RESUMEN
PURPOSE: To differentiate pancreatobiliary-type from intestinal-type periampullary carcinomas using combined magnetic resonance cholangiopancreatography (MRCP), contrast-enhanced MRI, and diffusion-weighted imaging (DWI). MATERIALS AND METHODS: MRI (3.0T) results of 41 patients with pathologically confirmed periampullary carcinoma were retrospectively assessed. Two radiologists, blinded to histologic type of each tumor, evaluated image findings independently. MRCP image features, enhancement pattern, and apparent diffusion coefficient (ADC) values were analyzed. Independent-sample t-test, chi-square, or Fisher's exact test were used to determine differential image findings between the pancreatobiliary-type and the intestinal-type group. Cohen's κ statistic or interclass correlation coefficient (ICC) were used to evaluate interobserver agreement between two observers. Univariate and multiple logistic regression analysis were performed to identify MRI features with predictive values. RESULTS: On the basis of hematoxylin-eosin staining, 27 patients were classified as having pancreatobiliary-type carcinomas, and 14 patients the intestinal type. The pancreatobiliary-type carcinomas more commonly showed progressive enhancement than the intestinal type (P = 0.003). The minimum ADC (ADCmin ) value of the pancreatobiliary-type group ([0.95 ± 0.21] × 10(-3) mm(2) /s) was significantly lower than the intestinal-type group ([1.10 ± 0.25] × 10(-3) mm(2) /s) (P = 0.047). For interobserver agreement, the κ values and ICCs for all parameters exceeded 0.8, indicating almost perfect agreement. At multiple logistic regression analysis, the enhancement pattern was the only significant independent predictor (P = 0.011, odds ratio [OR] = 0.105). When the enhancement pattern and ADCmin were used in combination, we could identify 70.4% of pancreatobiliary-type and 78.6% of intestinal-type carcinomas. CONCLUSION: Progressive enhancement and low ADCmin values suggest a pancreatobiliary-type periampullary carcinoma.
