This meta-analysis pooled phase III trials of immune checkpoint inhibitors added to chemoradiotherapy after resection of locally advanced head and neck squamous cell carcinoma.
Disease-free survival improved with a hazard ratio of 0.75 and overall survival with 0.74 — a 26% reduction in the risk of death. In patients with PD-L1 combined positive score of 1 or above, the pooled hazard ratio was 0.66, so the benefit concentrates where the biomarker is positive.
The toxicity is not a footnote. Any-grade adverse events affected 81% of patients, grade 3 or worse 43 to 45%, and 12 to 18% discontinued because of them. Immune-related events were significantly more frequent, with hypothyroidism prominent. In a population that has already had surgery and chemoradiotherapy, adding a treatment that stops one in six patients is a serious decision.
What this supports is checkpoint inhibition entering the postsurgical regimen for high-risk resected disease, with patient selection doing real work — CPS status, performance status, and whether the patient has the reserve for further toxicity. In Indian practice the constraint is cost before selection: checkpoint inhibitors are largely self-funded, and a 26% relative reduction in mortality needs to be presented alongside what it costs and what it does to the patient's remaining months, not instead of it.
- Test PD-L1 CPS: benefit was clearly larger at CPS 1 or above, HR 0.66 against 0.74 overall.
- Consent explicitly for a 43 to 45% grade 3 or worse event rate and a 12 to 18% discontinuation rate.
- Monitor thyroid function; hypothyroidism was the standout immune-related event.
- Weigh performance status after surgery and chemoradiotherapy — this is a third insult, not a first.
- Where funding is the barrier, present cost alongside the survival figure rather than after it.
The statistics, in plain English
A hazard ratio of 0.74 for overall survival means the rate of death at any given moment was about 26% lower, and the interval of 0.65 to 0.85 stays clear of 1.0. Hazard ratios describe the rate, not how much longer anyone lived — a 26% reduction can mean many extra months or few, depending on the underlying prognosis, and this analysis does not report median gains. The subgroup result at CPS 1 or above, HR 0.66, is the more useful clinical number but it is a subgroup: it was not the pre-specified primary comparison, and subgroups exaggerate more often than they mislead in the other direction. Weigh it against a grade 3 or worse rate above 40% in a population that has already been through surgery and chemoradiotherapy.
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