Neuroendocrine Tumor of the Pancreas : Biphasic CT versus MR Imaging in Tumor Detection

Introduction; Neuroendocrine tumors (islet cell tumors) are relatively rare pancreatic ne0piasms.k is clinically important that they have a high rate of malignancy with an incidence ranged from 60 to 92 % .The common noninvasive preoperative radiological approach for evaluating neuroendocrine tumors are represented by dynamic contrast-enhanced CT and MR imaging .As several previous studies indicated efficacy of biphasic CT in the pancreas, the recent consensus is that biphasic CT is the most sensitive method for detecting neuroendocrine tumors (1,2). While, several investigators have also emphasized advantages of MR imaging over CT, which are based on higher contrast resolution or higher sensitivity for contrast material enhancement of MR imaging, in the investigation of neuroendocrine tumors (3). And they have concluded that MR imaging is a promising tool for the detectability of neuroendocrine tumorsAt present, to our best knowledge, there is no report in which the capabilities of biphasic CT and MR imaging are compared each other in detecting neuroendocrine tumors, which is an interesting and challenging issue. Therefore, the purpose of this study is to compare the efficacy between biphasic CT and MR imaging with various pulse sequences for the detection of neuroendocrine tumors. Material and methods;1 9 patients (9 men and 10 women, age range from 40 to 69 years; mean 53 years) with 26 lesions of pathologically proved neuroendocrine tumors of the pancreas were retrospectively examined. All MR imaging examinations were performed with 1.5-T MR units (Signa Advantage or Horizon; GE Medical System) As this study was a retrospective study, there was moderate variability in MR units and sequence parameters that usually involved pulse sequence. All patients underwent spin-echo Tl and spin-echo T2 (15 patients) or fast spin-echo T2-weighted (4 patients) MR images. Additional MR examinations were also obtained unenhanced fat-saturated Tl-weighted MR images (14 lesions in 10 patients). Dynamic contrast-enhanced gradient-echo MR images (16 lesions in 13 patients) was performed with 0.1 mmol/kg gadolinium and followed by delayed-enhanced fatsaturated Tl-weighted MR images in 5-10 min (19 lesions in 17 patients). All patients underwent biphasic CT including unenhanced, arterial-dominant phase (AP; 20-28 set) and portal venous phase (PVP; 60-70 set) images. Contrast material was administrated at rates of 3.0-5.0 mUsec. The slice thickness was 7 mm and the table-incremental speed was 7 mm/set (1:l to 1.5 pitch). Analvsis;.l) The image quality of the all CT and MR; Two reviewers worked separately to evaluate the image quality of the all CT and MR images with each technique,to investigate whether the image quality may affect the results for the lesion detectability with each technique, A five-point (5-l) grading system (5= excellent, 4= marginally adequate, 3= good, 2= poor, l= nondiagnostic) was used assigning for the conspicuity of the pancreas. 2) Receiver Operating characteristic (ROC) analysis; ROC analysis was also employed independently and in random order by three