The impact of recipient body mass index on outcomes of pancreas transplantation: a systematic review
Harry de Souza1, Lilian Phillips-Kirkwood1, Henry C. Pleass1,2.
1Sydney Medical School, University of Sydney, Sydney, Australia; 2Department of Surgery, Westmead Hospital, Westmead, Australia
Introduction: Pancreas transplantation remains the gold standard, curative treatment for patients with insulin-dependent diabetes. The scarcity of available pancreata and high risks of the procedure necessitate careful and evidence-based patient selection criteria. Recipient selection criteria vary by country and institution, and often includes a body mass index (BMI) cut-off, generally falling between <30kg/m2 and <35kg/m2. The aim of this study is to present a systematic review of the current literature on the impact of preoperative BMI in adult pancreas transplant recipients on the outcomes of pancreas transplantation. Outcomes examined include short- and long-term graft and patient survival, and post-operative and disease related complications.
Method: An initial pool of 1025 articles were obtained multi-database search across Medline, Embase, the Cochrane Library, Scopus, and PubMed, as well as manual searches of the literature. Screening, as well as eligibility and bias assessments were conducted independently by the two first authors with 38 studies included for final analysis. Due to significant heterogeneity in reporting and analysis of BMI, population sizes of study cohorts, and outcome reporting, a meta-analysis of the data was not possible, and results are therefore presented as a narrative review.
Results: The impact of BMI on patient survival/death was discussed in 16 of the included 38 studies with 5 supporting its significance and 11 stating it was not significant. 23 of the included studies examined graft survival/loss, with 11 stating that BMI did have a significant impact on outcomes, and 12 stating that it did not. For both patient and graft survival, the groups of studies demonstrating a significant impact of BMI included most studies that were based on a large database, particularly the central OPTN database in the US – although the patient characteristics examined by each of these studies varies. The groups of studies reporting non-significance were predominantly single centre studies or used smaller databases. Average score on risk of bias for each group was similar. After comparing the size and quality of each study, the findings of this review trend towards BMI outside of the normal range having a significant impact on patient and graft survival. The impact of recipient BMI on a wide range of post-operative and diabetes-related outcomes were also examined across the 38 studies, and findings summarised in this review.
Conclusions: The results are limited by the heterogeneity of available literature, and there have been insufficient comparable high quality, recent, large population studies on this topic. Further primary studies are required to facilitate meta-analysis that will provide better evidence for the impact of recipient BMI on pancreas transplant outcomes and validate the continued use of BMI in patient selection for pancreas transplantation.
Kerry Hitos - advice on statistical analysis.
right-click to download