A Novel Risk Model for Predicting Dysphagia after Tongue Reconstruction: A Retrospective Multicenter Study in Japan

Abstract

Background: There is no consensus on the postoperative outcomes of tongue reconstruction. Therefore, the authors developed a novel risk model for predicting dysphagia after tongue reconstruction. Methods: This retrospective study was conducted by the Oral Pharyngeal Esophageal Operation and Reconstruction Analytical, or OPERA, group across 31 cancer centers and university hospitals in Japan. A total of 532 patients [390 (73.3%) men and 142 (26.7%) women; median age at surgery, 60 years (range, 15 to 88 years)] who were diagnosed with oral tongue squamous cell carcinoma and underwent tongue reconstruction following glossectomy between 2009 and 2013 were included. Independent risk factors were identified using univariate regression analysis and converted to a binary format for multivariate analysis. An integer value was assigned to each risk factor to calculate a total score capable of quantifying the risk of feeding tube dependence. Results: Overall, 54 patients (10.2%) required a feeding tube at the time of evaluation. Predictive factors for feeding tube dependence were advanced age, lower American Society of Anesthesiologists physical status, low body mass index, lower serum albumin, comorbid hypertension and diabetes, extended tongue defect, resection beyond the tongue, laryngeal suspension, postoperative radiation therapy, and no functional teeth. In multivariate logistic regression analysis, age greater than or equal to 58.5 years, postoperative radiation therapy, wider tongue defect, and body mass index less than 21.27 kg/m2 earned 6, 4, 3, and 2 points, respectively, for a maximum total score of 15. Conclusion: The authors’ risk model provides a mathematical tool for estimating the individual risk of postoperative feeding tube dependence before tongue reconstruction. CLINICAL QUESTION/LEVEL OF EVIDENCE: Risk, III.