Perioperative loss of psoas muscle is associated with patient survival in living donor liver transplantation.

Abstract

Potential conflict of interest: Nothing to report. TO THE EDITOR: With great interest, we read the article by Chae et al.1 assessing the association between perioperative decrease in the psoas muscle index (PMI) and patient mortality after living donor liver transplantation (LDLT). By multivariate regression analysis, they showed that a decline in PMI ≤ −11.7% from the day before surgery to postoperative day 7 was significantly associated with overall mortality during the follow‐up period of 3.3 ± 2.3 years. Thus, they concluded that a PMI change of ≤−11.7% could be used as an independent predictor of overall mortality after LDLT. Other than the limitations described in the Discussion section of the article by Chae et al., we noted other issues in this study that were not well addressed. First, mean preoperative hematocrit and albumin levels were not significantly different between low‐loss and high‐loss groups, but directly comparing 2 variables only has a limited clinical significance. In fact, preoperative anemia and hypoalbuminemia are highly prevalent in patients with severe chronic liver disease. The available evidence shows that preoperative anemia can reflect the severity of chronic liver disease and is more common with advanced liver disease. When the Model for End‐Stage Liver Disease scores are removed from the analysis, preoperative low hematocrit levels are associated with increased mortality following liver transplantation, suggesting that anemia may reflect a worse preoperative burden of disease.2 For patients with liver transplantation, moreover, preoperative anemia has been shown as an independent predictor of the need for intraoperative transfusions,3 which is a known risk factor for increased mortality following LDLT.4 The physiological functions of albumin include binding and transporting a large number of metabolites, controlling plasma volume by preserving colloid oncotic pressure, scavenging of free radicals, acting as a reservoir for nitric oxide, and affecting capillary membrane permeability. Preoperative hypoalbuminemia has been associated with the development of postoperative acute kidney injury, which is a frequent complication of liver transplantation and is associated with increased mortality after LDLT.5 Second, in postoperative findings, the authors only provided dialysis, early allograft dysfunction, all‐cause reoperation, and infection, but not other postoperative complications. In fact, postoperative acute kidney injury; gastrointestinal bleeding; pulmonary complications; sepsis; biliary tract complications including biliary stenosis or leakage; major vascular complications including isolated or combined hepatic artery, portal vein, and hepatic vein thrombosis or stenosis; and multiple‐organ failure have been shown as the independent risk factors of increased early‐ and late‐term mortality after LDLT.6 In a retrospective study, the multivariate regression analysis is useful for adjusting potential confounders and controlling selection biases, but a limitation of this statistical method is the assumption of a particular mathematical relation between the intervention and the measured outcome. To obtain the true inferences of multivariate regression analysis for adjusted hazard ratio of a measured outcome, all of the known risk factors affecting the measured outcome must be taken into the model. If an important risk factor is missed, multivariate adjustment for the hazard ratio of the measured outcome can be biased, and a spurious association between the intervention and the outcome of interest may be obtained.10 In this study, thus, not taking above postoperative complications into the model would have tampered with the inference of the multiple regression analysis when assessing associations between perioperative PMI changes and overall mortality in LDLT patients. Moreover, a multivariate regression analysis is unable to differentiate whether a decline in PMI ≤ −11.7% is a true determinant of overall mortality or simply a synthetical manifestation of worsened perioperative conditions that can significantly increase overall mortality of LDLT patients. Thus, we agree with the authors that future studies are still required to verify whether perioperative PMI change is really a modifiable risk factor affecting overall mortality after LDLT.