KEY POINTS
- This retrospective analysis of a prospectively maintained single-institution registry included 65 adults with recurrent or second primary head and neck cancer re-irradiated with pencil-beam scanning intensity-modulated proton therapy between January 2020 and December 2025.
- 45 patients (69.2%) were treated with radical intent and 20 (30.8%) postoperatively; the median prescribed dose was 66 Gy (relative biological effectiveness) in 30 fractions.
- Median follow-up was 9.3 months; median overall survival was 17.2 months, 12-month overall survival was 67.1%, median progression-free survival was 10.1 months, and median locoregional control was 28.1 months.
- Central/skull-base disease had worse progression-free survival than peripheral disease — 9.1 vs 14.7 months, p = 0.037 — but anatomical location lost prognostic significance after T-stage adjustment.
- Grade 3 or higher oral mucositis occurred in 7.7%, no grade 4–5 acute events were recorded, and late osteoradionecrosis occurred in 6.2%, though late toxicity data remain preliminary because follow-up was short.
CLINICAL TAKEAWAY
Pencil-beam scanning proton re-irradiation appears feasible for selected patients with recurrent or second primary head and neck cancer, with a low acute severe toxicity rate in this cohort. The strongest prognostic signal was advanced T-stage at re-irradiation, not central/skull-base location itself. The evidence is worth attention, but it remains limited by single-institution design, heterogeneous cases, short follow-up for late toxicity, and no matched photon comparator.