Externally validated digital decision support tool for time-to-osteoradionecrosis risk-stratification using...

Area of Science:

• Oncology
• Radiation Oncology
• Oral and Maxillofacial Surgery

Background:

• Existing studies on osteoradionecrosis of the jaw (ORNJ) predominantly use cross-sectional data, limiting understanding of temporal risk.
• Determining the time-to-event profile for ORNJ is crucial for long-term oral health monitoring in head and neck cancer (HNC) survivors.

Purpose of the Study:

• To develop and validate a predictive model for the time-to-event profile of ORNJ.
• To identify key dosimetric, clinical, and demographic predictors of ORNJ onset.
• To facilitate clinical implementation through a user-friendly interface.

Main Methods:

• Retrospective analysis of a clinical observational cohort of 1129 HNC patients with radiotherapy (RT) and 198 ORNJ cases.
• Development and internal validation of a Weibull Accelerated Failure Time model using dosimetric, clinical, and demographic data.
• External validation on an independent cohort of 265 patients (92 ORNJ cases) and usability testing of a developed graphical user interface (GUI).

Main Results:

• The model identified male gender, pre-RT dental extractions, and D25% radiation dose as significantly associated with a faster onset of ORNJ.
• Adjusted time ratios indicated faster onset for males (0.62), pre-RT extractions (0.73), and D25% (0.88).
• The model demonstrated strong internal calibration (Brier score 0.133, D-calibration p=0.998) and discrimination (C-index 0.72 at 72 months).

Conclusions:

• This study is the first to establish a direct relationship between radiation dose and the time to ORNJ onset.
• The findings provide a novel characterization of how radiation dose and patient factors influence the timing of ORNJ, a late radiation effect.
• The model offers insights into conditional risk during survivorship, beyond just the probability of developing ORNJ.