Intrapancreatic accessory spleen mimicking a neuroendocrine tumor: magnetic resonance findings and possible diagnostic role of different nuclear medicine tests

Abstract

Accepted: 3 September 2003 Published online: 13 November 2003 © Springer-Verlag 2003 Sir, A 52-year-old woman was referred to our Department of Radiology to investigate ultrasonographic diagnosis of abscess of the abdominal wall with MR examination. She had had appendicectomy 30 years previously and successive resections of the surgical scar with histological result of granulomatosis. At MR examination a nodule in the pancreatic tail was identified. This nodule was mildly hyperintense on T2-weighted fast-spin-echo sequences, hypointense on FSPGR T1-weighted sequences and showed conspicuous enhancement in the arterial phase during infusion of gadolinium (Fig. 1). It was round, with regular margins and had 1-cm maximum diameter. Laboratory data, tumor markers, serum insulin, gastrin, and calcitonin levels were normal. Signal intensity and contrast enhancement pattern oriented for nonfunctional neuroendocrine lesion. Indium-111 octreoscan revealed a correspondent focal tracer uptake in the pancreatic tail (Fig. 2). After distal pancreatectomy, the specimen revealed intrapancreatic splenic tissue. Islet cell tumors of the pancreas have a reported prevalence of less than 1 in 200,000 [1]; 95% of them are functioning but one-third are symptom free. The most common tumor is insulinoma (60–75%), followed by gastrinoma (20%) [1]. At MRI, neuroendocrine tumors are iso-hyperintense on T2-weighted sequences, hypointense on T1-weighted sequences, and show strong arterial enhancement during contrast injection as in CT [1]. Somatostatin receptors scintigraphy (Octreoscan) has high sensitivity (70–95%) in detection of gastrointestinal neuroendocrine tumors [2]. Splenic tissue shows a physiological uptake of 111In Octreoscan. It is probably due to the presence of specific somatostatin receptors on the surface of lymphocytes [2]; therefore, this can justify tracer uptake in an accessory intrapancreatic spleen, as in our case. One or more accessory spleens are present in approximately 10% of the population [3]. It is frequent to see accessory spleens near the splenic hilum, along the course of splenic vessels. Although intrapancreatic spleens are rarely depicted with imaging, they are not uncommon. In fact, in autoptical series the pancreatic gland is the second most common site of accessory spleens; 10–30% of them are localized within or near the pancreatic tail [3]. In our case the accessory spleen was unique, small (1 cm), and located deep inside the pancreatic tail, surrounded by normal pancreatic parenchyma. Our incorrect diagnosis was due to dimension, position, and singleEur Radiol (2004) 14:1322–1323 DOI 10.1007/s00330-003-2112-4 L E T T E R T O T H E E D I T O R