Magnetic resonance imaging (MRI) and clinical features of different parts of the pancreas involved in acute pancreatitis: a cross-sectional study

Background: Patients with acute pancreatitis (AP) have different sites of pancreatic involvement. The aim of this study was to investigate the differences in magnetic resonance imaging (MRI) findings and clinical features of different sites of involvement (subtypes) in AP, with a view to complement and complete the classification of AP based on anatomical imaging features. Methods: We consecutively collected data from inpatients with AP from January 2018 to October 2022 at a tertiary care hospital. The patients with AP were classified into three subtypes by MRI: type I mainly involved the head of the pancreas; type II mainly involved the body and tail of the pancreas; and type III involved the entire pancreas (head, body, and tail simultaneously). We examined the MRI findings and clinical features of the three subtypes, including their prevalence, gender, etiology, age, assessment of severity, prevalence of hypertension, diabetes mellitus, coronary artery disease, laboratory markers, prognosis, necrosis, and the incidence of complications. The three subgroups were analyzed using one-way analysis of variance (ANOVA), Kruskal-Wallis H-test, Chi-squared test or Fisher's exact probability method depending on the data distribution, and logistic regression and linear regression were used to determine the risk factors for poor short-term prognosis of AP and the number of days in hospital. Results were considered statistically significant at P<0.05. Results: Among the 240 patients recruited, the mean age was 51 +/- 15 years (range, 12-89 years); 146 (60.83%) were male and 94 (39.17%) were female. Biliary pancreatitis accounted for 45.00% (108/240), hyperlipidemic pancreatitis for 33.75% (81/240), alcoholic pancreatitis for 8.75% (21/240), and unknown etiology for 12.5% (30/240). Some 81.25% (195/240) of the cases were edematous pancreatitis, whereas 18.75% (45/240) were necrotizing pancreatitis. Overall, 75 patients (31.25%) had type I AP, 108 patients (45.00%) had type II AP, and 57 patients (23.75%) had type III AP. These three subtypes were significantly different in terms of etiology, incidence of diabetes, C-reactive protein (CRP), severity, incidence of necrosis, local complications, clinical and imaging severity scores, and prognosis (P<0.05). Total pancreatic involvement (Type III) was the most severe subtype, with hyperlipidemia as the main cause. Regression analysis revealed that subtype classification is an important risk factor for prognosis. Conclusions: We classified AP into three subtypes based on different sites of involvement and revealed the MRI features and clinical characteristics of each subtype of AP. The subtype classification helps to characterize AP from the imaging dimension and predict the prognosis. The results of this study could be a target for future studies to adopt new classification methods.