The efficiency of postoperative liver failure and mortality risk scores after liver resection for perihilar cholangiocarcinoma.

With great interest, we read the literature by Noji et al. entitled “Validation study of postoperative liver failure and mortality risk scores after liver resection for perihilar cholangiocarcinoma” (1), which was published in the latest issue of Hepatobiliary Surgery Nutrition. The authors have reached an important conclusion that postoperative mortality risk score (POMRS) and post-hepatectomy liver failure score (PHLFS) proposed by the Academic Medical Center (AMC) and Memorial Sloan Kettering Cancer Center (MSKCC) require optimization before use in clinical practice. However, after a careful reading of this study, we would like to address several fundamental flaws related to this article.

First, there are some obvious mistakes in Tabs. 2, 4, 5 in Ref. (1). Tab. 2 was listed to show the risks predicted by the post-hepatectomy liver failure (PHLF) risk score of Hokkaido University and AMC/MSKCC. To begin with, based on PHLF risk score of AMC/MSKCC (2), the predicted risk of PHLF of high risk with total points of 7 is 77% not 67% in Tab. 2 in Ref. (1). Furthermore, after a careful review, we noticed that the authors appeared to have made some apparent mistake inadvertently in Tabs. 4 and 5. The number of jaundice in no 90-day mortality group is 202 [81.8]/42 [18] rather than 202 [82]/47 [18] in Tab. 4. Meanwhile, there is an obvious typographic error in Tab. 4, the number of jaundice of 90-day mortality group is reported in Wiggers et al.’s (3) article with 29 [72.5]/1 [27.5]. In addition, the number of portal vein reconstruction in no 90-day mortality group is 40 [16.2]/207 [83.8] rather than 51 [18]/236 [82] in Tab. 4. Finally, the predicted risk of POMRS of low risk with total points of 1 is 1% not 2% in Tab. 5.

Second, the definition of mortality is different in Noji et al. and Wiggers et al.’s studies (1,3). The definition of mortality is 90-day in-hospital death after surgery in Noji et al.’s study (1). While mortality is defined as any cause of death within 90 days of resection in Wiggers et al.’s studies (3). This might be one reason potentially lead to a relatively lower area under the curve (AUC) level of POMRS (0.58). Thus, we suggest the authors choose the same criteria to get a more accurate and reliable conclusion. Meanwhile, the definition of preoperative cholangitis is different in Noji et al. and Wiggers et al.’s studies (1,3). Preoperative cholangitis should fulfill all 3 criteria including higher than 38 ℃, after new placement of a drainage catheter the temperature decreased, and/or liver abscess formation with fever in Noji et al. study (1). While preoperative cholangitis was defined as the patient had an episode of fever, abdominal complaints, and leucocytosis requiring (additional) biliary drainage, at any time in the preoperative course in Wiggers et al.’s paper (3). That’s why preoperative cholangitis of POMRS did not influence 90-day mortality in the present study.

Third, as results show in the present study, longer operative time and higher volume of operative bleeding, were associated with postoperative mortality and liver failure. Therefore, we recommend the author add operative time and operative bleeding as risk factors to improve POMRS and PHLFS model for calculating postoperative mortality and postoperative liver failure.

We thank all authors for their excellent contributions to validate the efficiency of POMRS and PHLFS for patients with perihilar cholangiocarcinoma after surgery. In our opinion, further high quality randomized controlled trials (RCTs) are still needed to further validate these findings.

The article’s supplementary files as