JCOG1008 is a phase II/III trial comparing three weekly cisplatin with weekly cisplatin for post-operative high risk patients with squamous cell carcinoma of head and neck. The patients with post-operative high risk factors, namely incomplete resection or extranodal extension are randomised in a one to one fashion. Arm A was three weekly cisplatin plus radiation and Arm B was weekly cisplatin plus radiation. In total 261 patients were enrolled in this trial.

The patient characteristics are generally well balanced and treatment delivery and compliance is also sufficient. The median radiation dose was 66Gy in both arms and a median total dose of cisplatin in the three weekly arm was 280 and 239 in the weekly arm. So the treatment delivery was sufficient in both arms.

As a result, for the primary endpoint of overall survival, Arm A with three weekly cisplatin plus radiation and Arm B with weekly cisplatin plus radiation, the weekly arm appeared to be better than the three weekly arm with a hazard ratio of 0.69 and the 99.1% confidence interval was 0.37-1.27 which was below the pre-specified non-inferiority hazard ratio of 1.32. The one-sided p-value for non-inferiority was 0.00272 which was also below the pre-specified p-value which was 0.00433. Hence, at the second planned interim analysis in the phase III part the Data Safety Monitoring Committee recommended terminating the trial early. As a result, the weekly cisplatin plus radiation was proved to be non-inferior to three weekly cisplatin.

Looking at the acute toxicities, the weekly arm appeared to be a better safety profile in terms of neutropenia, and dysphagia, nausea, infection, hyponatremia, renal impairment and hearing impairment, all were less frequent in the weekly arm.

So we can conclude this trial, the weekly cisplatin plus radiation in the post-operative setting is at least non-inferior to three weekly cisplatin plus radiation. However, this trial stopped early with the interim analysis so we have to observe further, for up to five years, for the final analysis.

Were there any important discussion points raised?

In the discussion during the ASCO virtual meeting some investigators pointed out some imbalance of the patient characteristics, for example T4 disease is more frequent in the three weekly arm and oropharyngeal cancer is more frequent in the weekly arm. So this may affect the results of this trial but the difference is not so robust. So I think this difference between the patient characteristics between arms is not such a big effect for the result but we can understand this criticism. We will perform the analysis with these factors adjusted.