Management of spontaneous portosystemic shunts at the time of liver transplantation: treatment or observation? Results of a systematic review

Background Optimal treatment of spontaneous portosystemic shunts (SPSS) during liver transplantation (LT) remains debated. We systematically reviewed the literature on definitions, treatment and outcomes of patients presenting SPSS undergoing LT. Methods According to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we used PubMed to retrieve all studies dealing with SPSS and LT between January 1987 and January 2020. The primary endpoints were definitions and outcomes according to the management of SPSS (treatment vs observation). Results Thirteen studies detailing the management of 962 SPSS were retrieved. Hemodynamically significant SPSS were defined as those having diameter >= 10 mm in 41% (n = 395) of patients. SPSS were splenorenal (42%), cavo-gastric (15.2%), umbilical (7.4%), mesenterico-caval (n = 31; 3.2%), mesenterico-renal (0.1%) and unreported (31.9%), respectively. At the time of LT 372 shunts (38.7%) were treated while 590 were observed (61.3%). During a follow-up time ranging from 4 months to 5 years, the reported overall survival (OS) at 1 year was not significantly different except for one study. Portal vein anastomosis complications (i.e. reduced flow, stenosis or thrombosis) were similarly reported in observed [n = 26 (4%)] and ligated SPSS [n = 10 (2%)] (p = 0.22) but the rate of relaparotomy was significantly higher in observed SPPS (16 vs 2; p = 0.01) to rescue post LT portal vein thrombosis (n = 6) and reduced portal flow and graft dysfunction (n = 10). Conclusions There was a heterogeneous management of SPSS during LT in the literature. Ligation of SPPS did not reduce vascular complications neither improved survival. A randomized prospective study might contribute to identify best management of SPSS at time of LT.