The Optimal Definition of Sarcopenia for Predicting Postoperative Pneumonia after Esophagectomy in Patients with Esophageal Cancer

Background Esophageal cancer (EC) is associated with malnutrition in the vast majority of patients, and this often leads to sarcopenia, which is characterised by loss of skeletal muscle mass (SMM). Although sarcopenia could be one of the risk factors for postoperative pneumonia (PP), the optimal definition of sarcopenia using skeletal muscle mass index (SMI) by bioelectrical impedance analysis (BIA) remains unknown for predicting PP after esophagectomy in patients with EC. Therefore, this study aimed to identify high-risk patients for PP after esophagectomy through evaluating SMI by BIA and set an appropriate cut-off value for this purpose. Method A total of 73 patients with EC who underwent subtotal esophagectomy with lymph node dissection at Osaka City University Hospital between 2017 and 2019 were reviewed retrospectively. The association between PP and perioperative factors including SMI by BIA were analysed. When SMI was lower than the cut-off values proposed by two study groups (Asian Working Group for Sarcopenia (AWGS) and original European Working Group on Sarcopenia in Older People (EWGSOP)) or SMM was less than 90% of standard, the patient was diagnosed with sarcopenia. Receiver operating characteristic analysis was performed to set the appropriate cut-off value of SMI, and a new criterion (modified EWGSOP) was formulated by using the value. Clinicopathological factors and postoperative complications between sarcopenia and non-sarcopenia groups were compared, which were classified by four different criteria: (1) AWGS, (2) original EWGSOP, (3) < 90% standard and (4) modified EWGSOP criteria. Results Nine patients (12.3%) were with PP grade III or higher. Total SMI as well as body mass index (BMI), transthyretin and % vital capacity (%VC) were found to be significantly associated with PP (Clavien-Dindo grade >= III). BMI, total and appendicular SMI of sarcopenic patients were found to be lower than those of non-sarcopenic patients. Low serum albumin and %VC were significantly associated with sarcopenia defined by modified EWGSOP criteria. The rate of PP was significantly higher in sarcopenic patients when the original and modified EWGSOP criteria were used (p = 0.0079 and 0.0015, respectively). A multivariate analysis revealed that sarcopenic state by modified EWGSOP criteria was the significant independent predictive factor of PP [p = 0.0031, hazard ratio (HR) = 10.1; 95% confidential interval (CI): 2.12-76.9]. Conclusion Preoperative sarcopenia by modified EWGSOP criteria could be the best indicator using BIA for predicting PP after esophagectomy in patients with EC.