Letter to Editor: Association Between Postoperative Liver Dysfunction and Blood‐Derived MicroRNA Signatures

Supported by grants from the National Natural Science Foundation of China (No. 81472775 and 81874211). Potential conflict of interest: Nothing to report. To the Editor: We read with great interest the article by Starlinger et al.1 This translational study analyzed preoperative plasma microRNA (miRNA) signature levels in two cohorts (exploratory and confirmatory) of 154 patients with hepatectomy. Random forest modeling showed that two top‐ranked miRNA pairs were identified (151a‐5p/192‐5p, 122‐5p/151a‐5p) with significant preoperative differences between individuals developing liver dysfunction (LD) and controls, and multivariate logistic regression analysis revealed that two miRNA ratios were independently associated with LD after curative resection (all P < 0.01). Therefore, the author concluded that the novel miRNA‐based biomarker represents a promising prognostic factor in patients undergoing partial hepatectomy, which might be helpful in guiding surgical decisions to the specific risk profile of patients. However, we would like to raise the following points: As the authors mention, patients with and without LD were matched based on basic characteristics and liver function. However, showed in their study, albumin levels differed significantly between the two groups (39.1 [31.5‐47.3] vs. 43.2 [37.0‐46.6] g/dL; P = 0.004), suggesting that there was an unbalanced background for comparisons between the two groups. Albumin is one of the main factors in Child‐Pugh classification, and it is an important marker of synthetic function. Decreases in serum albumin concentrations lead to increased perioperative morbidity and mortality rates.2 Therefore, we highly recommend that albumin should be matched between the two groups. In addition, the number of variables on patients’ baseline characteristics in this study seemed insufficient (e.g., operative time, operative estimated blood loss,3 need for vena cava or other vascular resection,4 and blood transfusion requirement5), even though these variables have an important influence on the postoperative morbidity and mortality, as demonstrated in many studies.3 Therefore, the reliability and accuracy of the conclusions of this study need further evaluation. Furthermore, whether analyzing miRNA would cost more money and cause longer wait times or even delay treatment is another important aspect that must be considered in translational medicine. In conclusion, we deeply appreciate the authors’ work. Moreover, we suggest that the level of albumin in preoperative groups should be matched. In addition, more details on the variables of patients’ characteristics should be included in this study. Clarification about the omissions we mention would greatly solidify the conclusions of the study by Starlinger et al.1