RESUMEN
Venovenous bypass (VVB) use during liver transplantation (LT) is notably variable among the centres and it is actually restricted to surgically complex cases, severely unstable recipients or grafts from high-risk donors. Historically, VVB was associated with the classical LT with caval cross clamping, while not much is known about the safety of this technique applied to piggyback LT. This retrospective observational study evaluated the effects of VVB applied to piggyback LT on mortality, hospital outcomes, postoperative graft and other organ dysfunction. We retrospectively collected data about recipient status, surgical complexity and graft quality of all the piggyback LTs performed at the Transplant Unit of IRCCS Azienda Ospedaliero-Universitaria di Bologna, Italy, from January 2012 to December 2022. A propensity score (PS) was built taking into account the variables possibly associated with either VVB choice and the investigated outcomes with the average treatment overlap method. PS-weighted general linear models (GLMs) were developed to investigate the adjusted effect of VVB use on the selected outcomes. The final analysis included 874 LT cases, of whom 74 (8.5%) underwent VVB. The effective sample sizes after PS-weighting were 280.2 and 64.3 patients in the no-VVB and VVB groups, respectively. PS-weighted GLMs did not show any differences regarding hospital and graft-related outcomes. However, significantly higher odds ratios for serum creatinine > 2 mg/dL and AKIN stage 2 or 3 during the first 24 h after ICU admission together with a higher renal replacement therapy need during ICU stay were reported for VVB exposure in the weighted analyses. This study suggests similar mortality and length of stay but a higher risk for postoperative acute kidney injury in patients undergoing piggyback LT with VVB.