Systematic review and meta-analysis of thrombocytopenia as a predictor of post-hepatectomy liver failure.

ABSTRACT

BACKGROUND: We performed a systematic review and meta-analysis to assess whether thrombocytopenia constituted a risk factor for post-hepatectomy liver failure (PHLF). METHODS: We searched MEDLINE and EMBASE from inception until February the 17th, 2018 for studies reporting cases of PHLF in patients with and without thrombocytopenia (defined as a platelet count below 100 or 150 (G/l)) and/or platelet counts in patients with and without PHLF. Pooled odd ratios for PHLF, as well as mean difference in platelet counts between patients with and without PHLF, were obtained by random effects models. Robustness was tested by subgroups and leave-one out sensitivity analyses. Heterogeneity was assessed using the Q-test and quantified based on I2 value. RESULTS: We included 15 studies representing 3966 patients. Pooled odds ratio for PHLF in thrombocytopenic patients was 3.71 (95% CI 2.51 to 5.48; I2 = 0%). Pooled odds ratio was 5.53 (95% CI 2.85 to 10.48) when pooling only studies based on preoperative platelet count, and 3.13 (95% CI 1.75 to 5.58) when pooling studies including only patients without liver cirrhosis. The pooled mean difference in platelet counts between patients with and without PHLF was -21.2 (G/l) (95% CI -36.1 to 6.4) in disfavor of patients with PHLF. When pooling only patients with various qualities of liver tissue, the pooled mean difference was 0.6 (G/l) (95% CI -21.1 to 22.2). CONCLUSION: Preoperative and/or postoperative thrombocytopenia constitute significant risk factors for PHLF in cirrhotic and non-cirrhotic patients.