A systematic review and meta-analysis of cold in situ perfusion and preservation for pancreas transplantation.Hameed AM, Wong G, et al.
Hpb, 2017; 19: 933-943.
To identify the most effective solution for in situ perfusion/preservation of the pancreas in donation after brain death donors, in addition to optimal in situ flush volume(s) and route(s) during pancreas procurement.
Embase, Medline and Cochrane databases were utilized (1980–2017). Randomized controlled trials (RCT; or quasi-RCTs) and/or observational articles were deemed eligible for this review, without language restriction. An article was only included if it presented data for a minimum of at least 10 patients/transplants per study group, and included information regarding perfusion fluid route(s), flush volume(s), back-table perfusion and final preservation of the pancreas. Two independent reviewers extracted study data. Risk of bias was assessed.
805 records were identified by the database search. After screening and eligibility assessment, 13 studies were included in qualitative synthesis (939 transplants). Of these, 4 observational studies were included in the meta-analysis.
Primary study outcomes included peak amylase and lipase in the first week post-transplantation, the number of pancreatitis episodes, and thrombotic graft loss. Other secondary outcomes of interest included C-peptide and HbA1C at last follow-up, acute rejection rates, graft survival, hospital length-of-stay, and surgical complications. Median ischemic times, donor/recipient ages, perfusion volumes, and graft survival were calculated based on the number of patients in each study group. Meta-analyses were conducted using studies with directly comparable groups.
Studies published between 1980–2017.
This is a well conducted systematic review of in situ preservation for pancreatic allografts. A broad search of several databases was undertaken, and two reviewers assessed papers and extracted data independently. The protocol was prospectively registered. The authors have completed a quality assessment of included studies and this illustrates the main weakness of the paper, being the underlying evidence on which it is based. To say that the review includes only RCTs and quasi-RCTs is not entirely correct as the majority of papers are cohort studies and only 3 relatively small RCTs were included; this represents the level of evidence in this specific field unfortunately. The authors acknowledge that the overall risk of bias for the included cohort studies is high and that the RCTs could not be thoroughly assessed due to a lack of information for the majority of domains. The meta-analyses presented had to be based solely on just a few cohort studies as there were insufficient RCTs with comparable groups eligible for meta-analysis. The only difference of note was a lower peak serum lipase when UW was used in comparison with HTK. There were no other notable differences in serum amylase, graft loss, hospital stay or pancreatic leak. The authors are rightly cautious about the conclusions that can be drawn from this evidence base.