Authors’ Reply: Laparoscopic Versus Open Liver Resection for Hepatocellular Carcinoma: A Case Controlled Study with Propensity Score Matching

Abstract

To the Editor, We are grateful to Drs. Tahir, Perin, and Balasubramanian for their comments on our recent publication titled ‘‘Laparoscopic Versus Open Liver Resection for Hepatocellular Carcinoma: A Case Controlled Study with Propensity Score Matching’’ [1]. We read the letter with great interest [2]. The questions in their letter to the editor [2] are important, and we have addressed them below to clarify the main points of our study. First, as they highlighted, our study should have been called a retrospective cohort study rather than a case-controlled study. We classified it with reference to a previous similar study, but admit that their point is correct [3]. Second, they pointed out that the cause of the reduced complications may not be laparoscopic surgery but rather the recent advances in perioperative management. We divided the study cohort into two groups according to the era of surgery and examined whether laparoscopic surgery remained a factor that reduced complications even when the era was taken into consideration as a variable in the multivariate analysis. As a result, it was revealed that both were independent complication-reducing factors, confirming that laparoscopic surgery was associated with reduced complications. There were no major changes in perioperative management at our institution during the study period except for the choice between open surgery and laparoscopic surgery. Thus, it cannot be completely ruled out that advances in surgical techniques, especially laparoscopic surgeries, may have affected the results. Third, as the reviewers’ pointed out, having a small number of surgeons exclusively undertaking surgeries would be an advantage for a single-institution study, but not at our facility. However, this may lead to increased external validity. Fourth, as they suggested, we compared overall survival (OS) and disease-free survival (DFS) in the propensity score matching cohort. We found that both were comparable between the LLR and the OLR groups (5-year OS: 69.9% in LLR and 76.1% in OLR; 5-year DFS: 34.5% in LLR and 35.6% in OLR). Finally, they pointed out that ‘‘association’’ and ‘‘causality’’ are not necessarily the same and that ‘‘causality’’ must be verified by prospective trials. We agree with this and included the following statement in the revised manuscript: ‘‘More careful analysis of the data, or ideally randomized controlled trials, are necessary to prove this concept.’’ We hope that our responses have adequately addressed their questions and again thank the reviewers for their thoughtful perspectives on our report on surgical outcomes for resection of HCC.