Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
Objective: To investigate the predictive role of tumor volume in the need for gastrostomy tubes and overall mortality in patients with advanced laryngeal and hypopharyngeal cancer.
Design: Multi-institutional observational cohort study
Methods: The records of 41 head and neck cancer centers across the United States and Canada were reviewed with a single institution serving as the coordinating site. The study included patients with newly diagnosed advanced squamous cell carcinoma of the larynx (cartilage-invading T3, T4) or hypopharynx (T2, T3) for which curative treatment was planned. Tumor and treatment characteristics were recorded alongside feeding, functional status, and patient-reported outcomes. From this cohort, primary imaging was obtained and reviewed by a single head and neck cancer neuroradiologist and tumor characteristics, subsite involvement, and other imaging features were recorded. For this abstract, we consider gastrostomy to be the primary outcome.
Results: From an original cohort of 279 patients, 156 patients had imaging available for review and were included in this study. Cohort was predominantly male (74%) with laryngeal primary tumors (88%), and mean age was 65.2 years. Fifty-nine percent underwent primary surgical laryngectomy or laryngopharyngectomy while 41% were treated with primary chemoradiation.
A multiple logistic regression model demonstrated that treatment modality (surgery vs. chemoradiation) and tumor volume were the most important predictors of gastrostomy use. Patients undergoing chemoradiation had a higher likelihood of requiring a gastrostomy than those receiving laryngectomy [OR 4.00 (95% CI: 1.66-9.64, p=0.002)]. Tumor volume significantly predicted gastrostomy dependence [OR 1.036 (95% CI: 1.003-1.069, p=0.029)], indicating a 4% increased risk of needing a gastrostomy per 1 mL increase in tumor volume. A predictive mathematical model was generated representing probability of gastrostomy requirement based on tumor volume threshold. Volumes of 18 mL, 53 mL, and 86 mL corresponded to 50%, 75%, and 90% predicted risk of gastrostomy dependence, respectively. Variables that were not predictive of gastrostomy included T stage, N stage, and anatomic site (hypopharynx vs larynx), and a detailed subsite analysis is pending. Additionally, tumor volume was predictive of 12-month mortality [OR 0.937 (95% CI: 0.882-0.996, p=0.035)] with each 1 mL tumor volume increase correlating to a 6.3% higher risk of death at 12 months.
Conclusions: This study underscores tumor volume as a significant predictor for both gastrostomy dependence and mortality in head and neck cancer, suggesting its importance as a clinical marker for optimizing patient management and counseling strategies. Further research is warranted to confirm these findings in larger cohorts.