Morbidity and mortality after major liver resection in patients with perihilar cholangiocarcinoma: A systematic review and meta-analysis

Background: Morbidity and mortality after hepatectomy for perihilar cholangiocarcinoma are known to be high. However, reported postoperative outcomes vary, with notable differences between Western and Asian series. We aimed to determine morbidity and mortality rates after major hepatectomy in patients with perihilar cholangiocarcinoma and assess differences in outcome regarding geographic location and hospital volume. Methods: A systematic review was performed by searching the MEDLINE and EMBASE databases through November 20, 2017. Risk of bias was assessed and meta-analysis and metaregression were performed using a random effects model. Results: A total of 51 studies were included, representing 4,634 patients. Pooled 30-day and 90-day mortality were 5% (95% CI 3%-6%) and 9% (95% CI 6%-12%), respectively. Pooled overall morbidity and severe morbidity were 57% (95% CI 50%-64%) and 40% (95% CI 34%-47%), respectively. Western studies compared with Asian studies had a significantly higher 30-day mortality, 90-day mortality, and overall morbidity: 8% versus 2% (P <.001), 12% versus 3% (P <.001), and 63% versus 54% (P =.048), respectively. This effect on mortality remained significant after correcting for hospital volume. Univariate metaregression analysis showed no influence of hospital volume on mortality or morbidity, but when corrected for geographic location, higher hospital volume was associated with higher severe morbidity (P =.039). Conclusion: Morbidity and mortality rates after major hepatectomy for perihilar cholangiocarcinoma are high. The Western series showed a higher mortality compared with the Asian series, even when corrected for hospital volume. Standardized reporting of outcomes is necessary. Underlying causes for differences in outcomes between Asian and Western centers, such as differences in treatment strategies, should be further analyzed. (C) 2019 Elsevier Inc. All rights reserved.