*Purpose: Dose-response relationship between preoperative and postoperative body mass index (BMI) and survival outcomes after liver transplantation (LT) has not been clearly defined yet.

*Methods: PubMed, Embase, Web of Science, and Cochrane databases were searched up to October 20th, 2019. We first performed pair-wise meta-analysis of optimal BMI (18.5-25 Kg/m²) versus underweight (<18.5), overweight (25-30), obesity class I (30-35), obesity class II (35-40) and obesity class III (>40) to pool the hazard ratios (HRs) and corresponding 95% confidence intervals (CIs). And dose-response analysis was used to estimate the relationship of graft- and patient-survival and BMI.

*Results: Twenty-three observational studies explored the association between preoperative BMI and survival outcomes (67981 individuals in 16 studies for patient survival and 36839 in 11 studies for graft survival), while only 1 reported postoperative BMI. In the pair-wise analysis, compared to optimal BMI, HRs of patient death in underweight, overweight, obesity I, obesity II, and obesity III were 2.13 (95%CI: 1.39-3.25, P<0.01), 0.96 (95%CI: 0.87-1.06, P=0.56), 1.06 (95%CI: 0.89-1.26, P=0.34), 1.36 (95%CI: 0.96-1.91, P=0.08), and 1.97 (95%CI: 1.19-3.28, P<0.01). Similarly, HRs of graft loss in underweight, overweight, obesity I, obesity II, and obesity III were 3.08 (95%CI: 1.47-6.44, P<0.01), 1.02 (95%CI: 0.89-1.18, P=0.43), 1.25 (95%CI: 0.85-1.84, P=0.14), 1.58 (95%CI: 1.08-2.30, P=0.03), and 2.90 (95%CI: 1.56-5.40, P<0.01). In the dose-response analysis, the non-linear model presents a U-shaped relationship between BMI and patient survival (P_nonlinearity <0.001). By reference to 17.5 kg/m², as BMI increased, the patient death risk decreased to 0.72 (95%CI: 0.62-0.84) in 27 kg/m² and then increased from 0.73(95%CI: 0.63-0.85) to 1.44 (95%CI: 1.09-1.90) in the 28.7-42 kg/m² range. For graft loss, a U-shaped relationship was also observed (P_nonlinearity <0.001). By reference to 17.5 kg/m², individuals in 26.7-28.0 kg/m² had the least risk with HR of 0.62 (95%CI: 0.48-0.80) and increased to 1.64 (95%CI: 1.03-2.61) in the BMI of 42 kg/m². Subgroup analysis by age, sex, sample size, and duration of follow-up presented similar results. We had no opportunity to explore the dose-response relationship between postoperative BMI and clinical outcomes for a lack of eligible studies.

*Conclusions: Preoperative underweight and severer obesity are both associated with significantly increased risk of graft loss and patient death. Overweight, especially BMI of 27-28 kg/m², may improve the survival outcomes most efficiently. More studies should be performed to explain the relationship between postoperative BMI and clinical outcomes after liver transplantation.

« Back to 2020 American Transplant Congress