Natural resolution or intervention for fluid collections in acute severe pancreatitis

Abstract

Abstract Background Revisions in terminology of fluid collections in acute pancreatitis have necessitated reanalysis of their evolution and outcome. The course of fluid collections in patients with acute pancreatitis was evaluated prospectively. Methods Consecutive adults with acute pancreatitis, who had contrast-enhanced CT (CECT) within 5–7 days of symptom onset, were enrolled in a prospective cohort study in a tertiary-care centre. Patients were treated according to standard guidelines. Follow-up transabdominal ultrasonography was done at 4-week intervals for at least 6 months. CECT was repeated at 6–10 weeks, or at any time if there were new or persistent symptoms. Asymptomatic collections were followed until spontaneous resolution. Risk factors for pancreatic pseudocysts or walled-off necrosis (WON) were assessed in multivariable analyses. Results Of 122 patients with acute pancreatitis, 109 were analysed. Some 91 patients (83·5 per cent) had fluid collections at baseline. Eleven of 29 with interstitial oedematous pancreatitis had acute peripancreatic fluid collections, none of which evolved into pseudocysts. All 80 patients with acute necrotizing pancreatitis had at least one acute necrotizing collection (ANC); of these, five patients died (2 after drainage), three underwent successful drainage within 5 weeks, and collections resolved spontaneously in 33 and evolved into WON in 39. By 6 months' follow-up, WON had required drainage in eight patients, resolved spontaneously in 23 and was persistent but asymptomatic in seven. Factors associated with increased risk of WON were blood urea nitrogen 20 mg/dl or more (odds ratio (OR) 10·96, 95 per cent c.i. 2·57 to 46·73; P = 0·001) and baseline ANC diameter greater than 6 cm (OR 14·57, 1·60 to 132·35; P = 0·017). Baseline ANC diameter over 6 cm was the only independent predictor of either the need for drainage or persistence of such collections beyond 6 months (hazard ratio 6·61, 1·77 to 24·59; P = 0·005). Conclusion Pancreatic pseudocysts develop infrequently in oedematous acute pancreatitis. Only one-quarter of ANCs either require intervention or persist beyond 6 months, whereas more than one-half of WONs resolve without any intervention within 6 months of onset. Baseline diameter of ANC(s) is an important predictor of outcome.