Hypertriglyceridemia associated acute pancreatitis in intensive care unit and therapeutic plasmapheresis

Objectives: hypertriglyceridemia induced acute pancreatitis (HTG-AP) is a severe condition seen in Intensive Care Unit (ICU), associated with high mortality. Recent clinical data suggest that therapeutic plasma exchanges (TPE) might improve HTG-AP’s course.Objective: to describe this emergent pathology in ICUs and to find HTG-AP’s prognosis factors.Methods: this retrospective, observational cohort study was conducted among 12 different ICUs in France from 2010 to 2020. Data were recorded from medical files for each patient. We evaluated factors influencing survival status at day 28 using a multivariable logistic regression analysis.Results: 95 adult patients with an HTG-AP diagnosis were enrolled in the study. Mortality rate was 13.3% at day 28. Death at day 28 was associated with worse SOFA (Odds Ratio (OR) =2.545, IQ [1.284-5.044], p=0.0074) and RANSON score (OR=1.313, IQ [1.028-1.677], p=0.029) at admission. Median ICU length-of-stay was 7 (1-96) days in the survival group versus 2 (1-97) days in the non-survival group (p=0.1009). The triglyceride (TG) level was not associated with mortality in our univariate and multivariate analysis (p=0.2283). Only 37% of patients were treated by TPE in this study, without significant impact on mortality (OR=1.011, IQ (0.973-1.05), p=0.1227). Conclusion: in this first French cohort of HTG-AP patients in ICU, TG level at admission and TPE are not associated with mortality. ; Contexte : les pancréatites aiguës d’origine hypertriglycéridémique (HTG-PA) sont une cause sévère d’hospitalisation en réanimation, associée à un taux de mortalité important. Les données récentes montrent que la plasmaphérèse pourrait avoir un impact positif sur l’évolution de cette pathologie. Objectifs : décrire cette pathologie en réanimation et identifier des facteurs pronostics.Méthode : cette étude rétrospective, observationnelle, multicentrique a été menée dans 12 centres en France de 2010 à 2020. Les données ont été récoltées à partir des dossiers médicaux de chaque patient. Nous avons analysé ...