ABSTRACT
Objective: To determine the efficacy of MRI pulse sequences with SENSE coil for depicting insulinoma. Methods: Thirteen patients with clinically suspected insulinoma, documented by pathologic examination of resected specimens, underwent MRI examination. An investigator determined the following data for each tumor: size, location, margination, signal intensity, and contrast enhancement pattern. In addition, the evidence of metastases was assessed. Results: Masses were located throughout the pancreas (head, n=2; neck, n=1; body, n=2; tail, n=6, and uncinate process, n=2). The mean maximal diameter was 13.7 mm (range, from 6 mm. to 30 mm.). The sensitivity of detecting tumor on different MR techniques were 30.7% on in-phase T1-weighted images, 61.5% on T2-weight images with fat suppression technique, 76.9% on pre-contrast 3D-GRE T1-weighted images, and 46.1% on dynamic 3D-GRE Gd images. The tumor detection rate obtained by both T2-weight images with fat suppression technique and pre-contrast 3D-GRE T1-weighted images were highest one (91.7%) with the same sensitivity of detecting tumor by the complete set of MR images. Two patients had metastatic liver disease at presentation. Conclusion: Recent advanced MR imaging is one of the sensitive imaging modalities for detecting the presence of insulinoma and metastatic disease. Combined pre-contrast 3D GRE sequence and T2-weighted images with fat suppression are the essential recent MR protocol.
ABSTRACT
OBJECTIVE: To determine the significance of small arterially-enhancing liver lesions seen with magnetic resonance imaging (MRI) in patients with chronic liver disease. MATERIAL AND METHOD: Our institutional review board approved this retrospective study, without requiring informed consent. Over a two-year period, 258 consecutive patients with cirrhosis or chronic hepatitis underwent multiphase three-dimensional, gadolinium-enhanced, breath-hold gradient-echo MRI. From this group, 29 patients underwent at least one follow-up MR study. When a small (< or = 20 mm) arterially-enhancing lesion was detected, the maximum diameter shape, signal intensities (T1-weighted and T2-weighted), and pattern of enhancement were evaluated to assess the associations between the imaging appearance on initial MR exam and subsequent behavior on follow-up imaging. Statistical testing was performed with JMP Statistical Software (SAS, Inc., Cary, NC) and StatXact 7 Statistical Software for Exact Nonparmetric Inference (Cytel, Inc., Cambride, MA). RESULTS: Sixty-five small (< or = 20 mm) arterially-enhancing lesions were detected in 29 patients. Ten of 65 lesions (15%) in nine patients were subsequently proven to represent hepatocellular carcinoma (HCC), while the remaining lesions either disappeared (46) or remained stable in size (9). Of the 10 lesions subsequently proven to represent HCC, eight lesions converted from hypo- or isointense to hyperintense on subsequent T2-weighted MRI (p < 0.001), seven lesions converted from hyper- or isointense to hypointense on subsequent T1-weighted images (p < 0.001), seven lesions demonstrated growth on subsequent MRI exam (mean increase in mean diameter = 1.4 cm), and five lesions subsequently developed rim enhancement that was not initially present. CONCLUSION: Small, arterially-enhancing lesions detected with MRI have a low likelihood of representing HCC, and MRI follow-up of such lesions is a reasonable approach. Lesions that increase in size, convert to hypointense on subsequent T1W images, convert to hyperintense in T2W images, or develop rim enhancement on follow-up MRI images are concerning and should prompt consideration of intervention.