Peripancreatic Vascular Involvement In Acute Pancreatitis: A MRI Study

Objective: To study MRI findings of peripancreatic vascularinvolvement in acute pancreatitis(AP) as well as correlations betweenvascular involvement and the severity of acute pancreatitis according to theAcute Physiology and Chronic Healthy Evaluation Ⅱ(APACHE Ⅱ) scoringsystem and the Magnetic Resonance severity index(MRSI).Materials and Methods: A total of326patients with AP admitted toour institution from August2009to August2013were included in this study.All cases signed informed consent. MRI plain scan and enhanced scan wereperformed within72hours after admission. Among them,122cases of APwere performed GE1.5T MR scanning and204cases were treated withGE3.0T MR scanning. The sequences of MRI scanning includedrespiratory-triggered axial fast recovery fast spin-echo T2-weighted image,axial single shot fast spin-echo T2-weighted image, coronal single shot fastspin-echo T2-weighted image, single shot fast spin-echo magnetic resonancecholangiopancreatography(MRCP), liver acquisition with volume acceleration.MRI findings of acute pancreatitis were noted. According to the magneticresonance severity index score (MRSI), the severity of acute pancreatitis wasgraded as severe (7to10points), moderate (4to6points) and mild (0to3points). APACHE Ⅱ score derived from the clinical data of patients gradedacute pancreatitis as mild(＜8points) and severe (≥8points). Theperipancreatic vascular involvement was noted in AP on MRI. Peripancreaticvessels observed included the splenic vein, splenic artery, superior mesenteric artery, superior mesenteric vein, portal vein, common hepatic artery andceliac trunk. A spectrum of vascular abnormalities included: artery invasion(vasculitis), vein invasion (vasculitis), pseudoaneurysm, venous thrombosis,left-sided portal hypertension etc. Spearman correlation of peripancreaticvascular involvement with the MRSI and the APACHEⅡ scores wereanalyzed.Results: In326patients with AP,81.9%of patients were edematous and18.1%of patients were necrotizing on MRI. There were38%（124/326）,55%（180/326）,7%（22/326） of patients had mild, moderate and severe APrespectively according to MRSI.16.9%（55/326）of AP patients had at leastone peripancreatic vascular involvement. The peripancreatic vascularinvolvement in AP patients on MRI included splenic vein thrombosis (7patients), superior mesenteric vein thrombosis (5patients), portal veinthrombosis (4patients), splenic artery pseudoaneurysm (3patients), splenicvein invasion (49patients), superior mesenteric vein invasion (24patients),portal vein invasion (41patients), splenic artery invasion (47patients),superior mesenteric artery invasion (36patients), common hepaticartery invasion (39patients), celiac trunk invasion (21patients), left-sidedportal hypertension (2patients).39%（23/59）of patients with necrotizing APhad peripancreatic vascular complication, which was higher than12%（32/267） of patients with edematous AP (P＜0.05). The prevalence amongsplenic vein thrombosis，superior mesenteric vein thrombosis，splenic veininvasion，splenic artery invasion at mild, moderate and severe AP accordingto MRSI had statistics differences and was positively correlated with theMRSI score (P＜0.05,0.3＜r＜0.5). Of326patients with AP, there were238 patients with mild AP and88patients with severe AP according toAPACHEⅡ. The prevalence of the vascular involvement between the twogroups had statistical difference, but had no correlation with the APACHEⅡscore (r＝0.114＜0.3,P＜0.05).Conclusion: Some patients with AP show peripancreatic vascularinvolvement on MRI, which included artery pseudoaneurysm, venousthrombosis, artery invasion, vein invasion and left-sided portal hypertension.The prevalence of splenic vein thrombosis， superior mesenteric veinthrombosis，splenic vein invasion and splenic artery invasion has a positivecorrelation with the severity of AP on MRI. The prevalence of peripancreaticvascular involvement has no correlation with the APACHEⅡ score.