Comparison of Somatostatin Analog and Meta-Iodobenzylguanidine Radionuclides in the Diagnosis andLocalization of Advanced Neuroendocrine Tumors

G. KALTSAS, M. KORBONITS, E. HEINTZ, J. J. MUKHERJEE, P. J. JENKINS,S. L. CHEW, R. REZNEK, J. P. MONSON, G. M. BESSER, R. FOLEY,K. E. BRITTON, AND A. B. GROSSMAN

Departments of Endocrinology (G.K., M.K., E.Z., J.J.M., P.J.J., S.L.C., J.P.M., G.M.B., A.B.G.),Diagnostic Radiology (R.R.), and Nuclear Medicine (R.F., K.E.B.), St. Bartholomew’s Hospital,London, United Kingdom EC1A 7BE

ABSTRACTA comparison has been made of [123I]meta-iodobenzylguanidine

([123I]MIBG) and [111In]pentetreotide scintigraphy in 54 patientswith a variety of neuroendocrine tumors of whom 46 patients hadmetastatic disease. [111In]Pentetreotide scintigraphy was more sen-sitive in detecting metastatic lesions, as demonstrated on computedtomography and/or magnetic resonance scanning, than [123I]MIBG:67% vs. 50% for carcinoid tumors (n 5 24), 91% vs. 9% for pancreaticislet cell tumors (n 5 12), 100% vs. 60% for medullary thyroid car-cinomas (n 5 5), and 75% vs. 100% for pheochromocytomas/paragan-gliomas (n 5 4). In only 2 patients were lesions seen with [123I]MIBGscanning that were not apparent with [111In]pentetreotide. With theexception of pancreatic islet cell tumors, both radionuclides exhibiteda similar sensitivity in detecting hepatic metastases, whereas in threepatients the two radionuclides exerted a complementary role as dif-ferent deposits exhibited uptake to only 1 or the other radionuclide.Hepatic metastases were the most important clinical predictor of a

positive scan for both radionuclides. Neither elevated 5-hydroxyin-doleacetic acid levels nor any other hormonal marker was predictiveof a positive scan. In 8 patients with clinical and/or hormonal evidenceof a neuroendocrine tumor but negative conventional radiology,[111In]pentetreotide scintigraphy was more sensitive than [123I]MIBG(37.5% vs. 12.5%) in detecting lesions.

In conclusion, scintigraphy with [111In]pentetreotide detects moremetastatic lesions than [123I]MIBG in patients with carcinoid andpancreatic islet cell tumors and medullary thyroid carcinomas;[123I]MIBG scintigraphy may be more sensitive for sympatho-adrenomedullary tumors. The radionuclides may exert a complemen-tary role in the detection and treatment of neuroendocrine tumors inoccasional patients, as areas of different pattern of uptake were iden-tified within the same patient. These data have implications not onlyfor staging such tumors, but also for identifying patients who mightbenefit from treatment using either [131I]MIBG or radioactive soma-tostatin analogs. (J Clin Endocrinol Metab 86: 895–902, 2001)

NEUROENDOCRINE tumors originate from cells pro-grammed to adopt a specific neuroendocrine pheno-

type (1). Although such tumors may vary in clinical presen-tation, location, and specific histology, all show characteristiccytoplasmic secretory granules on electron microscopy (2).Neuroendocrine tumors may also express active amine pre-cursor uptake-1 mechanisms and/or specific receptors at thecell membrane (2, 3). This has been used for both the detec-tion and treatment of such tumors, with the developmentand application of radiopharmaceuticals reliant upon suchspecific mechanisms.

Meta-iodobenzylguanidine (MIBG) is a catecholamine an-alog that uses the amine precursor uptake mechanism andmay thus be incorporated into vesicles or neurosecretorygranules in the cytoplasm; it has been shown to be highlysensitive for detecting tumors arising from the adrenal me-dulla and may be also taken up by nonadrenomedullaryneuroendocrine tumors (2–5). Radiolabeled octreotide, ananalog of somatostatin, is used in vivo to demonstrate tumors

that have somatostatin receptors on their surface. It has beensuggested that octreotide scanning may be more sensitive,although less specific, than MIBG in the detection of neu-roendocrine tumors other than pheochromocytomas, al-though direct comparisons were performed in differentpatients (2–5); however, these radionuclides may be com-plementary in the diagnosis of neuroendocrine tumors (2–6).In addition, considerable experience has been gained withthe use of [131I]MIBG for the treatment of a variety of neu-roendocrine tumors (4). Although the radionuclide profile of[111In]pentetreotide, the principal radiodiagnostic soma-tostatin analog, makes it a weak therapeutic agent, otherlabeled somatostatin analogs have recently been introducedinto clinical trial (5). However, until now the relative utilityof diagnostic radionuclides has been unclear, as the range ofneuroendocrine and related tumors in which formal com-parisons of [123I]MIBG and [111In]pentetreotide scintigraphyhave been made remains limited. In addition, information oncorrelations with other indicators of disease activity, e.g.symptoms and hormonal markers, is even more sparse, par-ticularly in patients with metastatic disease (3, 4, 6–9).

This study analyses the experience using both radiophar-maceuticals in a total of 54 patients. In 46 patients withdisseminated neuroendocrine tumors (carcinoid tumors,pheochromocytomas, medullary carcinomas of the thyroid,

Received April 11, 2000. Revision received September 11, 2000. Ac-cepted October 18, 2000.

Address all correspondence and requests for reprints to: Prof. A. B.Grossman, Department of Endocrinology, St. Bartholomew’s Hospital,London, United Kingdom ECIA 7BE. E-mail: a.b.grossman@mds.qmw.ac.uk.

0021-972X/01/$03.00/0 Vol. 86, No. 2The Journal of Clinical Endocrinology & Metabolism Printed in U.S.A.Copyright © 2001 by The Endocrine Society

895

islet cell tumors, and a pituitary carcinoma), both radiophar-maceuticals were used in the diagnosis and detection of theextent of the disease and in helping identify patients whomay benefit from treatment with either [131I]MIBG or, morerecently, labeled long-acting somatostatin analogs. A further8 patients with clinical and/or biochemical evidence of re-current neuroendocrine tumors or unidentified primary le-sions, not apparent on conventional radiology, were alsoincluded in the analysis. The results obtained from conven-tional radiology [computed tomography (CT)/magnetic res-onance (MR) imaging] were compared with those from theradiopharmaceuticals and were correlated with the presenceof symptoms and hormonal markers to identify possibleclinical or biological predictors of a positive scan. Prelimi-nary data on the first 7 patients of this cohort have previouslybeen reported (9). Brief details of the octreotide (but not[123I]MIBG) scans on some of the metastatic carcinoids andislet cell tumors have also been presented (10).

Subjects and MethodsSubjects

Fifty-four patients who had undergone imaging with both [123I]MIBGand [111In]pentetreotide scintigraphy were retrospectively analyzed.Forty-six patients (22 men and 24 women; mean age, 50 yr; range, 9–80yr) had histologically proven metastatic neuroendocrine tumors (in-cluding 1 patient with a metastatic prolactinoma); an additional 8 pa-tients had either histologically proven evidence of a neuroendocrinetumor in a single unusual site (2 breast carcinoids), clinical and/orbiochemical evidence of recurrence of a previously histologically proven

neuroendocrine tumor (n 5 5), or clinical and biochemical evidence ofa neuroendocrine tumor not seen on conventional imaging (n 5 1). Theorder of radionuclide imaging was varied according to logistic demands;in 32 patients the [123I]MIBG was given first, and in 22 the [111In]pente-treotide was the first imaging modality.

Metastatic neuroendocrine tumors

Carcinoid tumors (n 5 24). Twenty-four patients (12 men and 12 women;mean age, 54 yr; range, 27–80 yr) with a histologically proven diagnosisof carcinoid tumor were included in this group. The primary origin ofthe carcinoid tumors was known in 18 patients [bowel (11), bronchus (5),and pancreas (2)] and unknown in 6; in patients with an unknownprimary, the diagnosis was based on tissue biopsy from metastaticdeposits. Fifteen patients had imaging studies before any treatment wasapplied, and 9 had previous treatment (6 surgery alone, 1 surgery andradiotherapy, and 2 surgery and cytotoxic chemotherapy).

Islet cell carcinomas (n 5 12). Twelve patients (five men and seven women;mean age, 53 yr; range, 9–79 yr) had histologically proven islet celltumors (six gastrinomas, four vipomas, and two insulinomas); fourpatients had islet cell tumors in the context of multiple endocrine neo-plasia syndrome type 1. Six patients had undergone previous abdominalsurgery, and one had also received chemotherapy.

Medullary carcinoma of the thyroid (n 5 5). Five patients (two men andthree women; mean age, 33 yr; range, 27–53 yr) had histologically provenmedullary thyroid carcinomas; four patients had previously undergonethyroidectomy with neck dissection, and one had received a singletreatment of [131I]MIBG treatment (200 mCi).

Malignant pheochromocytomas and paragangliomas (n 5 4). Four patients(one man and three women; mean age, 37 yr; range, 25–53 yr) hadhistologically proven malignant sympatho-adrenomedullary tumors(three pheochromocytomas and one paraganglioma); all four patients

FIG. 1. Qualitativedifference in thepatternofuptakewith [123I]MIBG(upperpartof figure)and [111In]pentetreotide (lowerpartof figure) scintigraphyin the same patient with hepatic metastases from an ileal carcinoid. ➔, Positive pentetriotide uptake; ➩, positive [123I]MIBG uptake.

896 KALTSAS ET AL. JCE & M • 2001Vol. 86 • No. 2

had previously undergone surgery, whereas one patient had also re-ceived chemotherapy.

Malignant pituitary tumor (n 5 1). There was one patient with a malignantprolactinoma (male; age, 33 yr) who had undergone transsphenoidaland transfrontal surgery, external beam radiotherapy, and chemother-apy. He has been previously presented in more detail (11).

Nonmetastatic neuroendocrine tumors (n 5 8)

Eight patients were included in this group (three men and five wom-en; mean age, 49 yr; range, 31–85 yr). Seven patients had a previouslyhistologically proven neuroendocrine tumor (four carcinoids, one pheo-chromocytoma, and one medullary thyroid carcinoma), including onepatient with a pleural fibroma secreting proinsulin-like growth factor IIwho has also been previously reported (12); they underwent imaging foreither staging or detection of the site of the primary tumor (two patientshad breast carcinoids) or when seeking a recurrence in the face of clinicaland biochemical evidence of disease, but with negative conventionalimaging (n 5 5). One patient had biochemical evidence of a pheochro-mocytoma but negative conventional imaging studies.

Materials and Methods

The following clinical details and biochemical measurements wererecorded at the time of imaging according to an established investigationprotocol (13): symptoms of the carcinoid syndrome (diarrhea and/orflushing and/or bronchospasm); presence of headaches (hypertensivecrisis); palpitations and diaphoresis; weight loss of more than 10% ofbody weight; as well as symptoms related to any other relevant hor-monal secretion and/or symptoms associated with any local tumorcompressive effect: 24-h urinary 5-hydroxyindoleacetic acid excretion,serum and urinary catecholamines, serum a-fetoprotein, carcinoembry-onic antigen, hCGb, gastrointestinal hormones (vasointestinal peptide,pancreatic polypeptide, gastrin, glucagon, somatostatin, and neuroten-sin), calcitonin, and GHRH levels. Where relevant, the patients also hada five-point serum GH profile on a single day as an indicator of bio-chemical evidence of GH hypersecretion, and/or a midnight cortisoldetermination as evidence of autonomous cortisol secretion (13). In casesof ectopic humoral syndromes the relevant hormones were measured.All hormonal measurements were performed at the Department ofChemical Endocrinology, St. Bartholomew’s Hospital (London, UK), aspreviously published, except the gastrointestinal hormones, which weremeasured at the Department of Chemical Endocrinology, HammersmithHospital (London, UK) (13). The radiological assessment included con-ventional plain chest radiograph, CT and/or MR imaging of the abdo-men in all patients, and, when indicated, CT and/or MR scanning of thechest and/or bone and/or CT brain scanning. The positive findings werecompared with those obtained from the radionuclide scans.

CT and MR scanning

All scans were interpreted at the time of their performance by twoobservers who were blinded to the results of other investigations. Theywere then reviewed in conference by a radiologist (R.R.) who has specialexpertise in cross-sectional imaging. CT scans were carried out on an IGEHigh Speed Advantage (Milwaukee, IL) spiral CT scanner with 3–5 mmorgan-dependent collimation. The pitch was set at 1–1.5 and recon-structed on a high resolution algorithm depending on the organ. Non-ionic contrast medium was given at a rate of 3 mL/s to a total volume

of 100–150 mL. MR imaging was carried out on an IGE Signa 1.5T, theprecise sequence depending on the specific organ, but generally withspin echo T1- and T2-weighted, fat-suppressed T1-weighted, and gra-dient echo (SPER 90) imaging, with and without gadoliniumenhancement.

[123I]MIBG scintigraphy

Radioiodine-labeled MIBG was obtained from a commercial source(Amersham International, Gloucester, UK). Between 130 and 185 MBq[123I]MIBG were injected iv over 30 s. Imaging was obtained using a largefield of view camera (Simmons, Erlangen, Germany) set with a 15%window around the photo peak of 159 Kev, and a parallel hole, lowenergy, general purpose collimator. Images for 500,000 counts wereobtained at 10 min, 22 h, and 48 h postinjection covering the head, torso,and thighs, anteriorly and posteriorly. Images were stored on line in aHermes Sun workstation (Nuclear Diagnostics, Hagersten, Sweden).The thyroid uptake was blocked by 100–200 mg potassium iodideadministered orally for 2 days, starting the day before [123I]MIBGadministration and for 24 h thereafter.

[111In]DTPA-labeled-pentetreotide scintigraphy

An octreotide analog, pentetreotide (Octreoscan-111, Mallinckrodt,Inc., Petten, The Netherlands), was supplied as a radioactive preparationin single dose ampules. [111In]Pentetreotide was given as a bolus ivinjection at a dose up to 120 MBq. Whole body planar images wereobtained after 10 min and then after 4 and 21 h. Scanning was performedusing a g-camera with a large field of view (Counterbalance 3700, Sie-mens Gammasonics, Erlangen, Germany) interfaced to an on-line com-puter (DEC PDP11–44). Positive scans were defined as those in whichuptake of tracer occurred in areas not normally associated with itsaccumulation. No patient was receiving depot octreotide, and plainoctreotide treatment was stopped at least 48 h before the time of imaging,as the pharmacokinetic half-life of octreotide is of the order of 1–2 h.

All radionuclide studies were reported by the same physician withoutknowledge of the results of previous investigations. The scintigraphicresults were classified as positive if it was possible to prove the presenceof a tumor radiologically (by means of CT and/or MR scanning) orsurgically at a site of uptake or as negative. The term false positive wasused when there was apparent uptake on the scan but additional ra-diological examination did not reveal the presence of a metastatic de-posit at that time or on subsequent follow-up. A relationship was soughtbetween scintigraphic results using the two radionuclides and the an-atomical origin of the tumor, the site of deposits, the previously appliedlocal or systemic treatment, the clinical symptomatology for carcinoidsyndrome and pheochromocytoma, the presence of ectopic humoralsyndromes and hepatic metastases, and the presence of hormonal mark-ers. All comparisons were made using the x2 test, with statistical sig-nificance taken as P , 0.05.

ResultsMetastatic neuroendocrine tumors

Overall, scintigraphy with [111In]pentetreotide was posi-tive in 36 patients (78%) compared with [123I]MIBG in 20patients (43%; P , 0.05). Twenty-seven patients in total hadhepatic metastases; [111In]pentetreotide scintigraphy was

TABLE 1. A direct comparison of [123I]MIBG and [111In]-pentetreotide scintigraphy in displaying a variety of metastatic neuroendocrinetumors

Scan One site All sites Site other than liver Livera New lesion False 1

[123I]MIBG 20 (43.5) 11 (24) 7 (15) 15 (55) 3 (7) 4 (9)[111In]pentetreotide 36 (78) 26 (56.5) 21 (46) 22 (81) 10 (22) 3 (7)

The sensitivities in detecting lesions at any single site are shown, as well as their sensitivity in demonstrating all lesions located on CT and/orMR imaging. These are also analyzed according to whether the lesions are within or without the liver. New lesions are defined as those notpreviously suspected clinically, but confirmed on conventional imaging. Finally, false positives are defined as scintigraphically positive areasthat show no abnormality on conventional imaging, but may, in fact, represent very early areas of functional activity below the limit of resolutionof current CT or MR imaging modalities. Percentages are given in parentheses.

aIncluding patients with metastatic islet cell tumors.

RADIONUCLIDE DIAGNOSIS OF NEUROENDOCRINE TUMORS 897

TABLE 2. Clinical, hormonal, radiological, and radionuclear imaging findings in 46 patients with metastatic neuroendocrine tumors andin a patient with a metastatic prolactinoma

Patient no. Age (yr) Diagnosis Previous therapy Symptoms/signs

1 73 Carcinoid NAD Abdominal distention, massive liver

2 41 Carcinoid Surgery, CCNU/5FU Hemoptysis, acromegaly, carcinoid

3 41 Carcinoid NAD Carcinoid, WL, heart failure(tricuspid regurgitation), pulmonary HT

4 32 Carcinoid NAD Abdominal pain, hepatomegaly

5 68 Carcinoid NAD Carcinoid6 32 Carcinoid NAD Cushing’s syndrome (ectopic ACTH secretion)7 53 Carcinoid NAD Carcinoid, massive liver8 44 Carcinoid NAD Carcinoid, WL9 50 Carcinoid NAD LN enlargement

10 27 Carcinoid Surgery, RT WL, obstructive jaundice

11 70 Carcinoid NAD Abdominal discomfort, massive liver

12 55 Carcinoid Surgery Carcinoid, WL

13 71 Carcinoid Surgery Carcinoid, postmeal flush14 72 Carcinoid NAD Recurrent hypoglycemia, pulmonary

HT, hepatomegaly15 50 Carcinoid Surgery Carcinoid, hepatomegaly tricuspid regurgitation

16 28 Carcinoid NAD Asymptomatic, breast nodule17 77 Carcinoid NAD Carcinoid, hemoptysis, WL

18 52 Carcinoid Surgery Hemoptysis, liver mets

19 56 Carcinoid NAD Abdominal discomfort, mass

20 80 Carcinoid NAD Carcinoid, WL, massive liver21 49 Carcinoid Surgery, doxorubicin

and a interferonCarcinoid, WL, hepatomegaly

22 58 Carcinoid Surgery Abdominal discomfort, WL, hepatomegaly23 76 Carcinoid Surgery Protrusion L eye24 44 Carcinoid NAD Carcinoid25 53 Pheo Adrenal surgery Intrascapular pain, liver mets26 25 Paraganglio Adrenal surgery Hoarseness voice, carotid mass27 29 Pheo Adrenal surgery Malignant HTN, palpitations, liver mass28 44 Pheo Retroperitoneal surgery,

CCNU/5FURecurrent abdominal mass

29 27 Medullary Surgery, [131I]MIBG Recurrent neck mass, LN enlargement, liver mets

30 53 Medullary Thyroidectomy and LNenlargement

Recurrent neck mass, LN enlargement

31 33 Medullary NAD Cushing’s (ectopic ACTH production)

32 45 Medullary Thyroidectomy Asymptomatic, ? LN enlargement33 28 Medullary Thyroidectomy, LN Thyroid mass34 41 Gastrinoma, MEN1 Surgery Epigastric pain

35 54 Insulinoma Surgery 5FU-a interferon WL, abdominal distention36 75 Insulinoma NAD Hypoglycemias

37 58 Gastrinoma, MEN 1 NAD Hypergastrinemic complaints38 49 Gastrinoma, MEN1 Surgery Deteriorating vision

39 57 Gastrinoma, MEN 1 Surgery, CCNU/5FU Recurrent hypergastrinemia

40 36 Gastrinoma NAD Cushing’s (ectopic ACTH), WL, recurrent pancreatitis

41 9 Gastrinoma Surgery Recurrent peptic ulcers42 79 VIPome Surgery Abdominal mass, WL, diarrhea, flushing

43 52 VIPome Surgery Recurrent diarrhea44 42 VIPome NAD Recurrent diarrhea, WL45 68 VIPome NAD Diarrhea46 33 Pituitary Surgery, RT Visual field defects

898 KALTSAS ET AL. JCE & M • 2001Vol. 86 • No. 2

TABLE 2. Continued

Hormonal Radiology [123II]MIBG [111In]Pentetreotide

1 AFP, CEA Liver mets No uptake 1 Uptake liver, L chest parahilararea

1 5HIAA, GH, GHRH,calcitonin

Liver mets, mets skull, sternum,ribs, thoracic/lumbar spine

No uptake 1 Uptake liver, skull, subclavianspine, breast

1 5HIAA Liver mets, 4 3 3.5 cm enhancingpelvic mass

1 Uptake liver and upperabdomen

1 Uptake liver, central abdomen,orbit

Liver & thoracic spinal mets, Lsuprarenal & vertebral mass

1 Uptake liver, R adrenal 1 Uptake liver & porta hepatis

1 Calcitonin Abnormal thickening mesentery No uptake No uptake1 Cortisol (ACTH) Small lesion posterior chest No uptake No uptake1 5HIAA Liver mets 1 Uptake liver 1 Uptake liver1 5HIAA Liver mets 1 Uptake liver, midline 1 Uptake liver, midline

Soft tissue mass thoracic region No uptake 1 Uptake L hilumLiver mets, pancreatic mass,

inferior venae cavae compressionNo uptake No uptake

1 5HIAA Liver mets 1 Uptake liver (1), ? focus upperabdo

1 Uptake liver (11), ? focusoutside liver coeliac axis

Thickening rectal wall, sarcoileacjoint mets

1 Uptake midline abdo, ? 1 liveruptake

1 Uptake midline abdo, 1 liveuptake

1 5HIAA, calcitonin Liver mets, ? peritoneal deposits 1 Uptake liver 1 Uptake liver1 Insulin Liver mets, pancreatic masses,

LMN1 Uptake liver (11) 1 Uptake liver (1)

1 5HIAA Liver mets, thickening mesentery 1 Uptake liver, paraaortic 1 Uptake liver, paraaortic, Rhilum

1 5HIAA Liver mets 1 Uptake liver No uptake1 5HIAA Mass L chest, liver mets,

paraesophageal massNo uptake 1 Uptake liver, mediastinum

1 5HIAA Mass chest, liver mets, spinal mets 1 Uptake liver 1 Uptake liver, chest, bones,scrotum

1 Gastrin, SMS, calcitonin Retroperitoneal mass, paraaorticLMN

No uptake No uptake

1 5HIAA, calcitonin Liver mets, distortion IVC 1 Uptake liver 1 Uptake liver1 5HIAA Liver mets 1 Uptake liver No uptake1 5HIAA Liver mets No uptake No uptake

Eye, pleura, bony mets No uptake 1 Uptake L orbit, chest, spine, ribs1 5HIAA Liver mets No uptake No uptake1 Catecholamines Liver mets 1 Uptake liver (11) 1 Uptake liver (1)1 Catecholamines Carotid masses 1 Uptake carotid area 1 Uptake carotid area1 Catecholamines Liver mets, spinal mets 1 Uptake liver, spine 1 Uptake liver1 Catecholamines Suprarenal mass, bony rhets 1 Uptake liver, bony mets 1 Uptake liver, bony mets

1 Calcitonin Soft tissue mass neck,mediastinum, liver mets

1 Uptake neck, chest 1 Uptake neck, chest

1 Calcitonin Soft tissue mass neck 1 Uptake neck 1 Uptake neck

1 Calcitonin Soft mass thyroid, multiple noduleslung

No uptake 1 Uptake neck

1 Calcitonin Soft tissue mass neck No uptake 1 Uptake neck1 Calcitonin, CEA Soft tissue mass neck 1 Uptake neck 1 Uptake neck1 Gastrin Mass upper pole L kidney 1 Uptake L kidney, mediastinum 1 Uptake L kidney, mediastinum

pancreas1 Insulin Liver mets, pancreatic mass No uptake 1 Uptake liver1 Insulin Tumor head pancreas, adrenal

involvementNo uptake No uptake

1 Gastrin, PP, b hCG Mass head pancreas, ? duodenum No uptake 1 Focal uptake pancreas1 Gastrin, VIP, 5 HIAA Eye, chest, liver, bony mets No uptake 1 Uptake mediastinum, liver,

bone, R orbit1 Gastrin, PP, CT Subaortic LN, extensive gastric

varicesNo uptake 1 Focal uptake midline abdomen

1 Gastrin Mass tail pancreas, LNenlargement

No uptake 1 Focal uptake tail pancreas

1 Gastrin Hepatic mets No uptake 1 Uptake liver, L lung1 VIP, neurotensin Pancreatic mass, paraaortic lymph

nodesNo uptake 1 Uptake tail pancreas, paraaorti

LN, liver1 VIP, PP Liver mets No uptake 1 Uptake liver1 VIP, SMS Liver mets, paraaortic lymph node No uptake 1 Uptake liver1 hCGB, VIP, PP, glucagon Hepatic mets, pancreatic mass No uptake 1 Uptake liver1 PRL Pituitary tumor, extention orbit,

cortexNo uptake 1 Uptake pituitary

Carcinoid, Carcinoid syndrome; mets, metastases; WL, weight loss; LN, lymph node; IVC, inferior venae cavae; RT, radiotherapy; L, left; R, right.

RADIONUCLIDE DIAGNOSIS OF NEUROENDOCRINE TUMORS 899

positive in 22 patients (81%), and [123I]MIBG scintigraphywas positive in 15 patients (55%). However, in 3 patients inwhom both scans were positive there were complementaryareas of uptake (Fig. 1). The overall performance of [111In]-pentetreotide and [123I]MIBG scintigraphy is shown inTable 1.

Carcinoid tumors (n 5 24)

Sixteen patients with metastatic carcinoid tumors had pos-itive [111In]pentetreotide scintigraphy (67%), and 12 positive[123I]MIBG scintigraphy (50%). The presence of a positivescan was not affected by the primary origin of the tumor,previous localized treatment (surgical resection and/or focalradiotherapy), or chemotherapy (Table 2). The presence ofnonlocal clinical symptoms also did not predict a positivescan, although patients with the carcinoid syndrome moreoften had a positive than a negative scan (12 of 18 vs. 2 of 8;P 5 0.09). No association was found with individual com-ponents of the carcinoid syndrome and a positive scan. Eigh-teen patients had liver metastases (75%): [111In]pentetreotidescintigraphy was positive in showing hepatic metastases in14 of 18 (78%), whereas [123I]MIBG scintigraphy was positivein 12 of 18 (67%); scintigraphy with both nuclides was neg-ative in 2 patients. However, in 2 patients [111In]pentet-reotide scintigraphy was negative, and [123I]MIBG was pos-itive, whereas in 4 patients [123I]MIBG scintigraphy wasnegative, and [111In]pentetreotide was positive. The presenceof increased levels of all hormones measured was not relatedto a positive scan; 12 of the 16 patients with positive scin-tigraphy with either radiopharmaceutical had at least 1 el-evated hormonal marker. Of the 24 patients with carcinoidtumors, 8 had negative [111In]pentetreotide scintigraphy, and6 of these had a raised tumor marker (Table 2).

Scintigraphy with [111In]pentetreotide detected six newlesions, which had not clinically previously been suspected,whereas [123I]MIBG detected only one, also shown by[111In]pentetreotide. Scintigraphy with [111In]pentetreotidewas particularly sensitive in detecting orbital lesions (twopatients were found subsequently to have choroidal depos-its). In two patients the two radionuclides exerted a com-plementary role as different deposits exhibited uptake toonly one or another radionuclide.

Two false positive results with each of [111In]pentetreotideand [123I]MIBG scintigraphy were noted.

Islet cell carcinomas (n 5 12)

Eleven patients had positive [111In]pentetreotide scintig-raphy (91%), whereas only one patient with a gastrinomaassociated with the multiple endocrine neoplasia syndrometype 1 syndrome had positive [123I]MIBG scintigraphy (9%);in this patient scintigraphy with both radiopharmaceuticalsrevealed a previously unsuspected lesion in the mediasti-num. Scintigraphy with [111In]pentetreotide also revealedclinically nonsuspected lesions in the orbit and abdomen;both scans were negative in one patient with an insulinoma.In a patient with elevated serum somatostatin levels, scin-tigraphy with [111In]pentetreotide was positive (Table 2).With the exception of one patient with an insulinoma whoalso had negative scintigraphy with [111In]pentetreotide, all

patients responded clinically and biochemically to systemicoctreotide treatment.

Medullary thyroid carcinoma (n 5 5)

All five patients had positive scintigraphy with [111In]-pentetreotide and three with [123I]MIBG; scintigraphy withboth radiopharmaceuticals detected a previously unsus-pected deposit in one patient, whereas in one patient withneck metastases the two radionuclides exhibited a differentpattern of uptake. Previous surgical treatment or treatmentwith [131I]MIBG did not affect the uptake of the radiophar-maceuticals. Subsequent treatment with octreotide did notaffect the elevated hormonal levels. Scintigraphy with[123I]MIBG was falsely positive in one case where apparentadrenal uptake was not confirmed on CT imaging.

Pheochromocytomas and paragangliomas (n 5 4)

All patients had elevated plasma and urinary catechol-amine levels; all four patients with pheochromocytomas hadpositive [123I]MIBG, and three showed positive [111In]pente-treotide scintigraphy; [123I]MIBG scintigraphy detected twopreviously unsuspected lesions, whereas scintigraphy withboth radiopharmaceuticals was false positive on two differ-ent occasions, both showing apparent hepatic metastases notseen on CT or MR scanning. Patients with false positive scansdid not develop obvious lesions on CT and/or MR scanningon subsequent follow-up.

Hepatic metastases

In total, 27 patients had hepatic metastases (18 carcinoids,6 pancreatic tumors, 2 pheochromocytomas, and 1 medullarythyroid carcinoma); [111In]pentetreotide scintigraphy waspositive in 22 (81%), compared with 15 (55%) for [123I]MIBG(P , 0.05). Scintigraphy with [111In]pentetreotide and[123I]MIBG exhibited a 76% vs. 71% sensitivity in detectingradiologically demonstrable hepatic metastases in patientswith carcinoid tumors, medullary thyroid carcinomas, andpheochromocytomas combined (P . 0.05). In the patientwith malignant prolactinoma, scintigraphy with [123I]MIBGwas negative, but scintigraphy with [111In]pentetreotide re-vealed uptake in the pituitary carcinoma, but not in themetastases.

Nonmetastatic neuroendocrine tumors

Scintigraphy with [111In]pentetreotide was positive inthree patients (37.5%), whereas [123I]MIBG was positive inonly a single patient (12.5%); this patient also had positive[111In]pentetreotide scintigraphy. Scintigraphy with [111In]-pentetreotide was positive in two patients with breast car-cinoid tumors, but did not reveal evidence of another pri-mary tumor or metastases; it also failed to locate a lesion intwo patients with clinical suspicion of recurrence of a pre-viously histologically proven carcinoid tumor of the bron-chus and abdomen. Scintigraphy with both [111In]pentet-reotide and [123I]MIBG was negative in two patients withbiochemical evidence of primary or recurrent pheochromo-cytomas. However, scintigraphy with [111In]pentetreotidewas positive in one patient who presented with recurrent

900 KALTSAS ET AL. JCE & M • 2001Vol. 86 • No. 2

hypoglycemia and was found to have pleural fibroma withneuroendocrine features secreting pro-IGF-II. Octreotidetreatment in this patient failed to relieve the hypoglycemia.

Discussion

Imaging with radiolabeled octreotide analogs and MIBGhas been extensively used in the detection and treatment ofneuroendocrine tumors (1–3, 14, 15). Radiolabeled MIBG hasbeen shown to have a more than 90% sensitivity and veryhigh specificity for the diagnosis of adrenomedullary tu-mors, but a sensitivity of only 25–70% in detecting otherneuroendocrine tumors (2, 3, 7, 15). Radiolabeled octreotidehas been shown to have a detection rate of 67–91% for allneuroendocrine tumors (3, 15), but is less specific, as uptakecan be demonstrated in many other tumors, granulomas, andautoimmune diseases (2). Although it has been suggestedthat these radionuclides may exert a complementary role inthe imaging of neuroendocrine tumors, previous details havefor the most part derived from studies performed in differentpatients (3, 4, 7). Very little work has been published onformal comparisons in individual patients (3, 6–9, 14–21); inaddition, details on clinical and biochemical predictors of apositive scan are even more sparse and occasionally contra-dictory (6, 7, 22, 23).

The results of this study in patients with a variety ofneuroendocrine tumors clearly suggest that [111In]pentet-reotide scintigraphy is in general more sensitive in detectingneuroendocrine tumors than radiolabeled MIBG (2–4, 6).However, occasionally scintigraphy with [123I]MIBG mayappear to play a complementary role by demonstrating up-take in nonoctreotide avid lesions, mainly in the context ofhepatic metastases.

The few comparative studies between [123I]MIBG and[111In]pentetreotide scintigraphy in carcinoid tumors haveshown a higher overall [111In]pentetreotide sensitivity for thedetection of either one or all metastatic sites (7, 16); however,a complementary role of [123I]MIBG scintigraphy was notedas either different intensity of uptake or uptake in non-octreotide-avid regions (6–8). In our patients, although scin-tigraphy with [111In]pentetreotide was the more sensitivetechnique overall, the sensitivities of the two techniques indetecting liver metastases were comparable, e.g. 78% for[111In]pentetreotide vs. 67% for [123I]MIBG scintigraphy.Scintigraphy with [111In]pentetreotide was more sensitive indetecting extraabdominal (9) and previously unsuspectedlesions (2, 7, 8), whereas a particular predilection of thisradiopharmaceutical was noted in identifying orbital lesions;unlike [111In]pentetreotide, no previously unsuspected le-sion was detected solely with [123I]MIBG scintigraphy (7). Nocorrelation of a positive scan was found with the anatomicalsite of origin of the carcinoid tumor (7), the presence offlushing (7), or the presence of the carcinoid syndrome (6).Previous surgical or systemic therapy did not alter the uptakeof these radiopharmaceuticals and should not therefore be areason for excluding [111In]pentetreotide or [123I]MIBG scin-tigraphy. Although a variety of hormonal markers were in-vestigated, no specific hormonal predictors of a positive scanwere identified.

The majority of endocrine pancreatic tumors can be visu-

alized using [111In]pentetreotide scintigraphy (3, 24, 25);[123I]MIBG has been reported as having an overall low sen-sitivity of 25% (3, 24). Only 1 of the 12 patients with malignantpancreatic tumors demonstrated an uptake in scintigraphywith [123I]MIBG, whereas scintigraphy with [111In]pentet-reotide was almost always positive, even in patients withhigh circulating somatostatin levels.

Studies in different patients with comparable tumors haveshown scintigraphy with [111In]pentetreotide to be twice aslikely to demonstrate lesions as [123I]MIBG in patients withmedullary thyroid carcinomas (2, 3). Our observations com-bined with the few existing formal comparative studies haveshown that [111In]pentetreotide scintigraphy is more likely tovisualize primary tumors and lymph node and bone metas-tases, but less successful in detecting liver metastases,whereas [123I]MIBG has limited diagnostic effectiveness (20,26). These findings confirm the superior sensitivity of radio-labeled octreotide, but indicate that both scans may still beusefully applied on occasions in individual patients, as he-patic metastases with a different pattern of uptake may beidentified by the two radionuclides.

In the small number of reports including comparisonsbetween [123I]MIBG and [111In]pentetreotide scintigraphy inpatients with malignant pheochromocytomas, [123I]MIBGwas noted to have an overall higher detection rate, whereasscintigraphy with [111In]pentetreotide may occasionally re-veal [123I]MIBG-negative lesions (18, 21); a higher rate of falsepositive [111In]pentetreotide scintigraphy has also been de-scribed (18). Similarly, in this relatively small group of pa-tients scintigraphy with [123I]MIBG was more sensitive, butthe complementary role of [111In]pentetreotide was estab-lished as differential location of uptake in hepatic metastaseswas seen. These findings may have clinical implications, inthat [111In]pentetreotide scintigraphy could be adjunctive tothe other diagnostic methods used for staging malignantpheochromocytomas.

Although somatostatin receptors have been described inthe pituitary, mainly in GH-secreting tumors, prolactinomasin general have a relative poor [111In]pentetreotide uptake(27). Although experience with scintigraphy in malignantpituitary tumors is limited, [111In]pentetreotide scintigraphyhas revealed previously undetected lesions, established thesite of metastases, and revealed tumor recurrence at fol-low-up (28). However, in our patient, despite the fact that[111In]pentetreotide revealed uptake in the pituitary region,no response to systemic octreotide administration was noted.

In patients with a strong suspicion of a neuroendocrinetumor but in whom all imaging modalities were negative,scintigraphy with [111In]pentetreotide identified more le-sions than [123I]MIBG, although the detection rate was stilllow. In such patients, hormonal measurements are morepredictive, and other imaging modalities, e.g. venous sam-pling, may be necessary to reveal these indolent tumors (25).Both radiopharmaceuticals showed occasional positive up-take despite negative imaging on conventional radiology,defined as false positives. However, it is possible that scin-tigraphy is actually correctly detecting lesions that are func-tioning but too small to be seen on MR/CT scanning. It ispossible that positron emission tomography scanning mayalso be useful in such situations (29).

RADIONUCLIDE DIAGNOSIS OF NEUROENDOCRINE TUMORS 901

Targeted radiotherapy with radionuclides has been effec-tive in treating tumors because of the localization of theb-emitting radionuclide, either within the tumor cell or in itsproximity, compared with nontarget cells. [131I]MIBG, basedon a positive [123I]MIBG uptake, has been used in the treat-ment of malignant pheochromocytomas, neuroblastomas,carcinoid tumors, and medullary thyroid carcinomas; theoverall consensus is that although complete responses havebeen observed in only a minority of cases, partial responsesare common, whereas in most cases clinically useful hor-monal and symptomatic responses may be obtained (3, 5).Radiolabeled somatostatin analogs represent a separatenovel therapeutic modality of neuroendocrine tumors. Thecurrent data suggest that therapy with somatostatin analogsradiolabeled with b-emitters will be applicable to a largerpopulation of patients with neuroendocrine tumors, and willgenerally locate the same lesions identified by [123I]MIBG(30). This study suggests that lesions not shown by scintig-raphy with [111In]pentetreotide will only rarely be picked upby scintigraphy with [123I]MIBG, although occasional pa-tients, mainly with hepatic metastases, may show comple-mentary uptake with [123I]MIBG and [111In]pentetreotidescintigraphy: in such patients therapy with both labeled ra-dionuclides might be feasible.

In conclusion, scintigraphy with [111In]pentetreotide ingeneral detects more metastatic lesions than [123I]MIBG inpatients with neuroendocrine tumors. In occasional patientsscintigraphy with [123I]MIBG demonstrated lesions not evi-dent with [111In]pentetreotide. No reliable clinical and/orbiological predictors other than the presence of hepatic me-tastases exist. Although both were effective in identifyinghepatic metastases, there were qualitative differences in thepattern of uptake. It is concluded that scintigraphy with[123I]MIBG may indicate areas of uptake in a significant pro-portion of patients with neuroendocrine tumors, whichwould suggest the feasibility of [131I]MIBG therapy. Whenthis is negative, [111In]pentetreotide is more likely to dem-onstrate lesions in all types of neuroendocrine tumors, otherthan those originating from sympatho-adrenomedullary tis-sue, and may indicate those in whom treatment with b-emit-ting somatostatin analog radiopharmaceuticals might beindicated.

References

1. Langley K. 1994 The neuroendocrine concept today. Ann NY Acad Sci.733:1–17.

2. Hoefnagel CA. 1995 MIBG and radiolabelled octreotide in neuroendocrinetumors. Q J Nucl Med. 39(Suppl 1–4):137–139.

3. Krenning EP, Kwekkeboom DJ, Pauwels S, Kvols LK, Reubi JC. 1995 So-matostatin receptor scintigraphy. In: Freeman L, ed. Nuclear medicine annals.New York: Raven Press; 1–50.

4. Shapiro B. 1995 Ten years of experience with MIBG applications and thepotential of new radiolabeled peptides: a personal overview and concludingremarks. Q J Nucl Med. 39(Suppl 1–4):150–155.

5. Wiseman GA, Kvols LK. 1995 Therapy of neuroendocrine tumors with ra-diolabelled MIBG and somatostatin analogues. Semin Nucl Med. 25:272–278.

6. Taal BG, Hoefnagel CA, Valdes Olmos RA, Boot H. 1996 Combined diag-nostic imaging with 131I-metaidobenzylguanidine and 111In-pentetreotide incarcinoid tumours. Eur J Cancer. 32:1924–1932.

7. Nocaudie-Calzada M, Huglo D, Carnaille B, Proye C, Marchanddise X. 1996Comparison of somatostatin analogue and metaiodobenzylguanidine scintig-raphy for the detection of carcinoid tumours. Eur J Nucl Med. 23:1448–1454.

8. Carnaille B, Nocaudie M, Pattou F, et al. 1994 Scintiscans and carcinoidtumors. Surgery. 116:1118–1122.

9. Bomanji J, Mather S, Moyes J, et al. 1992 A scintigraphic comparison ofiodine-123-metaidobenzylguanidine and an iodine labeled somatostatin ana-log (Tyr-3-octreotide) in metastatic carcinoid tumors. J Nucl Med.33:1121–1124.

10. Ur E, Bomanji S, Mather SJ, et al. 1993 Localisation of neuroendocrine tumorsand insulinomas using radiolabelled somatostatin analogues, 123I-Tyr3-oct-reotide and 111In-pentetreotide. Clin Endocrinol (Oxf). 38:501–506.

11. Kaltsas GA, Mukherjee JJ, Plowman PN, Monson JP, Grossman AB, BesserGM. 1998 The role of cytotoxic chemotherapy in the management of agressiveand malignant pituitary tumors. J Clin Endocrinol Metab. 83:4233–4238.

12. Drake WM, Miraki F, Siddiqi A, et al. 1998 Dose-related effects of growthhormone on IGF-I and IGF-binding protein-3 levels in non-islet cell tumourhypoglycemia. Eur J Endocrinol. 139:532–536.

13. Trainer PJ, Besser GM. 1995 The Bart’s endocrine protocols, 1st Ed. London:Churchill Livingstone.

14. Kaltsas GA, Putignano P, Mukherjee JJ, et al. 1998 Carcinoid tumours pre-senting as breast cancer: the utility of radionuclide imaging with 123I-MIBG and111In-DTPA pentetreotide. Clin Endocrinol (Oxf). 49:685–689.

15. Krenning EP, Kwekkeboom DJ, Bakker WH, et al. 1993 Somatostatin receptorscintigraphy with [111InDTPA d Phe1] and [123ITyr3] octreotide: the Rotterdamexperience with more than 1000 patients. Eur J Nucl Med. 20:716–731.

16. Ramage JK, Williams R, Buxton-Thomas M. 1996 Imaging secondary neu-roendocrine tumours of the liver: comparison of I123 metaidobenzylguanidine(MIBG) and In111-labelled octreotide (Octreoscan). Q J Med. 89:539–542.

17. Lauriero F, Rubini G, D’Addabbo F, Rubini D, Shettini F, D’Addabbo A.1995 I-131 MIBG scintigraphy of neuroectodermal tumors:Comparison be-tween I-131 MIBG and In-111 DTPA-octreotide. Clin Nucl Med. 20:243–249.

18. Tenenbaum F, Lumbroso J, Schlumberger M, et al. 1995 Comparison ofradiolabeled octreotide and meta-iodobenzylguanidine (MIBG) scintigraphyin malignant pheochromocytoma. J Nucl Med. 36:1–6.

19. Lastoria S, Maurea S, Vergara E, et al. 1995 Comparison of labeled MIBG andsomatostatin analogs in imaging neuroendocrine tumors. Q J Nucl Med.39:145–149.

20. Rufini V, Salvatori M, Saletnich I, et al. 1995 Radiolabelled somatostatinanalog scintigraphy in medullary thyroid carcinoma and carcinoid tumor. QJ Nucl Med. 39(Suppl 1–4):140–144.

21. Kopf D, Bockisch A, Steinert H, et al. 1997 Octreotide scintigraphy andcatecholamine response to an octreotide challenge in malignant phaeochro-mocytoma. Clin Endocrinol (Oxf). 46:39–44.

22. Kwekkeboom DJ, Krenning E, Bakker WH, Yoe Oei H, Kooij PPM, LambertsSWJ. 1993 Somatostatin analog scintigraphy in carcinoid tumors. Eur J NuclMed. 20:283–292.

23. Shalaby-Rana E, Majd M, Andrich MP, Movassaghi N. 1997 In-111pente-treotide scintigraphy in patients with neuroblastoma. Comparison with I-131MIBG, N-Myc oncogene amplification, and patient outcome. Clin Nucl Med.22:315–319.

24. Hoefnagel CA. 1994 Metaiodobenzylguanidine and somatostatin in oncology:role in the management of neural crest tumors. Eur J Nucl Med. 21:561–581.

25. Scott BA, Gatenby MD. 1998 Imaging in the diagnosis of endocrine neoplasia.Curr Opin Oncol. 10:37–42.

26. Kwekkeboom DJ, Reubi JC, Lamberts SWJ, et al. 1993 In vivo somatostatinreceptor imaging in medullary thyroid carcinoma. J Clin Endocrinol Metab.76:1413–1417.

27. Tofani A, Cucchi R, Pompili A, et al. 1995 111In-octrotide scintigraphy inpituitary adenomas. Q J Nucl Med. 39(Suppl 1–4):94–97.

28. Kaltsas GA, Grossman AB. 1998 Malignant pituitary tumors. Pituitary.1:69–81.

29. Arnold DR, Villemagne VL, Civelek CA, Dannals RF, Wagner HN, Udels-man R. 1998 FDG-PET: a sensitive tool for the localisation of MIBG-negativepelvic pheochromocytomas. Endocrinologist. 8:295–298.

30. Oberg K. 1999 Neuroendocrine gastrointestinal tumors-a condensed overviewof diagnosis and treatment. Ann Oncol. 10(Suppl 2):S3–S8.

902 KALTSAS ET AL. JCE & M • 2001Vol. 86 • No. 2