Factors affecting outcome in acute hypertriglyceridemic pancreatitis treated with plasma exchange: an observational cohort study.

Abstract

OBJECTIVES: The optimal therapy for hypertriglyceridemic acute pancreatitis, especially the role of plasma exchange (PE), is not entirely clear. The aim of our large, single-center, observational, cohort study was to analyze the factors affecting outcome in hypertriglyceridemic pancreatitis treated with PE. METHODS: We included 111 episodes of hypertriglyceridemic pancreatitis treated with PE, which occurred in 103 different patients. The Acute Physiology and Chronic Health Evaluation (APACHE) II score, triglycerides, delay to first PE, and PE treatment details were retrospectively obtained from the patients' records. The main outcome measures were length of hospitalization and in-hospital mortality. RESULTS: The patients were 47±9 years old and the median APACHE II score at first PE was 4 (inter-quartile range (IQR) 2-7). There was a seasonal variation in the incidence of hypertriglyceridemic pancreatitis, and the recurrence rate was 1.6% per year. Triglycerides at presentation did not correlate with APACHE II or influence the outcome. The mean reduction in triglycerides during PE was 59% (from 44±31 to 18±15 mmol/l), which was twice the reduction observed during conservative treatment (27% daily). The median hospital stay was 16 days (IQR 10-24) and in-hospital mortality was 5%. The median delay to first PE was 35 hours (IQR 24-52), and there was no difference in mortality in the early and late PE groups (7% vs. 6%, p = 0.79). The group with citrate anticoagulation during PE had a significantly lower mortality than the group with heparin anticoagulation (1% vs. 11%, p = 0.04), and citrate was an independent predictor also in the multivariate model (p = 0.049). CONCLUSIONS: PE effectively reduced serum triglycerides faster than could be expected with conservative treatment. The delay in PE therapy did not influence survival. We found that citrate anticoagulation during PE was associated with reduced mortality, which should be confirmed in a randomized study.