy, quick, and harmless procedure that is valuable in the localization of primary endocrine pancreatic tumors and their often radiologically and clinically unrecognized metastases. Future prospective controlled studies comparing this procedure with other radiological investigative techniques should demonstrate its sensitivity and specificity and determine the place of somatostatin receptor imaging in the localization of GEP endocrine tumors. Copyright C1 1992 by W.B. Saunders Company

G

ASTROENTEROPANCREATIC (GEP) endocrine tumors include islet-cell tumors such as gastrinomas, vasoactive intestinal polypeptide tumors (VIPomas), glucagonomas, insulinomas, and carcinoids. Many of these tumors grow slowly and remain undetected for a long period of time.le4 On detection, the proportion of malignant tumors varies according to the origin of the neoplastic cells, and ranges from approximately 10% for insulinomas5 to 60% for gastrinomas and nonsecreting islet-cell tumors4 With carcinoid tumors, malignancy largely depends on the size of the primary tumor, while metastatic spread is almost the rule in patients with the carcinoid syndrome. Although the various syndromes associated with hormonal hypersecretion provide clues as to the presence of a GEP tumor, delays of years frequently occur between the appearance of the first symptoms and the actual diagnosis. The localization of these tumors often causes major problems; 55% to 70% of insulinomas are smaller than 2 cm in diameter at diagnosis5 and more than 30% of gastrinomas are smaller than 1 cm. VIPomas are mostly of pancreatic origin but can also arise from nervous structures outside the pancreas, eg, ganglioneuromas, and up to 30% of gastrinomas are located in the duodenal wall, as are some forms of somatostatinomas. An additional challenge is multiple endocrine neoplasia type I (MEN-K), as multiple endocrine tumors are found within the pancreas in most cases.6 In this review, we analyze the literature with regard to the value of different radiological techniques in the localization of GEP endocrine tumors, and also describe the preliminary results of somatostatin receptor scintigraphy.

puted tomography (CT)-guided fine-needle biopsy for cytology and/or histology finalizes the diagnosis in most cases. The sensitivity and specificity of this procedure is close to 80%, thus providing a valuable clue to the endocrine nature of such tumors.8 Immunohistochemistry using antisera to hormones or endocrine markers such as chromogranin A often help to ascertain the ultimate diagnosis.’ RADIOLOGICAL PROCEDURES IN THE PATIENT WITH HORMONAL HYPERSECRETION

The biggest challenge is presented by the patient with an overt clinical syndrome related to hormonal hypersecretion, in whom ultrasound does not show the presence of metastatic spread. The localization of the primary tumor in these patients is crucially important, as surgical therapy is universally considered to be the first (curative) choice. A great number of radiological techniques have been studied in prospective and retrospective studies in different forms of GEP endocrine tumors. The standard with which each of these techniques has to be evaluated is the finding at surgery and, in exceptional cases, at autopsy. However, this is not an absolute standard;” surgery may fail to localize very small tumors, even in cases of indisputable hormonal hypersecretion, and islet-cell hyperplasia may, in exceptional cases, be the pathophysiological basis of the hormonal syndrome. Therefore, the use of seven radiological techniques is here evaluated. Ultrasound

The detection rate ofsolitary insulinomas varies from 25% to 60%““‘; however, this rate decreases to 15% in the case of multiple tumorsJ5 Gastrinomas are frequently missed at ultrasound investigation, with approximately only 25% being

LARGE TUMOR MASS

Ultrasonography of the abdomen is the established firstline procedure to evaluate patients presenting with gastrointestinal symptoms. Consequently, a considerable number of GEP endocrine tumors are readily detected by the demonstration of liver metastases or enlarged paraaortal lymph nodes. This applies especially to patients with nonsecreting islet-cell carcinomas’ and metastic carcinoids,’ and to a fair percentage of patients with gastrinomas, VIPomas, and glucagonomas. In principle, percutaneous ultrasound- or comMetabolism, Vol 41, No 9. Suppl 2 (September),

1992: pp 11 l-1 15

From the Departments ofMedicine and Nuclear Medicine, Erasmus University, Rotterdam, The Netherlands; and the Digestive Specialties Department, H&pita1E. Merriot, Lyon, France. This paper (with virtually the same content) will also publish in a supplement to Digestive Diseases and Sciences (in press). Address reprint requests to S. W.J. Lamberts, MD, Professor of Medicine, Department of Medicine, University Hospital Dijkzigt, 40 Dr. Molewaterplein, 3015GD Rotterdam, The Netherlands. Copyright 0 1992 by W.B. Saunders Company 00260495/92/4109-2001$03.00/O 111

LAMBERTS,

112

detected.‘* Ultrasound detection of tumors smaller than 2 cm is rare, and 7 mm in diameter is probably the absolute threshold. CT

CT overcomes a number of practical limitations of ultrasound, since especially the pancreas and its tail can be visualized despite intestinal gas. Except for readily visualized large tumors, the detection of GEP endocrine tumors is mainly based on their hypervascularization.” Results of studies indicate that 30% to 70% of small solitary insulinomas and gastrinomas are detected at CT’5,‘9-2’;these figures are lower in patients with the MEN-I syndrome.15 Although most investigators used 4 mm sections, CT rarely detects tumors with a diameter less than 2 cm unless a dramatic blush occurs after injection of contrast medium”; intraarterial injection of contrast medium is probably not required.** Arteriography

The need for a meticulous procedure with arteriography is even greater than that for ultrasound and CT. Although hypervascularization differentiates endocrine tumors from most other cancers, the capillary blush can be very subtle. Hyperselective injections into the pancreatic arteries are in most instances necessary, using both frontal and oblique views, and digital subtraction should be used whenever available. Sensitivity varies from 53% to 90% in solitary insulinomas’0,‘5~16 and from 13% to 88% in gastrinomas.‘8.23 with the highest success rates being reached in small series. An average rate of 50% to 60% seems to be the real figure, but sensitivity decreases drastically in patients with multiple tumors. Transhepatic Portal Venous Sampling

Transhepatic portal venous sampling (TPVS) is invasive, involving cannulation of peripancreatic veins via the right hepatic vein. Limitations to the interpretation of results are (1) poor precision of the threshold gradient above which local hormonal hypersecretion is established, (2) laminar flow in large veins ofthe portal system, and (3) variation in hormone release during the sampling period. In experienced hands, TPVS has been reported to be fairly efficient, with sensitivity in insulinoma patients ranging from 46% to 100%.24-26However, false-positive results have been reported in 5% to 33% of patients,*‘.*’ and the interpretation of data is very difficult in patients with MEN-I and in the case ofdiffuse hyperplasia, ie, nesidioblastosis. The localization of gastrinomas can be facilitated by the local intraarterial injection of small boluses of secretin.

AND KRENNING

Intraoperative Ultrasound

Intraoperative ultrasound (IOUS) represents a major advancement in the localization of GEP endocrine tumors. In several large series of insulinoma patients, sensitivity was close to 80%,“.‘5.‘6 and it reached almost 100% if IOUS and palpation of the pancreas were combined; however, IOUS is superior to palpation.‘6 Nevertheless, multiple tumors were recognized in approximately only 50% of cases.'5 Sensitivity for pancreatic gastrinoma detection is high (85%), but extrapancreatic gastrinomas still seem to be best detected by palpation. ’ ’ Endoscopic Ultrasound (Echoendoscopy)

Preliminary results on the use of echoendoscopy in the diagnosis of islet-cell tumors are promising.3’ In a series of patients with solitary tumors undetected by ultrasound and CT, 80% were visualized by echoendoscopy. with the tumors having a mean diameter of only 1.3 cm.33 Echoendoscopy allows detailed exploration of the tail of the pancreas and the periampullar area. SOMATOSTATIN

RECEPTOR

SCINTIGRAPHY

Somatostatin receptors have been shown to be present on a variety of tumors arising in tissues that normally contain them. Large numbers of high-affinity receptors have been found on most metastatic islet-cell tumors and carcinoids.34.35 Long-term octreotide therapy controls hormonal hypersecretion from endocrine pancreatic tumors and carcinoids in most patients, and clinical symptomatology greatly improves.36 There is also evidence for control of tumor growth in some patients during somatostatin analogue treatment.36 These results lead to an improvement in the quality of life for these patients and make the clinical introduction of octreotide a major breakthrough in the treatment of endocrine cancers.37 The presence of large numbers of high-affinity somatostatin receptors on many of these tumors led us to explore whether it is possible to detect receptor-positive tumors in vivo using a radioactive iodine-labeled analogue.38 Tyr3-octreotide is a somatostatin analogue with biological activities similar to octreotide. We coupled Tyr3-octreotide to 123Iand injected 37 to 555 MBq ‘231-Tyr3-octreotide intravenously in patients suspected of having somatostatin receptor-positive tumors, while planar or emission-computed tomographic images were made with a gamma camera. Following the bolus injection Table 1.

Detection

of GEP Endocrine Tumors by Somatostatin Receptor Imaging

Tumor Type

Nuclear Magnetic Resonance Imaging Islet-cell

The value of nuclear magnetic resonance imaging (NMR) is at present unclear. In a prospective study of patients with gastrinomas, specificity was lOO%, but sensitivity was only 20%. No tumors smaller than 1 cm in diameter were detected, and only 50% of those larger than 3 cm were recognized.30 Hence, NMR may not be very successful in the visualization of islet-cell tumors.”

CHAYVIALLE,

No. of Tumors Vlsualtzed

tumors

VlPomas Gastrlnomas Glucagonomas

l/l IO/l 1 (91%)

lnsulinomas

l/J 5/8

“Nonfunctioning” Islet-cell tumors

212

Somatostatinomas Carclnolds

112 37/39

(63%)

(95%)

SOMATOSTC\TlN

RECEPTOR SCINTIGRAPHY

Fig 1. “‘In-DTPA-octreotide scan in a 36-year-old man with organic hyperinsulinism. Ultrasound, CT, and highly selective arteriography had not localized the insulinoma. (A) Frontal gamma-camera photograph of the abdomen taken 24 hours after isotope administration showed a normal liver, two normal kidneys, and a hot spot at the level of the left kidney. (B) SPECT images made through this region showed the insulinoma to be localized in the tail of the pancreas just behind the left kidney.

of radioiodinated Ty?-octreotide, rapid accumulation of radioactivity was seen in the liver. Approximately 50% of the activity was cleared from the blood pool within 2 minutes after injection, and it was possible to localize a variety of tumors and their metastases. The procedure localized the primary tumor and/or previously unknown metastases in seven of nine patients with islet-cell tumors. In five of these tumors, we could subsequently investigate the surgically removed tumor.39 There was a close relationship between the in vitro detection of somatostatin receptors using autoradiography and the gamma-camera pictures obtained after injection of Iz31-Tyr3-

113

octreotide. This indicates that ligand binding to the tumor in vivo represents binding to specific somatostatin receptors. In addition, we performed preoperative in vivo hormonal studies and in vitro experiments with cultured tumor cells. Again, there was a close relationship between the presence of somatostatin receptors and the in vivo and in vitro effects of octreotide on hormonal secretion by these tumors. This means that a positive scan predicts a beneficial effect of octreotide therapy on hormonal hypersecretion. A further argument that membrane receptors for somatostatin were visualized with this scintigraphic technique is that in vivo competition was observed between radiolabeled and unlabeled octreotide with regard to binding to somatostatin receptor-positive tumors.40 After the initial success of the technique, we were concerned about its several drawbacks: the radionuclide 123Iis not readily available in many parts of the world, its short half-life of approximately 12 hours hampers its use, very high quality specifications of the isotope are required to ensure its successful use in scintigraphy, and it is very expensive. Another problem, especially important in the visualization of isletcell tumors and abdominal carcinoids, is that ‘231-Tyr3-octreotide is excreted via the liver, gallbladder, and bile ducts into the gastrointestinal tract, and makes these parts of the body so “hot” with radioactivity that small, primary tumors in these regions can easily be missed, especially if no single photon emission computed tomography (SPECT) pictures are made. To circumvent these drawbacks, we designed an alternative peptide, diethylenetriaminepentaacetic acid (DTPA)-octreotide, which is very easily conjugated with I “In. ” 'In-DTPAoctreotide has a much longer half-life of nearly 3 days and is excreted via the kidneys.4’ Our experience in more than 250 patients with various tumors has shown that ” ‘In-DTPAoctreotide visualizes somatostatin receptor-positive tumors even better than ‘231-Tyr3-octreotide. Physiological organ accumulation of ” 'In-DTPA-octreotide is seen in the kidneys, spleen, liver, bladder, and in the thyroid and pituitary glands. Table 1 shows the results of in vivo scintigraphy in 25 patients with endocrine pancreatic tumors and 39 patients with carcinoids. In the first 40% of these patients, receptor imaging was performed with “31-Tyr3-octreotide, and in the other 60% it was performed with ” 'In-DTPA-octreotide. The technique turned out to be highly successful, and primary endocrine pancreatic tumors and their previously unrecognized metastases were visualized in 20 of the 25 patients (80%); metastases were scintigraphically evident in 11 patients. The visualization of insulinomas with this technique was more difficult, and tumors were located in only five of eight patients. In vitro autoradiography showed that all insulinomas contained receptors for somatostatin- 14 and -28, and that octreotide receptors were absent on the tumors that could not be visualized in vivo. An example of a positive scan in an insulinoma patient is shown in Fig 1. In vivo somatostatin receptor visualization of carcinoid tumors was also highly successful. Using “‘I-Tyr3-octreotide, both the primary tumor and metastatic carcinoids could be visualized in 12 of 13 patients.42 Again, the presence of SOmatostatin receptors on carcinoid tissue was predictive of a

LAMBERTS,

CHAYVIALLE,

AND KRENNING

Fig 2. “‘In-DTPA-octreotide scan in a 64-year-old man in whom a primary ileal carcinoid tumor had been removed at the age of (A) Frontal gamma-camera photograph of the abdomen after 24 hours shows an enlarged liver with a somatostatin receptor-positive spot in the left lobe of the liver. (B) In the chest (anterior), a somatostatin receptor-positive area is shown in the hilus of the right lung near the lateral basis of the right lung; additional CT images confirmed the presence of tumor tissue, while a biopsy of the nodus of the liver lobe showed the presence of carcinoid tumor tissue.

subsequent clinical and biochemical improvement during octreotide therapy.42 In a total group of 39 patients evaluated to date with both of the isotope-labeled somatostatin analogues, 37 primary tumors and/or previously unknown metastases (95%) were visualized (Table 1, Fig 2). Islet-cell tumors and carcinoids are mostly malignant tumors originating from endocrine cells throughout the body, but especially from pancreatic islet cells, intestinal mucosa, and the lung. The primary tumor and metastases are often

57. hot and left

difficult to localize with current radiological techniques.’ Somatostatin receptor scintigraphy is an easy, noninvasive, diagnostic technique for this type of tumor, and it helps considerably in localizing both the primary tumor and matastases. Future prospective controlled studies comparing this procedure with other currently used radiological investigative techniques should demonstrate its sensitivity and specificity. and help to determine the place of somatostatin receptor imaging in the diagnosis of patients with GEP endocrine tumors.

REFERENCES

1. Moertel CG: An odyssey in the land of small tumors. J Clin Oncol 5:1503-1522, 1987 2. Solcia E. Capella C. Fiocca R. et al: The endocrine system and related tumors. Gastroenterol Clin North Am 18:671-693, 1989 3. Eriksson B, Skogseid B, Lundqvist G. et al: Medical treatment and long-term survival in a prospective study of 84 patients with endocrine pancreatic tumors. Cancer 65:1883-1890, 1990 4. Kent RB, van Heerden JA, Weiland LH: Nonfunctioning islet cell tumors. Ann Surg 193:185-190, 1981 5. van Heerden JA, Edis AJ, Service FJ: The surgical aspects of insulinomas. Ann Surg 189:677-682, 1979 6. Obcrg K, Skogseid B, Eriksson B: Multiple endocrine neoplasia type 1 (MEN-l). Acta Oncol 28:383-387, 1989 7. Fugazzola C. Procacci C, Andreis IAB, et al: The contribution of ultrasonography and computed tomography in the diagnosis of nonfunctioning islet cell tumors of the pancreas. Gastrointest Radio1 15:139-144. 1990 8. Ekberg 0, Bergenfeldt M, Aspelin P, et al: Reliability of ultrasound-guided fine-needle biopsy of pancreatic masses. Acta Radio1 29:535-539, 1988

9. Wilander E: Diagnostic pathology of gastrointestinal and pancreatic neuroendocrine tumors. Acta Oncol 28:363-368, 1989 10. Rossi P, Allison DJ, Bezzi M: Endocrine tumors of the pancreas. Radio1 Clin North Am 27: 129- 161, 1989 11. Norton JA, Cromack DT. Shawker TH, et al: Intraoperative ultrasonographic localization of islet cell tumors: A prospective comparison to palpation. Ann Surg 207: 160-168. 1988 12. Dunnick NR. Long JA, Krudy AG, et al: Localizing insulinomas with combined radiographic methods. AJR 135:747-752, 1980 13. Kuhn FP. Giinther R, Ruckert K, et al: Ultrasonic demonstration of small pancreatic islet cell tumors. J Clin Ultrasound 10: 173-175, 1982 14. Giinther R, Klose KJ, Ruckert K: Islet cell tumors: Detection of small lesions with CT and US. Radiology 148:485-488, 1983 15. Galiber AK, Reading CC, Charboneau JW, et al: Localization of pancreatic insulinoma: Comparison of pre- and intraoperative US with CT and angiography. Radiology 166:405-408, 1988 16. Grant CS, van Heerden J, Charboneau JW, et al: Insulinoma: The value of introperative ultrasonography. Arch Surg 123:843-848. 1988 17. Paivlsalo M. Mlktiir5nen H, Siniluoto T, et al: Ultrasound

SOMATOSTA.TlN

RECEPTOR SCINTIGRAPHY

compared with computed tomography and pancreatic arteriography in the detection of endocrine tumors of the pancreas. Eur J Radio1 9:173-178. 1989 18. Doppman JL, Miller DL, Chang R, et al: Gastrinomas: Localization by means of selective intraarterial injection of secretin. Radiology 174:25-29, 1990 19. Rossi P, Baert A, Passariello R, et al: CT of functioning tumors of the pancreas. AJR 144:57-60, 1985 20. Krudy AC, Doppman JL, Jensen RT: Localization of isletcell tumors by dynamic CT: Comparison with plain CT, arteriography, sonography and venous sampling. AJR 143585-589, 1984 21. Wank SA, Doppman JL, Miller DL, et al: Prospective study of the ability of CT to localize gastrinomas in patients with ZollingerEllison syndrome. Gastroenterology 92:905-912, 1987 22. Ahlstriim H, Magnusson A, Grama D, et al: Preoperative localization of endocrine pancreatic tumors by intra-arterial dynamic computed tomography. Acta Radio1 31:171-175, 1990 23. Maton PN. Miller DL, Doppman JL, et al: Role of selective angiography rn the management of patients with the Zollinger-Ellison syndrome. Gastroenterology 92:9 13-9 18, 1987 24. Roche A, Raissonier A. Gillon-Savouret MC: Pancreatic venous sampling and arteriography in localizing insulinomas and gastrinomas: Procedure and results in 55 cases. Radiology 145:621-627, 1982 25. Martensson H, Bottcher G, Sundler F, et al: Localization and peptide content of endocrine pancreatic tumors. Ann Surg 2 1:607614, 1990 26. Norton JA, Shawker TH, Doppman JL, et al: Localization and surgical treatment of occult insulinomas. Ann Surg 212:615620, 1990 27. Roche A, Capeau J, Hahmi P: Methodes radiologiques de localisation des tumeurs endocrines du pancreas. Gastroenterol Clin Biol 7:49-58. 1983 28. Cherner JA. Doppman JL, Norton JA, et al: Selective venous sampling for gastrin to localize gastrinomas: A prospective assessment. Ann Intern Med 105:841-847. 1986 29. Williams DM: Intraarterial secretin injection for gastrinoma localization Radiology 174:9- 11, 1990 30. Frucht H. Doppman JL, Norton JA, et al: Gastrinomas: Comparison of MR imaging with CT, angiography, and US. Radiology 17 1:713-7 17, 1989

115

3 1. Stark DD, Moss AA, Goldberg HI: Nuclear magnetic resonance of the liver, spleen, and pancreas. Cardiovasc Intervent Radio1 8:329341, 1986 32. Amouyal P, Amouyal G. Hautecouverture M, et al: Apport de l’echoendoscopie dans le diagnostic topographique des insulinomes. Resultats intermediaires dune etude prospective et multicentrique. Gastroenterol Clin Biol 15:A88, 199 1 33. Rijsch T. Boyce G, Sivak MV, et al: Preoperative localization of pancreatic endocrine tumors not detected by ultrasound and CT. American Gastroenterology Association Meeting, Chicago, IL, May 1991 (abstract no. 2199) 34. Reubi J-C, Maurer R, von Werder K. et al: Somatostatin receptors in human endocrine tumors. Cancer Res 47:551-558, 1987 35. Reubi J-C, Hacki WH, Lamberts SWJ: Hormone-producing gastrointestinal tumors contain high density of somatostatin receptors. J Clin Endocrinol Metab 65: 1127-I 134, 1987 36. Kvols LK. Moertel CG, O’Connell MJ. et al: Treatment of the malignant carcinoid syndrome. Evaluation of a long-acting somatostatin analogue. N Engl J Med 3 15:663-666, 1986 37. Lamberts SWJ. Krenning EP, Reubi J-C: The role of somatostatin and its analogs in the diagnosis and treatment of tumors. Endocr Rev 12:450-482, 1991 38. Krenning EP, Bakker WH, Breeman WAP, et al: Localisation of endocrine-related tumors with radioiodinated analogue of somatostatin. Lancet 1:242-244. 1989 39. Lamberts SWJ, Hofland LJ, van Koetsveld PM, et al: Parallel in vivo and in vitro detection of functional somatostatin receptors in human endocrine pancreatic tumors: Consequences with regard to diagnosis, localization, and therapy. J Clin Endocrinol Metab 7 I : 566-574, 1990 40. Bakker WH, Krenning EP. Breeman WAP. et al: Receptor scintigraphy with a radioiodinated somatostatin analogue: Radiolabelling, purification, biologic activity, and in vivo application in animals. J Nucl Med 31:1501-1509, 1990 41. Bakker WH. Alberts R, Bruns C, et al: [“‘In-DTPA-DPHE]‘octreotide, a potential radiopharmaceutical for imaging of somatostatin receptor-positive tumors: Synthesis, radiolabeling and in vitro validation. Life Sci 49: 1583- I59 1, 199 1 42. Lamberts SWJ, Bakker WH, Reubi J-C. et al: Somatostatinreceptor imaging in the localization of endocrine tumors. N Engl J Med 323:1246-1249, 1990