Response to: Comment on: Textbook Outcomes following Open Live Donor Right Hepatectomy and Open Right Hepatic Lobectomy for Cancer in 686 patients: Re-defining the Benchmark.

Abstract

We read the letter to the editor by Liu et al regarding our recently published study on right hepatectomy (RH) textbook outcomes (TO) with great interest.1,2 We want to thank Liu et al for acknowledging our work and for their insightful comments, which we are hoping to address below. First, the authors mention that in their experience, “real-world” RH for cancer is more commonly performed for hepatocellular carcinoma (HCC), while in our study, the most common indication was colorectal cancer liver metastasis (CRLM) in 58.4% of patients. Consequently, we observed a low rate of cirrhosis (5.4%) in our RH population, whereas Liu et al argue that the rate of cirrhosis could be higher in their patient population where RH is more often performed for HCC. We acknowledge that liver resection is often performed for HCC in Eastern centers, but in Western countries CRLM remains the most common indication for liver resection, including RH.2–5 The authors point that the presence of cirrhosis could affect the patient’s ability to achieve a TO is well taken. In fact, this could be one of the reasons why we observed a much higher rate of TO (92.2%) in live donor patients, where the prevalence of cirrhosis was zero, compared to TO in cancer patients (53.7%), where cirrhosis was present in 5.4% of patients. Second, we agree with the authors that abnormal BMI, either too low, indicating malnutrition, or too high, indicating obesity, is an important marker of patient’s baseline health. Liu et al have previously published on the association of abnormal BMI with failure of TO after liver resection for HCC.6 Although, in our study, logistic regression did not reveal BMI to be associated with TO in either the cancer or the live donor patients, this could be due to fact that we analyzed BMI as a continuous variable. However, similar to the discussion about cirrhosis above, live donor patients had lower BMI on average compared cancer patients, and this difference could partially account for the higher TO rates in the live donor cohort. We certainly aim to investigate the association of BMI with TO rates in a more granular fashion in the future, by grouping BMI similarly to the methodology described by Liu et al. Third, the authors make the interesting suggestion of using advanced statistical methods, such as propensity score matching, to balance the baseline clinicopathologic factors of the cancer and live donor patients and then compare TO rates between the 2 groups. They argue that the baseline large differences in the 2 populations may not make direct TO comparisons statistically appropriate without matching. Although we definitely considered the above when designing our study, we decided that our main goal was to report on our entire and unedited experience with 686 RH patients over a decade and we did not want to exclude patients that could not be matched. Consequently, the data reported in our article represent true “real-world” TO rates after RH performed either for cancer or live donor purposes in a high-volume Western center. We acknowledge that the live donor group consisted of younger, healthier, and thinner patients, and we discuss how these factors likely contributed to the much higher TO rates in that group. However, we believe that this unique population of patients that underwent RH while otherwise healthy, gave us the opportunity to uncover the “best-achievable” TO results following this operation. In summary, we want to thank Liu et al for their thoughtful appraisal of our work and we are content that our study is generating productive discussion amongst liver surgeons with the ultimate goal of improving outcomes for our patients. Acknowledgments E.D., D.G., and A.H. contributed equally to this work.