Clinical Outcomes in A Multi-center Cohort Involving 919 Patients with Hypertriglyceridemia-associated Acute Pancreatitis

Objectives: Hypertriglyceridemia-associated acute pancreatitis (HTG-AP) is one of the most common etiologies of acute pancreatitis (AP) worldwide. Compared to other etiologies, patients with HTG-AP may develop more severe AP, but previous studies yielded controversial conclusion due to the lack of adequate adjustment for the confounders. Therefore, this study aimed to examine the possibility and risk factors of developing severe AP in HTG-AP. Methods: Data from patients with an established diagnosis of AP were collected from January 2013 to December 2023 using a pre-designed data collection form and were gathered from five tertiary cross-regional centers of China. HTG-AP was defined as serum triglyceride (TG) levels >500 mg/dl and excluded other etiologies. The possibility and risk factors of severe AP were assessed by multivariable logistic regressions after adjusting potential confounders. A prediction model was established and validated. Results: Between 2013 to 2023, we identified a total of 6996 patients with AP, of whom 4378 were included in the final analysis. Compared to other etiologies, patients with HTG-AP had a higher risk of developing severe AP (odds ratio [OR]: 1.897; 95% confidence interval [CI]: 1.380-2.608; p<0.001) and organ failure. HTG-AP patients showed higher possibility for developing respiratory and circulation failure but renal failure compare to other etiologies. In HTG-AP patients, risk factors for severe AP included age, fasting blood glucose, white blood cell (WBC) counts, and the presence of pleural effusion. Triglyceride level was found not significantly associated with severity in HTG-AP patients. A prediction model incorporating these risk factors demonstrated an AUC of 0.837 in the training and 0.883 in the testing set, with adequate calibration. Conclusions: Using a multi-center cross-regional cohort, we demonstrated that HTG-AP had a higher risk of developing severe AP and organ failure. A risk prediction model for predicting severe AP was developed and effectively stratified patients.