The landscape of definitive radiotherapy discontinuation: a 5-year institutional audit and clinical profiling

PURPOSE: Definitive radiotherapy discontinuation is a critical key performance indicator (KPI) in oncological care. This study aimed to determine the frequency of definitive radiotherapy discontinuation and to characterise the clinical profile of a cohort treated between 2020 and 2024. MATERIALS AND METHODS: We conducted a retrospective descriptive study of 16,292 fractionated radiotherapy courses initiated between January 2020 and December 2024. Definitive discontinuation was defined as the cessation of treatment before completing the planned dose for any reason. We analysed demographic, clinical, and dosimetric variables, and classified the reasons for discontinuation into predefined categories. Statistical analysis included descriptive statistics to characterise the discontinuation cohort. RESULTS: A total of 279 definitive discontinuations (1.71%) were recorded. The mean age was 66.0 ± 13.9 years (60.2% male). The discontinuation rate was significantly higher for palliative than for curative courses (3.49% vs. 1.08%). Causes of discontinuation differed significantly by intent: death from oncological causes predominated in the palliative group (49.0%), whereas non-oncological complications were the leading cause in the curative cohort (33.1%, p < 0.001). Treatment persistence analysis revealed that palliative courses were interrupted significantly earlier in their planned schedule (median completion 50.0%) compared to curative courses (64.3%, p = 0.012). A compromised performance status (KPS < 70) was present in 29.7% of courses. COVID-19 caused only eight direct discontinuations (2.8%). The 2020 temporal peak (2.45%) was driven almost exclusively by the palliative cohort (5.24%), while curative discontinuation rates remained stable (~ 1.0%). The discontinued cohort was clinically characterised by advanced age, compromised performance status, metastatic disease, palliative treatment intent, and treatment initiation delays. CONCLUSION: The definitive discontinuation rate was low compared with previously reported series. Oncological factors were the predominant causes, especially in palliative treatments. Differentiating between unavoidable events (e.g. rapid oncological deterioration) and potentially avoidable ones (e.g. toxicity or patient decision) transforms this metric into a practical quality indicator. Although direct COVID-19 discontinuations were minimal, the pandemic highlighted the vulnerability of palliative pathways during systemic crises. When appropriately stratified and interpreted, this indicator may provide a framework for inter-institutional comparison and benchmarking, supporting targeted quality improvement strategies such as optimisation of patient selection, increased use of hypofractionated regimens in limited-prognosis patients, and enhanced supportive care for high risk and vulnerable subgroups.