Abstract 4523: Personalized risk prediction for treatment intolerance in operable HPV-negative head and neck cancer: Beyond the eyeball test

Abstract Introduction: Head and neck cancer (HNC) treatment is highly complex, with substantial toxicity risks from curative therapies, often including surgery, radiation, and chemotherapy. Poor treatment tolerance can hinder optimal recovery, impacting patient outcomes. Current predictive tools lack precision in forecasting treatment intolerance for HNC patients undergoing curative surgery. Frailty and sarcopenia, especially muscle quantity biomarkers, are increasingly recognized as crucial predictive factors. This study aims to develop a comprehensive risk index incorporating sarcopenia, frailty, and patient-specific factors to better predict intolerance. Methods: An ambispective observational study was conducted with 542 patients undergoing curative-intent HNC surgery from 2015 to 2024. Patient data on demographics, frailty, sarcopenia (via cervical paraspinal muscle index or CPSMI), tumor staging, and planned adjuvant therapy were collected. Using logistic regression, we evaluated predictive factors for treatment intolerance, defined as major adverse events (MAEs) or incomplete treatment. Risk model performance was validated internally and externally and compared against ASA, mFI, and RAI. Results: In 542 patients, lower CPSMI (first tertile) correlated significantly with intolerance (OR 1.85, 95% CI 1.10-3.13). High-density muscle was protective, while low-density muscle predicted intolerance (OR 1.88, 95% CI 1.18-2.99). Sarcopenia and frailty were independently predictive, with the mFI also strongly correlating with intolerance (OR 4.92, 95% CI 1.88-12.9). The new risk index demonstrated superior predictive accuracy, showing a 0.04 increase in the C-statistic compared to existing indices (p=0.02). Conclusion: The multimodal risk index effectively predicts treatment intolerance in HNC surgery patients. By integrating sarcopenia, frailty, and patient-specific factors, the tool enables better preoperative counseling and intervention planning, with particular utility for high-risk individuals. Further research on prehabilitation strategies may enhance treatment tolerance in vulnerable patients, supporting a proactive approach to HNC care. Citation Format: Marco Antonio Mascarella, Keith Richardson, Nader Sadeghi, Alex Mlynarek, Michael Hier, Khalil Sultanem, Christina Tsien, Khashayar Esfahani, Marie-Jeanne Kergoat. Personalized risk prediction for treatment intolerance in operable HPV-negative head and neck cancer: Beyond the eyeball test [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4523.