MRI for hepatocellular carcinoma: a primer for magnetic resonance imaging interpretation.

Hepatocellular carcinoma is among the most prevalent solid organ cancers, and, unlike many cancers, may be diagnosed non-invasively by imaging criteria [1] with the preferred modality recently shifting from multiphasic computed tomography (MDCT) to magnetic resonance imaging (MRI). The purpose of this article is to help facilitate radiologists and radiology trainees in the transition to MRI by providing a step-wise approach to exam interpretation to improve the MRI detection of HCC. A methodical, consistent approach to navigating a HCC screening MRI exam, in conjunction with the LI-RADS framework for characterization, should lead to improved HCC detection and diagnosis.

Diagnostic Performance of Contrast-Enhanced MRI With Secretin-Stimulated MRCP for Non-Calcific Chronic Pancreatitis: A Comparison With Histopathology.

OBJECTIVES: Diagnosis of non-calcific chronic pancreatitis (NCCP) in patients presenting with chronic abdominal pain is challenging and controversial. Contrast-enhanced magnetic resonance imaging (MRI) with secretin-stimulated MRCP (sMRCP) offers a safe and noninvasive modality to diagnose mild CP, but its findings have not been correlated with histopathology. We aimed to assess the correlation of a spectrum of MRI/sMRCP findings with surgical histopathology in a cohort of NCCP patients undergoing total pancreatectomy with islet autotransplantation (TPIAT). METHODS: Adult patients undergoing TPIAT for NCCP between 2008 and 2013 were identified from our institution's surgery database and were included if they had MRI/sMRCP within a year before surgery. Histology was obtained from resected pancreas at the time of TPIAT by wedge biopsy of head, body, and tail, and was graded by a gastrointestinal pathologist who was blinded to the imaging features. A fibrosis score (FS) of 2 or more was considered as abnormal, with FS ≥6 as severe fibrosis. A multivariate regression analysis was performed for MRI features predicting fibrosis, after taking age, sex, smoking, alcohol, and body mass index (BMI) into consideration. A quantitative receiver operating characteristic (ROC) curve analysis was performed and Spearman rank correlation coefficient (r) was calculated. RESULTS: Fifty-seven patients (females=49, males=8) with NCCP and MRI/sMRCP were identified. ROC curve analysis showed that two or more MRI/sMRCP features provided the best balance of sensitivity (65%), specificity (89%), and accuracy (68%) to differentiate abnormal (FS≥2) from normal pancreatic tissue. Two or more features provided the best cutoff (sensitivity 88%, specificity 78%) for predicting severe fibrosis (FS≥6). There was a significant correlation between the number of features and severity of fibrosis (r=0.6, P<0.0001). A linear regression after taking age, smoking, and BMI into consideration showed that main pancreatic duct irregularity, T1-weighted signal intensity ratio between pancreas and paraspinal muscle, and duodenal filling after secretin injection to be significant independent predictors of fibrosis. CONCLUSIONS: A strong correlation exists between MRI/sMRCP findings and histopathology of NCCP.