Evaluation of somatostatin-receptor scintigraphy for detecting neuroendocrine tumors.

BACKGROUND: Somatostatin-receptor scintigraphy (SRS) has gained attention as an imaging modality for neuroendocrine tumors (NETs). The purpose of this study was to present one of the first American series evaluating the ability of SRS to detect local and distant disease caused by NETs. METHODS: Medical records were reviewed from 35 patients who underwent a total of 38 studies using 111In-pentetreotide between 1993 and 1995. Twenty-two patients had islet cell tumors, seven had carcinoid tumors, and six had other NETs. RESULTS: The overall sensitivity of SRS was 74% for detecting local disease (primary tumor +/- regional lymph node metastases) in all NETs, excluding insulinoma, 75% in gastrinoma, 0% in insulinoma, 78% in other islet cell tumors, and 50% in carcinoids. For detecting distant disease, the overall sensitivity of SRS was 67% for all NETs, excluding insulinoma, 100% for gastrinoma, 50% for other islet cell tumors, and 80% for carcinoids. Specificity and positive predictive value were 100% for all tumors. Negative predictive value ranged from 33% to 100%. CONCLUSIONS: A positive SRS study strongly predicts the presence of tumor (100% positive predictive value in our study). However, unlike the European reports of very high sensitivity (80% to 88%), we found that SRS had a lower sensitivity (67% for all NETs excluding insulinoma and 71% for noninsulinoma gastroenteropancreatic NETs). Thus negative SRS in patients with NETs must be viewed cautiously, because the false-negative rate is high, and this limits the use of this method in the most difficult patients.