- Design
- PRISMA systematic review with descriptive synthesis
- Population
- 8 studies of treatments for radiation-induced fibrosis after head and neck cancer, from 4078 screened
- Primary outcome
- Fibrosis improvement by study-specific measures
- Effect
- Pentoxifylline-tocopherol: 83% achieved ≥50% reduction in fibrotic surface area (uncontrolled); pravastatin ≥30% thickness reduction in 35.7%
A systematic search of four databases for treatments of radiation-induced fibrosis after head and neck cancer returned 4078 records and ended with eight studies. That number is the finding.
What exists is promising and uncontrolled. In one uncontrolled trial, 83 per cent of patients on pentoxifylline-tocopherol achieved a 50 per cent or greater reduction in fibrotic surface area; pravastatin reduced fibrosis thickness by 30 per cent or more in 35.7 per cent. Amifostine and topical superoxide dismutase were also reported. On the procedural side, lipotransfer was associated with physical improvement in 97 per cent of 38 patients and microcurrent therapy improved cervical mobility in 92 per cent of 26.
The outcome measures differ between every study — surface area, thickness, mobility, patient report — so nothing can be ranked against anything else. Pentoxifylline-tocopherol remains the reasonable first thing to try because it is cheap, oral and the best documented, not because it has been shown superior.
- Offer pentoxifylline and tocopherol as a first-line trial, and record a measurable baseline before starting.
- Set a review point — three to six months — rather than continuing indefinitely.
- Refer for structured physiotherapy alongside any drug; the mobility studies are as strong as the pharmacological ones.
- Discuss lipotransfer with the reconstructive team for localised, functionally limiting fibrosis.
- Tell patients honestly that the evidence base is eight small studies.
Why it matters
A common, disabling late effect of treatment has almost no controlled evidence behind anything offered for it.
Don't overread it
Most of the improvement rates come from uncontrolled series; none establishes that one treatment is better than another.
The statistics, in plain English
Response rates from uncontrolled series — 83 per cent, 97 per cent, 92 per cent — have no comparator, so they include whatever would have happened anyway, including the natural course and regression to the mean in patients recruited at their worst. With eight heterogeneous studies and different outcome measures, no pooled estimate was possible and none should be inferred.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for ent & head and neck surgery, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free