J Clin Oncol 23: by American Society of Clinical Oncology INTRODUCTION

Transcription

1 VOLUME 23 NUMBER 1 JANUARY JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T Detection of Liver Metastases From Endocrine Tumors: A Prospective Comparison of Somatostatin Receptor Scintigraphy, Computed Tomography, and Magnetic Resonance Imaging Clarisse Dromain, Thierry de Baere, Jean Lumbroso, Hubert Caillet, Agnès Laplanche, Valerie Boige, Michel Ducreux, Pierre Duvillard, Dominique Elias, Martin Schlumberger, Robert Sigal, and Eric Baudin From the Department of Radiology, Institut Gustave-Roussy, Villejuif Cedex, France. Submitted January 6, 2004; accepted September 7, Authors disclosures of potential conflicts of interest are found at the end of this article. Address reprint requests to Clarisse Dromain, MD, Department of Radiology, Institut Gustave-Roussy, 39, rue Camille Desmoulins, Villejuif Cedex, France; dromain@igr.fr by American Society of Clinical Oncology X/05/ /$20.00 DOI: /JCO A B S T R A C T Purpose To compare the respective sensitivity of somatostatin receptor scintigraphy (SRS), computed tomography (CT), and magnetic resonance imaging (MRI) in the detection of liver metastases from well-differentiated gastroenteropancreatic endocrine tumor (WDGEP ET) patients. To define predictive factors for high-sensitivity SRS. Patients and Methods Sixty-four patients with WDGEP ET underwent SRS with abdominal single-photon emission computed tomography (SPECT), spiral CT, and 1.5-T MRI within a 15-day interval, the order of which was randomized. Two readers analyzed images of each modality, blindly and independently. Results Hepatic metastases were present in 40 of the 64 patients and confirmed by pathology after liver biopsy or surgery in 32 and eight patients, respectively. SRS, CT, and MRI detected a total of 204, 325, and 394 metastases, respectively. The number of detected metastases was significantly higher with MRI than with CT (P.02) and SRS (P 10 4 ) and higher with CT than with SRS (P 10 4 ). SRS was negative in seven patients with a positive CT and/or MRI. More lesions were detected in 10 patients by SPECT compared with static views. The median metastasis size was significantly correlated (P.04) with the sensitivity of SRS. Conclusion MRI seems to have an edge over CT and SRS for the detection of liver metastases from endocrine tumors. We recommend the systematic performance of liver MRI at WDGEP ET initial staging and before major therapeutic events. The low performance of SRS was mainly explained by the impact of the metastasis size on the detection capacity of SRS. J Clin Oncol 23: by American Society of Clinical Oncology INTRODUCTION Well-differentiated gastroenteropancreatic (WDEGP) neuroendocrine tumors (ET), recently reclassified as endocrine tumors by the 2000 WHO classification, 1 all express specific neuroendocrine markers. They also share common clinical features including hormone secretion, their association with an inherited syndrome, and a long-term natural history in a subgroup of welldifferentiated endocrine carcinoma. The two main prognostic parameters in ET are pathological differentiation and stage. 2-5 Imaging of well-differentiated endocrine carcinoma has several goals, the more important among them being detection of the primary tumor staging and the identification of tumors forming part of the spectrum of an inherited syndrome in a subgroup of ET. 70

2 Detection of Liver Metastases In patients with gastroenteropancreatic endodermalderived ET, clinicians have taken advantage of somatostatin receptor scintigraphy (SRS) with radiolabeled octreotide, a functional imaging modality for tumor staging. SRS allows visualization of the somatostatin receptor, and especially subtype 2, which is expressed by the great majority of welldifferentiated endocrine carcinoma. 6 Introduced in 1989, SRS has led to upstaging in 10% to 43% of patients with a modification of previous therapeutic options in some cases The specificity of SRS has also been shown to be high. 12 Finally, most authors consider that SRS should be considered as a complementary conventional imaging tool in patients with ET. The presence of liver metastases as well as the degree of liver involvement are crucial to assess both prognosis and therapeutic management particularly the choice of a liver locoregional therapy like surgery, embolization, or radiofrequency. 2-5 With conventional imaging, early arterial phase imaging is important for the detection of hepatic metastases because they are highly vascularized. 13,14 In a previous computed tomography (CT) imaging study of hepatic metastases from ET, Paulson et al 13 found 30% more metastatic lesions during the arterial phase including 6% of hepatic metastasis patients exclusively evidenced on the hepatic arterial phase. We found similar results with magnetic resonance imaging (MRI) in 37 patients with liver metastases from ET. A typical hypervascular pattern was found in 73% of patients and the hepatic arterial phase MRIs revealed the greatest number of metastases in 70% of patients. 15 Standardization of conventional imaging is lacking in reported comparisons of SRS and conventional imaging techniques used to stage ET. The objective of our singlecenter prospective study was to compare the respective sensitivity of CT, MRI, and SRS in the detection of liver metastases from WDGEP ET and to define parameters influencing the sensitivity of SRS. PATIENTS AND METHODS Patients From September 2000 to October 2002, 66 consecutive patients with an endocrine tumor with or without known liver metastases were enrolled in this prospective study after referral to our institute. Inclusion criteria were confirmed WDGEP ET pathological diagnosis in our institution and absence of contraindication for CT or MRI. The liver was screened for metastases by means of CT, MRI, and SRS within a 15-day interval with randomization of the order of the three imaging modalities to avoid bias because of time effect or imaging technique interferences. Our institutional review board approved the study, and all patients gave their informed consent. MRI could not be performed in one patient because the patient did not turn up for the appointment, and CT was performed without contrast injection in one patient because of acute renal failure. These two patients were subsequently excluded from the study; therefore, a total of 64 patients (32 females and 32 males with a mean age of 57 years; range, 15 to 90 years), with a diagnosis of WDGEP ET, and who were studied with the three imaging techniques, constituted our study group. Imaging Techniques CT examinations was performed with a HiSpeed spiral scanner (GE Medical Systems, Milwaukee, WI). Spiral CT images were obtained before and after a monophasic injection of 100 ml of nonionic contrast material (Xenetix 300, Guerbet, France). A CT injector (Medrad, Pittsburgh, PA) was used to deliver contrast medium via a catheter inserted into the antecubital vein at a rate of 3 ml/sec. After injection of contrast material, three spiral CT acquisitions were obtained during the hepatic arterial phase, the portal venous phase, and the equilibrium phase at 30, 70, and 300 seconds, respectively, after the initiation of the injection. Scanning was performed at 120 kv and 270 ma. Contiguously reconstructed sections (pitch of 1:1) were obtained through the liver with 5-mm collimation. Each spiral CT acquisition through the liver was accomplished during a breath-hold. MRI examinations were performed with a 1.5-T whole-body imager (Signa LX; General Electric Medical Systems, Milwaukee, WI). All MRIs were acquired in the axial plane with a phased-array body multicoil. Slice thickness was 7 mm, with a 2-mm intersection gap for all pulse sequences. Fat-suppressed T2-weighted imaging was obtained with a respiratory-triggered fast spin-echo sequence (4,000 to 8,000/102): effective TR/effective TE, 16 echo train length, four signals acquired, 10-millisecond interecho spacing, matrix, khz received bandwidth, 280 to 400 mm field of view, 20% respiratory trigger point, 40% trigger window, and gradient moment nulling in the frequency-encoding direction. Saturation bands above and below the imaging volume were used to attenuate flow-related artifacts throughout MRI. Breath-hold single-shot fast spin-echo sequences were obtained using the following parameters: /90 (effective TR/effective TE), half-fourier acquisition, 104 echo train length, one signal acquired, matrix, 62.5 khz received bandwidth, and 40-cm field of view. Dynamic contrast-enhanced MRIs were obtained with an initial delay of 10 seconds and then at three consecutive 30-second intervals and 5 minutes after the initiation of a bolus injection of 0.1 mmol/kg of Gd-DOTA (Dotarem, Guerbet, France) into the antecubital vein. T1-weighted sequences were acquired with fast multiplanar spoiled gradient-recalled echo imaging (125/4.2) TR/TE, 60 degrees flip angle, one signal acquired, matrix, 62.5 khz received bandwidth, 40-cm field of view, and 25-second acquisition time in a single breath-hold. SRS was performed after intravenous injection of Indium-111- DTPA-Phe1-octreotide (pentetreotide; OctreoScan; Mallinckrodt Medical, Petten, the Netherlands; 170 to 220 MBq). Using the dedicated kit containing the In-111 chloride solution and a special needle, quality control revealed a labeling yield of more than 98% for this radiopharmaceutical. To reduce digestive artifacts, an adequate colonic preparation was administered (64 g of macrogol 4000 in the evening after the injection and again the next morning before 24-hour imaging). A large field-of-view gamma camera equipped with a medium-energy collimator was used (Axis double detector gamma camera; Philips Medical Systems, Best, the Netherlands). Acquisition was performed using both 111 In photopeaks (171 and 245 kev). As recommended, 16 static anterior and posterior spot views covering the whole body ( word matrix, at least 10 minutes per view or 300,000 preset counts for the head and neck and 500,000 for the rest of the body), were acquired at 4 hours, 24 hours, and when needed, at

3 Dromain et al hours. Abdominal single-photon emission computed tomography (SPECT) was performed at 24 hours with 64 projections ( word matrix, 1 minute per projection) and iterative reconstruction. An additional thoracic SPECT was performed when necessary. Image Analysis Images of each modality were analyzed by two readers (radiologists and nuclear physicians) per modality, blindly and independently. CTs and MRIs were all interpreted on a digital workstation (Pathspeed; General Electric Medical Systems, Milwaukee, WI), with total freedom for window and level adjustments and for the magnification of each image at the time of the analysis. SRS images were analyzed on films. Radiologists and nuclear physicians were unaware of clinical and biological findings or of any imaging studies concerning the patients. They knew, however, that all patients had an endocrine tumor and were being examined to evaluate the presence and extent of hepatic metastases. The six readers scored the number of metastases depicted at each examination. Only presumed metastatic lesions for which the confidence rate was high were taken into account. A patient was considered positive for one liver imaging modality when at least one liver lesion meeting the malignancy criteria was depicted in agreement with both observers and at least one typical lesion was discovered on biopsy to confirm the WDGEP ET liver metastasis diagnosis. A patient was considered negative when no liver lesion meeting the malignancy criteria was depicted in agreement with both observers. In this study, the terms positive and negative refer to the liver status. All images of each imaging modality were taken into account to establish the presence or absence and the number of hepatic metastases. When more than 15 metastases were depicted, 16 were considered for the statistical analysis. The final number of metastases for each modality, determined by the consensus of two readers, was obtained during a second reading session. The size, location (made segment by segment according to the Couinaud numbering system with drawing), and the pattern of enhancement of each metastasis were also noted for CT and MRI. For SRS, only the total number of lesions was taken into account. All patients with presumed liver metastases detected with at least one imaging modality experienced either a biopsy (n 32) or a surgery (n 8). To assess a relative sensitivity of each imaging modality: (1) we arbitrarily set the total number of liver metastases by summing the number of metastases depicted for each patient by the highest sensitive modality (reference number), (2) we calculated the ratio between the total number of metastases determined by each modality to the reference number. The extrahepatic tracer uptake was also recorded for SRS. A specific analysis of the group of patients treated with the somatostatin analog at the time of the imaging protocol was also performed. Statistical Analysis To assess interobserver agreement in interpreting images, the statistic was used to measure the degree of agreement between the two observers for each imaging technique. A value attaining 0.40 signified positive but poor agreement, a value of 0.41 to 0.75 indicated good agreement, and a value exceeding 0.75 indicated excellent agreement. Wilcoxon s signed-rank test was used for the statistical analysis of the number of metastases detected by each imaging technique and a P value of less than.05 was considered significant. To define parameters influencing the relative sensitivity of SRS, two groups of patients were compared using the 2 test: patients in the high-sensitivity SRS group had a difference of less than 5 in the number of liver metastases detected with SRS and the best imaging modality whereas patients in the low-sensitivity SRS group had false-negative SRS results or a difference of 5 or more liver metastases detected with SRS and the best imaging modality. Parameters taken into account for this analysis were sex and age, the primary tumor site, the absence or presence of hormonal secretion, previous surgical removal of the primary tumor, the median size of metastases on the morphological examination that detected the higher number of metastases, and the pattern of enhancement of hepatic metastases on CT and MRI. RESULTS Patients Patients characteristics and results are summarized in Table 1. Primary ET sites were the small bowel (n 21), pancreas (n 15), lung (n 11), colon and rectum (n 3), unknown (n 5), appendix (n 3), thymus (n 2), duodenum (n 2), larynx (n 1), and esophagus (n 1). Multiple endocrine neoplasia (MEN) type 1 was present in two patients with a pancreatic tumor. Hormonal secretion was present in 37 (59%) patients, including four clinically functioning pancreatic ET (one insulinoma, one gastrinoma, two glucagonoma). Fourteen patients were screened for the initial diagnosis and staging of the primary tumor whereas 50 patients were being investigated during follow-up. Twenty-six patients had previously received treatment including systemic chemotherapy (n 19), somatostatin analogs (n 5), and interferon (n 1). Only 13 patients were treated at the time of the imaging protocol with somatostatin analogs in 12 patients and chemotherapy in one patient. Interobserver Agreement Interobserver agreement was excellent for MRI and CT with a value of 84% and 78%, respectively, and was good for SRS with a value of 69%. SRS discrepant results between observers refer to the number of liver metastases at abdominal SPECT in all patients but one. Liver Imaging Results Of the 64 consecutive patients with a diagnosis of ET eligible in the study, 40 patients (62.5%) had hepatic metastasis detected by at least one imaging modality (Fig 1). The total number of metastases detected on SRS, CT, and MRI was 204, 325, and 394 in 33, 39, and 39 patients, respectively (Table 2). MRI detected significantly more liver metastases than CT (P.02) and SRS (P 10 4 ), and CT detected significantly more liver metastases than SRS (P 10 4 ). The relative sensitivities of SRS, CT, and MRI were 49.3%, 78.5%, and 95.2%, respectively. Based on a per-patient analysis, the maximum number of metastases was found with MRI in 14 patients (35%), with CT in six patients (15%), and with SRS in one patient (2.5%). When imaging modalities were analyzed, CT was negative in two patients (5%) with a positive SRS and in one patient (2.5%) with a positive MRI (Fig 2). MRI was 72 JOURNAL OF CLINICAL ONCOLOGY

4 Detection of Liver Metastases Patient No. Table 1. Individual Characteristics of 64 Patients With Well-Differentiated Endocrine Tumors Age (years) Sex Primary Hormonal Secretion Liver Metastases No. of Metastases SRS CT MRI 1 75 F Lung 0 Y M Unknown CgA, calcitonin Y M Pancreas Calcitonin N F Ileum 5HIAA Y F Pancreas Insulin Y M Thymus ACTH, CgA Y F Unknown 0 Y F Unknown CgA N M Lung 0 N F Ileum 0 Y F Duodenum CgA Y F Larynx Calcitonin N F Rectum 0 Y M Ileum 0 N M Appendix 0 N M Pancreas CgA N M Pancreas Gastrin Y M Ileum 5HIAA Y F Ileum 0 N F Ileum 5HIAA Y M Ileum 5HIAA, CgA Y M Pancreas CgA N F Colon 0 Y M Ileum 0 Y F Ileum CgA Y F Lung 0 Y M Pancreas CgA N M Lung CgA, 5HIAA Y F Ileum 5HIAA Y M Unknown 0 N F Pancreas Glucagon Y M Lung 0 N F Duodenum 0 N F Ileum 0 Y F Appendix 0 N F Ileum CgA, 5HIAA Y M Pancreas 0 Y M Lung CgA, 5HIAA Y M Ileum CgA, 5HIAA Y F Ileum CgA, 5HIAA Y M Ileum CgA, 5HIAA N F Ileum CgA N F Ileum 0 Y F Pancreas 0 Y F Pancreas CgA Y F Lung CgA, 5HIAA Y F Lung 0 N F Ileum CgA N M Pancreas 0 N M Pancreas 0 Y M Lung Calcitonin, CgA Y F Unknown CgA Y M Pancreas 0 Y F Ileum 5HIAA Y M Esophagus 5HIAA N M Pancreas Glucagon, CgA Y F Rectum 0 N M Ileum 5HIAA Y M Ileum CgA, 5HIAA Y M Appendix 0 N F Lung 0 N M Thymus CgA N M Pancreas Calcitonin, CgA Y M Lung 0 Y Abbreviations: SRS, somatostatin receptor scintigraphy; CT, computed tomography MRI, magnetic resonance imaging; F, female; M, male; Y, yes; N, no; CgA, Chromogranin A; 5HIAA, 5 hydoxyindolacetic acid; ACTH, corticotropin. When more than 15 metastases were depicted, 16 were considered for the statistical analysis. 73

5 Dromain et al Fig 1. Liver screening in a 50-year-old man with an ileum endocrine tumor. (A) Enhanced computed tomography, (B) unenhanced T2-weighted magnetic resonance imaging (MRI), and (C) arterial phase-enhanced T1-weighted MRIs demonstrated a typical pattern of hypervascular metastases, also well shown on (D) axial tomographies of somatostatin receptor scintigraphy. negative in a patient with a positive SRS. SRS failed to detect hepatic metastases in seven patients (17.5%) in whom MRI or CT was positive for hepatic metastases. In the 12 patients treated with somatostatin analogs, SRS was positive in nine with positive CT and MRI results and negative in three patients with negative CT and MRI results. The number of liver metastases detected in the nine patients with positive findings was 55, 94, and 94 on SRS, CT, and MRI, respectively. None of these 12 patients experienced an objective response with somatostatin analog therapy. Table 2. Number of Metastases According to Imaging Modalities Reviewer 1 Reviewer 2 Consensus CT MRI SRS Abbreviations: CT, computed tomography; MRI, magnetic resonance imaging; SRS, somatostatin receptor scintigraphy. The median size of depicted metastases was 12 mm for CT and 10.5 for MRI. Metastases were hypervascular in 30 patients (77%) and hypovascular in nine patients (23%) at both CT and MR imaging. A miliary pattern with numerous small hepatic lesions scattered throughout the liver was present in nine patients (22.5%; Fig 3). Among the 204 metastases detected on SRS, 190 were detected on static anterior and posterior views whereas 203 metastases were detected on abdominal SPECT. Abdominal SPECT increased the number of metastases detected with SRS in 10 patients in whom 13 additional lesions had not been detected with the static views. SRS also detected extrahepatic tracer uptake in 32 patients (51%). Predictors of High-Sensitivity SRS The median size of metastases was the only factor significantly correlated with the sensitivity of SRS for the detection of liver metastases (P.04). Indeed, a positive SRS classified as a high-sensitivity SRS result was found in 22% of patients with a median metastasis size of less than 7 mm, in 35% of patients with a median metastasis size of 8 to 14 mm, and 64% of patients with a median metastasis size of 74 JOURNAL OF CLINICAL ONCOLOGY

6 Detection of Liver Metastases Fig 2. Liver screening in a 29-year-old woman with pancreatic insulinoma. (A) Arterial and (B) portal phase-enhanced computed tomography images show no obvious metastases whereas (C and D) arterial phase-enhanced magnetic resonance imaging shows several hypervascular metastases. (E) Somatostatin receptor scintigraphy planar view is falsely negative. more than 15 mm (Table 3). A trend was seen toward better sensitivity of SRS in the detection of liver metastases when the primary tumor was located in the ileum compared with primary tumors located in the lung but the difference was not statistically significant (Table 4). No significant difference was found in the sensitivity of SRS between metastases 75

7 Dromain et al Fig 3. Liver screening in a 70-year-old man with an unknown endocrine tumor. (A) Arterial phase-enhanced computed tomography image shows minute contrast uptake disappearing on (B) portal phase image. (C) T2-weighted magnetic resonance imaging demonstrates metastatic hepatic miliary well. (D) Axial tomographies of somatostatin receptor scintigraphy was falsely negative. classified as hypovascular and hypervascular nor between biologically functional and nonfunctional tumors. DISCUSSION Controversy persists regarding the most accurate diagnostic method for the detection of hepatic metastases from ET 10,11,16,17 To our knowledge, our study is the first to compare prospectively three standardized imaging modalities for the detection of WDGEP ET metastases to the liver. Indeed, in most studies comparing conventional imaging and SRS, the parameters of the imaging technique used were not specified, namely the performance of either CT or MRI during the arterial phase after injection of contrast medium. Moreover, in our study, each imaging modality was performed within a short time period, according to the Table 3. Correlation Between the Sensitivity of SRS and the Median Size of Metastases Median Size of Metastases High- Sensitivity SRS Low- Sensitivity SRS Total No. % No. % No. % 7 mm mm mm Total Abbreviation: SRS, somatostatin receptor scintigraphy. Table 4. Correlation Between the Sensitivity of SRS and the Primary Tumor Site Primary Tumor High- Sensitivity SRS Low- Sensitivity SRS Total No. % No. % No. % Lung Pancreas-duodenum Ileum Miscellaneous Total Abbreviation: SRS, somatostatin receptor scintigraphy. 76 JOURNAL OF CLINICAL ONCOLOGY

8 Detection of Liver Metastases same protocol for each patient and was in keeping with current methodological recommendations. The results of this study highlight the superiority of MRI over CT and SRS in the detection of hepatic metastases from ET. MRI detected 190 hepatic metastases missed by SRS and 69 missed by CT. Based on a per-patient analysis, liver uptake was negative in 17.5% of patients with SRS whereas CT and MRI results were positive. CT and MRI results were negative when SRS was positive in only one patient with one hepatic metastasis. MRI depicted a far greater number of liver metastases than CT despite its lower spatial resolution and even when liver metastases exhibited a miliary pattern. This is probably due to the high-contrast resolution of MRI, especially on T2- weighted images and on the hepatic arterial phase enhanced T1-weighted images. 15 Standardization of imaging techniques is a key point of the sensibility, of both CT and MR imaging, in the detection of liver metastases from WDGEP ET. Indeed, as already mentioned, the arterial phase plays a major role to standardize the detection of endocrine liver metastases and must be systematically performed during CT and MR imaging in both initial and follow-up screening. On MRI, we also recommend performance of two T2-weighted sequences: one with a high lesion-to-liver contrast-noise ratio for solid lesion (fast spin-echo sequence) and another with high contrast for tissues with a long T2 relaxation time (single-shot sequence). The first has a high sensibility in the detection of liver metastases, whereas the second has a high specificity allowing optimal distinction between hypervascular metastases from hemangiomas. 15 Therefore, we advocate liver MRI as a critical tool in patients with WDGEP ET. Our practical recommendation is to systematically perform a liver MRI, and not a CT, in the initial and preoperative screening of metastatic spread in WDGEP ET patients. MRI plays a major role in detecting liver metastases whereas SRS may play a complementary role in detecting extrahepatic disease as claimed by other studies It is also our personal experience that liver MRI should be performed during a patient s follow-up examination. Indeed, unenhanced MRIs such as unenhanced T1- weighted and unenhanced T2-weighted images allow a good detection and delineation of liver metastases and have the advantage over enhanced CT or MR images to be reproducible with time because they are not influenced by the uptake of contrast agent, variable with the physiological conditions. Taking into account the cost of MRI compared with CT, however, a practical recommendation during follow-up would be to evaluate the patient with WDGEP ET liver with both techniques first and then to choose the most accurate. These results are at variance with one study comparing MRI and CT in the detection of liver metastases in 24 patients with gastrinoma. 10 In that study, SRS was the most sensitive liver imaging method compared with any of the conventional imaging methods. SRS results found in gastrinoma patients, therefore, may not be applicable to other gastroenteropancreatic ET. The high frequency of positive subtype 2 somatostatin immunoreactivity in gastrinoma compared with other gastroenteropancreatic tumors may partly explain these discrepancies. 18 The only parameter significantly associated with high-sensitivity SRS was the median size of metastases. This result has already been reported for the detection of primary gastrinoma arising in the duodenum and pancreas. 19 The size effect could be caused by the amount of radionuclide uptake required for detection, as suggested when different radionuclides were used. 19 The type or density of the somatostatin receptors expressed by the tumor may be a predominant additional factor. The recent somatostatin analog radionuclide with improved subtype 2 receptor affinity but also great internalization properties may overcome these drawbacks. 20,21 These new somatostatin radiolabeled compounds, tailored for positron imaging technology, hold great promise for an enhancement of the sensitivity of SRS Conversely, 2-[fluorine-18]-fluoro-2-deoxy-D-glucose (FDG)-positron emission tomography is of limited value for the imaging of well-differentiated endocrine tumors. 25 With regard to SRS acquisition methods and in agreement with previous reports, our study confirms that SPECT is more sensitive for the detection of focal lesions than planar imaging. 17 SPECT is endowed with the capacity to display cross-sectional images and has better contrast resolution than planar imaging, which accounts for its enhanced sensitivity. Despite the poor performance of SRS in the present study, its specificity is reported to be high at 88%. 12 Furthermore, SRS explores the total body and could be useful for the selection of patients for somatostatin analogs, radiolabeled somatostatin analog therapy, or probe-guided surgery. Finally, contrary to previous studies where SRS showed a greater likelihood of being positive in patients with elevated 5HIAA, we found no correlation between the performance of SRS and the presence of hormonal secretion 26 There are, however, several limitations in our study. First, although ET-derived liver metastases were pathologically proven in all patients, a detailed lesion-bylesion pathological analysis was not possible because most of the patients in this series had advanced metastatic ET to the liver and, therefore, were not submitted to hepatic resection. This must be taken into account in the interpretation of the relative sensitivities of the three imaging techniques. A second potential limitation is that 12 patients were treated during the imaging protocol with somatostatin analogs that could have affected the performance of SRS. Indeed, it has been suggested that somatostatin analogs may result in low or no uptake of the radioligand because of occupancy, competition, or 77

9 Dromain et al downregulation of the receptors by the unlabeled ligand. However, subsequent studies have shown that somatostatin analog treatment modifies the biodistribution of 111 Inpentetreotide, significantly increasing the tumor-tobackground ratio In our study, no patient treated with somatostatin analogs at the time of imaging had positive CT and MRI when the SRS result was negative. The number of liver metastases detected by SRS in this group of treated patients was lower than the number of metastases detected by CT and MRI but with a similar difference to that found in the group of untreated patients. Thus, somatostatin analog therapy does not seem to have had a great influence on the performance of SRS. Finally, specificity could not be evaluated because a liver biopsy was not performed in patients with negative liver imaging. In conclusion, MRI depicted by far the greatest number of hepatic metastases in patients with WDGEP ET. This imaging modality, therefore, appears to be the procedure of choice for initial staging and liver evaluation before surgery, embolization, or radiofrequency in patients with ET. The low performance of SRS was mainly explained by the effect of the size of liver metastases on its detection capacity. Acknowledgment We thank Lorna Saint Ange for editing. Authors Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. REFERENCES 1. Solcia EKG, Sobin LH. Histological typing of endocrine tomours, in Springer (ed): WHO Classification, International Histological Classification of Tumours (ed 2). Berlin, Germany, Springer, Madeira I, Terris B, Voss M, et al: Prognostic factors in patients with endocrine tumours of the duodenopancreatic area. Gut 43: , Janson ET, Holmberg L, Stridsberg M, et al: Carcinoid tumors: Analysis of prognostic factors and survival in 301 patients from a referral center. Ann Oncol 8: , Rindi G, Bordi C, Rappel S, La Rosa S, Stolte M, Solcia E: Gastric carcinoids and neuroendocrine carcinomas: Pathogenesis, pathology, and behavior. World J Surg 20: , Yu F, Venzon DJ, Serrano J, et al: Prospective study of the clinical course, prognostic factors, causes of death, and survival in patients with long-standing Zollinger-Ellison syndrome. J Clin Oncol 17: , Hofland LJ, Lamberts SW, van Hagen PM, et al: Crucial role for somatostatin receptor subtype 2 in determining the uptake of [111In-DTPA-D- Phe1]octreotide in somatostatin receptor-positive organs. J Nucl Med 44: , Westlin JE, Janson ET, Arnberg H, Ahlstrom H, Oberg K, Nilsson S: Somatostatin receptor scintigraphy of carcinoid tumours using the [111In- DTPA-D-Phe1]-octreotide. Acta Oncol 32: , Scherubl H, Bader M, Fett U, et al: Somatostatin-receptor imaging of neuroendocrine gastroenteropancreatic tumors. Gastroenterology 105: , Jamar F, Fiasse R, Leners N, Pauwels S: Somatostatin receptor imaging with indium-111- pentetreotide in gastroenteropancreatic neuroendocrine tumors: Safety, efficacy and impact on patient management. J Nucl Med 36: , Gibril F, Reynolds JC, Doppman JL, et al: Somatostatin receptor scintigraphy: Its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas A prospective study. Ann Intern Med 125: 26-34, Chiti A, Fanti S, Savelli G, et al: Comparison of somatostatin receptor imaging, computed tomography and ultrasound in the clinical management of neuroendocrine gastro-enteropancreatic tumours. Eur J Nucl Med 25: , Gibril F, Reynolds JC, Chen CC, et al: Specificity of somatostatin receptor scintigraphy: A prospective study and effects of false-positive localizations on management in patients with gastrinomas. J Nucl Med 40: , Paulson EK, McDermott VG, Keogan MT, DeLong DM, Frederick MG, Nelson RC: Carcinoid metastases to the liver: Role of triple-phase helical CT. Radiology 206: , Foley WD, Mallisee TA, Hohenwalter MD, Wilson CR, Quiroz FA, Taylor AJ: Multiphase hepatic CT with a multirow detector CT scanner. AJR Am J Roentgenol 175: , Dromain C, de Baere T, Baudin E, et al: MR imaging of hepatic metastases caused by neuroendocrine tumors: Comparing four techniques. AJR Am J Roentgenol 180: , Kwekkeboom DJ, Krenning EP: Somatostatin receptor scintigraphy in patients with carcinoid tumors. World J Surg 20: , Schillaci O, Spanu A, Scopinaro F, et al: Somatostatin receptor scintigraphy in liver metastasis detection from gastroenteropancreatic neuroendocrine tumors. J Nucl Med 44: , Kulaksiz H, Eissele R, Rossler D, et al: Identification of somatostatin receptor subtypes 1, 2A, 3, and 5 in neuroendocrine tumours with subtype specific antibodies. Gut 50:52-60, Alexander HR, Fraker DL, Norton JA, et al: Prospective study of somatostatin receptor scintigraphy and its effect on operative outcome in patients with Zollinger-Ellison syndrome. Ann Surg 228: , de Jong M, Breeman WA, Bakker WH, et al: Comparison of (111)In-labeled somatostatin analogues for tumor scintigraphy and radionuclide therapy. Cancer Res 58: , Reubi JC, Schar JC, Waser B, et al: Affinity profiles for human somatostatin receptor subtypes SST1-SST5 of somatostatin radiotracers selected for scintigraphic and radiotherapeutic use. Eur J Nucl Med 27: , Hofmann M, Maecke H, Borner R, et al: Biokinetics and imaging with the somatostatin receptor PET radioligand (68)Ga-DOTATOC: Preliminary data. Eur J Nucl Med 28: , Kwekkeboom DJ, Bakker WH, Kam BL, et al: Treatment of patients with gastroentero-pancreatic (GEP) tumours with the novel radiolabelled somatostatin analogue [(177)Lu-DOTA(0),Tyr(3)]octreotate. Eur J Nucl Med Mol Imaging 30: , Wester HJ, Schottelius M, Scheidhauer K, et al: PET imaging of somatostatin receptors: Design, synthesis and preclinical evaluation of a novel 18F-labelled, carbohydrated analogue of octreotide. Eur J Nucl Med Mol Imaging 30: , Adams S, Baum R, Rink T, Schumm- Drager PM, Usadel KH, Hor G: Limited value of fluorine-18 fluorodeoxyglucose positron emission tomography for the imaging of neuroendocrine tumours. Eur J Nucl Med 25:79-83, Kalkner KM, Janson ET, Nilsson S, Carlsson S, Oberg K, Westlin JE. Somatostatin receptor scintigraphy in patients with carcinoid tumors: Comparison between radioligand uptake and tumor markers. Cancer Res 55: , 1995 (suppl) 27. Dorr U, Rath U, Sautter-Bihl ML, et al: Improved visualization of carcinoid liver metastases by indium-111 pentetreotide scintigraphy following treatment with cold somatostatin analogue. Eur J Nucl Med 20: , Janson ET, Kalkner KM, Eriksson B, Westlin JE, Oberg K: Somatostatin receptor scintigraphy during treatment with lanreotide in patients with neuroendocrine tumors. Nucl Med Biol 26: , Leners N, Jamar F, Fiasse R, Ferrant A, Pauwels S: Indium-111-pentetreotide uptake in endocrine tumors and lymphoma. J Nucl Med 37: , Hofland LJ, van Koetsveld PM, Waaijers M, Zuyderwijk J, Breeman WA, Lamberts SW: Internalization of the radioiodinated somatostatin analog [125I-Tyr3]octreotide by mouse and human pituitary tumor cells: Increase by unlabeled octreotide. Endocrinology 136: , JOURNAL OF CLINICAL ONCOLOGY