S557
Clinical – Head & neck
ESTRO 2026
curves, and Brier score. Decision curve analyses were conducted to evaluate the clinical utility of the models across different threshold probabilities. Results: A multicenter cohort of 732 HNC patients with long- term thyroid function data was assembled, with a median follow-up duration of 7.2 years. Ten NTCP models were identified from the literature for external validation, including six logistic regression models, two Lyman-Kutcher-Burman models, and two nomograms. The performance of the candidate models varied, with accuracies ranging from 0.462 to 0.755 and AUC from 0.570 to 0.796. Nine models systematically underestimated hypothyroidism risk, particularly at low predicted risk ranges (Figure 1). Models incorporating thyroid function reserve metrics, such as thyroid volume or baseline TSH level, demonstrated superior calibration and discrimination. One model which included age, mean thyroid dose, and baseline TSH level, achieved the highest accuracy (0.755), highest discrimination (AUC 0.796) and lowest Brier score (0.173). Decision curve analysis indicated a net benefit over treat-all or treat-none strategies at threshold probabilities above 10% (Figure 2).Figure 1
definitive randomized results are available. Future progress will likely depend on integrating high- precision radiotherapy with molecularly targeted and immunotherapeutic approaches and on adaptive trial designs capable of accommodating the marked biological heterogeneity of salivary gland carcinomas. Keywords: Salivaly gland, prediction, 1008 Mini-Oral 1154 Multicenter validation of NTCP models for radiation-induced primary hypothyroidism in head and neck cancer survivors with long-term endocrine outcomes James C.H. Chow 1 , Kenneth C.W. Wong 2 , Jing Cai 3 , Jeffrey C.F. Lui 1 , Yiu Wing Cheung 1 , Jiaming Wu 3 , Anthony H.P. Tam 1 , Gavin T.C. Cheung 4 , Ka Man Cheung 1 , Kwok Hung Au 1 , Victor H.F. Lee 5 1 Department of Clinical Oncology, Queen Elizabeth Hospital, Hong Kong SAR, China. 2 Department of Clinical Oncology, Prince of Wales Hospital, Hong Kong SAR, China. 3 Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong SAR, China. 4 Department of Oncology, United Christian Hospital, Hong Kong SAR, China. 5 Department of Clinical Oncology, The University of Hong Kong, Hong Kong SAR, China Purpose/Objective: Numerous normal tissue complication probability (NTCP) models have been developed to predict radiation-induced primary hypothyroidism, but their external validity and clinical applicability remain unclear. This multicenter study aimed to externally validate and compare published NTCP models using long-term endocrine outcomes and to assess their predictive performance and clinical utility. Material/Methods: A comprehensive systematic literature search was conducted using standardized search strategies to identify published NTCP models for radiation-induced primary hypothyroidism. External validation was performed using a multicenter cohort of head and neck cancer (HNC) patients treated with intensity- modulated radiotherapy at four oncology centers. Cases were defined as patients with elevated serum thyroid-stimulating hormone (TSH) levels, while controls were patients who remained biochemically euthyroid for at least 5 years following radiotherapy. Clinical data and radiation dose-volume parameters of the thyroid gland were extracted. Model validation was conducted following the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines. Model predictive performance was compared using multiple metrics, including accuracy, discrimination, calibration
Made with FlippingBook - Share PDF online