Objective: Although surgical resection is a mainstay in the management of esophageal carcinoma (EC), its postoperative outcomes remain unsatisfactory. To optimize surgical strategies for EC, a simple method of stratifying patients according to risk factors is desired. Controlling nutritional status (CONUT), the prognostic nutritional index (PNI), transthyretin and transferrin are nutritional parameters used to predict the long-term outcomes of EC patients. We aimed to comparatively evaluate the prognostic significance of these four markers, measured preoperatively, in patients with operable EC.
Methods: In total, 224 patients undergoing surgical resection for EC were retrospectively reviewed. Overall/cancer-specific survivals (OS/CSS) were estimated applying the Cox proportional hazard model to univariate and multivariate analyses. PNI, transthyretin and transferrin levels were treated as continuous variables in these analyses.
Results: Preoperative CONUT had significant associations with tumor location, depth and preoperative irradiation. The other three markers all showed significant relationships with age and tumor depth. On univariate Cox regression analysis, preoperative CONUT, PNI, transthyretin and transferrin all correlated significantly with OS and CSS. On multivariate Cox regression analysis, the preoperative transthyretin level was identified as an independent predictor of OS (HR 0.51 per 10 mg/dL increase, 95% CI 0.29-0.88, p = 0.017) and CSS (HR 0.50, 95% CI 0.27-0.91, p = 0.027) as well as tumor depth, nodal metastasis and preoperative irradiation, while the other three parameters were not.
Conclusions: Preoperative transthyretin, as a continuous variable, independently predicted both OS and CSS in resectable EC patients, appearing to be the best prognosticator among conventional nutrition-related parameters.
Keywords: Esophageal carcinoma; Esophagectomy; Nutrition; Outcome; Transthyretin.