Design: randomised phase II/III non-inferiority trial, final analysis
Population: 261 patients with postoperative high-risk locally advanced head and neck squamous cell carcinoma
Primary outcome: overall survival, non-inferiority margin hazard ratio 1.32
Effect: 5-year overall survival 71.2% weekly versus 58.7% three-weekly, hazard ratio 0.76 (95% CI 0.52 to 1.12) at median follow-up 5.6 years

JCOG1008 randomised 261 patients with postoperative high-risk locally advanced squamous cell carcinoma of the head and neck to chemoradiotherapy with cisplatin 100 mg/m² every three weeks for three cycles, or 40 mg/m² weekly for seven cycles. The interim analysis had already shown non-inferiority for overall survival; this is the final analysis at a median follow-up of 5.6 years.

Five-year overall survival was 58.7% in the three-weekly arm and 71.2% in the weekly arm, hazard ratio 0.76 (95% CI 0.52 to 1.12) against a pre-specified non-inferiority margin of 1.32. Non-inferiority is confirmed. Five-year relapse-free survival and local relapse-free survival were both numerically better with the weekly schedule, and no late adverse event differed between arms by more than 10 percentage points.

What matters here is that the long-term data did not undo the interim result, which is the usual worry with a non-inferiority trial reported early — a schedule that looks equivalent at two years can lose ground when late relapses accumulate. It did not, and late toxicity did not diverge either.

For Indian practice this confirms rather than changes what most units do. Weekly cisplatin is already the common schedule here, chosen for tolerability, for the ability to deliver it in day care, and because dose interruptions are easier to manage. That pragmatic choice now has five-year survival data behind it rather than only a tolerability argument. The point still to make in the multidisciplinary meeting is cumulative dose: the weekly schedule only works if the seven cycles are actually delivered.

Record planned and delivered cumulative cisplatin dose, not just the schedule — the weekly regimen depends on completing seven cycles
Use the five-year figures, not the interim ones, when counselling about survival
Do not read the numerically better survival in the weekly arm as superiority; the trial tested non-inferiority
Keep audiometric and renal monitoring on the weekly schedule; cumulative toxicity is the same drug
Confirm the high-risk indication — positive margins or extranodal extension — before committing to concurrent chemoradiotherapy at all

Why it matters

The schedule most units already use for tolerability now has five-year survival data rather than only a convenience argument.

The statistics, in plain English

A hazard ratio of 0.76 with an upper limit of 1.12, well inside the pre-specified margin of 1.32, is what confirms non-inferiority. The apparent 12.5 percentage point survival advantage for the weekly arm is not a demonstrated benefit: the confidence interval crosses 1.0, so the honest statement is 'no worse', not 'better'. With 261 patients, a superiority claim would need a far narrower interval. The absence of any late adverse event differing by more than 10 percentage points is reassuring but was not a formally powered comparison.