Radiation-induced optic neuropathy following proton therapy for intracranial, skull base, and head and neck tumors in adult patients: a systematic review and meta-analysis

Abstract Introduction Radiation-induced optic neuropathy (RION) is a rare but serious complication of radiotherapy, leading to potentially irreversible vision loss. This systematic review and meta-analysis aimed to determine the prevalence of RION in adult patients treated with proton therapy for intracranial, skull base, or head and neck tumors, and to identify predictors of RION. Methods Studies published after year 2000 reporting on the prevalence of RION in ≥ 10 patients who underwent proton therapy for the aforementioned tumors were eligible for inclusion. Study quality assessment was performed using the ROBINS-I tool. The primary outcome was the prevalence of RION (any grade). The secondary outcomes were predictors of RION. Results A total of 29 studies comprising 2385 patients were included, of which 1 study was prospective. Treatment modalities were proton-only radiotherapy ( n = 15), planned combined proton-photon radiotherapy ( n = 11), or unplanned combined proton-photon radiotherapy ( n = 3). The most frequently included tumors were head and neck cancers ( n = 10), meningiomas ( n = 7) and chordoma/chondrosarcoma ( n = 4). Overall risk of bias was serious in 11 (37%) studies or moderate in 18 (63%) studies. Median follow-up time was 45 months (interquartile range [IQR]: 30–62). The pooled prevalence of RION was 5% (95% confidence interval [CI]: 4–7%). Subgroup analyses showed a significant higher prevalence of RION in studies published between 2000 and 2014 vs. 2015–2025 (8% vs. 5%, respectively, p = 0.032) and in proton-only radiotherapy compared to planned or unplanned combined photon-proton radiotherapy (7% vs. 4% and 2%, respectively, p = 0.025). Conclusion RION occurs in approximately 5% of patients after proton therapy. The pooled estimate should be interpreted cautiously because of substantial clinical and methodological heterogeneity between studies.