3701 - Risk Factors for Developing Radiation Related Late Toxicities after Treatment for Head and Neck Cancers and Impact on Survival

J. Mohan Mathew¹, J. Ringash¹, J. Su², M. Woodside¹, S. V. Bratman¹, J. Cho¹, E. Hahn¹, A. Hosni¹, A. J. Hope¹, J. Kim¹, A. McPartlin¹, B. OSullivan¹, C. J. Tsai¹, J. Waldron¹, J. Kwan¹, A. Spreafico³, D. Goldstein⁴, S. H. Huang¹, and P. Wong¹; ¹Department of Radiation Oncology, Princess Margaret Cancer Centre, University of Toronto, Toronto, ON, Canada, ²Department of Biostatistics, Princess Margaret Cancer Centre, University of Toronto, Toronto, ON, Canada, ³Department of Medical Oncology, Princess Margaret Cancer Centre, University of Toronto, Toronto, ON, Canada, ⁴Department of Otolaryngology-Head & Neck Surgery, Princess Margaret Cancer Centre, University of Toronto, Toronto, ON, Canada

Purpose/Objective(s): Identification of severe radiation-related late toxicity (RLT) risk factors in patients treated for head & neck cancers (HNCs) could guide mitigation strategies & follow-up routines. Impact of RLT on long term overall survival (OS) is also not well studied in HNC survivors. We investigate potential predisposing factors & association with long term outcomes for patient with severe RLTs. Materials/

Methods: We performed a retrospective case-control study, using data from our institutional prospective head & neck anthology of outcome system & case registry of the Adult Radiation Late Effect Clinic. We included HNC patients treated from 2003 – 2020 who received radiation (RT) = 50 Gy without any treatment failure. Those with documented RTOG grade 3-4 RLT (= 6 months post-RT) were classified as cases. Those without documented grade 3/4 RLT & follow-up of = 2 years served as controls. Cases could have one or more RLTs & date of RLT was recorded at the 1^st occurrence. Baseline characteristics were compared between cases & controls. Actuarial rates of LT & OS were calculated using competing risk & Kaplan-Meier methods, respectively. Multivariable logistic regression analyses (MVAs) were performed to identify risk factors (odds ratio, OR) for RLT & prognostic factors (hazard ratio, HR) for OS.
Results: Our analysis included 4797 patients (1048 cases & 3749 controls), with a median follow-up of 5.5 years. The most common RLTs observed were: osteoradionecrosis (n=355), dysphagia or feeding tube dependence (n=196) & fibrosis (n=140) which represented two thirds (66%) of RLTs. 5-year RLT rate was 18% (17-19). MVA identified the following potentially modifiable risk factors given in Table 1. Additional non modifiable risk factors included younger age [OR 0.99 per year (0.98-1.00) P=0.016], female sex [1.3 (1.1-1.5) p=0.01], advanced stage [TNM-7 stage II vs I: OR 1.6 (1.1-2.4) p=0.016, III vs I: 1.50 (1.01-2.2) p=0.045 & IV vs I: 1.7 (1.2-2.5) p=0.008 respectively] & primary site (vs oropharynx) [nasopharynx (OR 1.5 (1.2-2) p=0.001, oral cavity (2.3 (1.8-2.9) p<0.001) & hypopharynx (1.9 (1.3-2.7) p=0.002]. MVA for OS showed that RLT patients had higher mortality risk vs control group [HR 1.3 (1.1-1.6) p=0.001] after adjusting for age, gender, stage, site & performance status.
Conclusion: We found non IMRT treatment technique, neck irradiation, dose per fraction >2 Gy, use of chemotherapy & being a current smoker at diagnosis as significant modifiable & younger age, female sex, advanced cancer stage and primary site as significant non modifiable risk factors for developing RLTs. This information could be used to guide patient discussion, and select patients for RLT monitoring. Abstract 3701 – Table 1

Variable	OR (95% CI)	p-value
RT Tech (Non-IMRT vs IMRT)	2.4 (1.9, 3.0)	<0.001
RT Neck Volume		<0.001
Unilateral vs None	2.2 (1.5, 3.2)	<0.001
Bilateral vs None	2.7 (1.9, 3.9)	<0.001
Fraction Size (>2.0 vs =2.0 Gy)	1.5 (1.04, 2.1)	0.028
Chemotherapy (Yes vs No)	1.3 (1.1, 1.6)	0.009
Smoking (Current vs ex/none)	1.2 (1.0, 1.4)	0.048