Once head and neck cancer has recurred or become metastatic, for the majority of patients there are not curative local therapies and the majority of these patients receive systemic therapy, usually chemotherapy or cetuximab as we heard, and can expect a median survival of under a year. The standard regimen is the EXTREME regimen, which is a combination of cisplatin, 5-fluorouracil and cetuximab, but it is recognised that the immune checkpoint inhibitors, including pembrolizumab and nivolumab, have activity in head and neck cancer and both pembrolizumab and nivolumab are FDA approved in the second line setting.

In thinking about how to begin to study immune checkpoint inhibition in the first line setting, that is to say as the first systemic therapy that a patient receives when the cancer recurs or becomes metastatic, we considered the fact that PD-L1 expression in tumour has been associated with a higher response rate, better sensitivity to immune checkpoint inhibitors as well as the fact that chemotherapy has been proposed as a rational combination with immune checkpoint inhibitors – it can disrupt tumour architecture and maybe improve accessibility of the immune system to the tumour and it induces rapid disease control and has a higher response rate than immune checkpoint inhibitors given as monotherapy.

So we wished to determine whether pembrolizumab alone or in combination with chemotherapy would be superior to treatment with the standard of care regimen, EXTREME, in an all comers patient population as well as looking at those patients who were biomarker enriched because their tumours expressed the marker PD-L1. The primary endpoints of the study were overall survival and progression free survival in patients with a PD-L1 combined positive score of greater than or equal to 20; a combined positive score of greater than or equal to 1 and all patients enrolled. Let me just define for you what the combined positive score was: it was a ratio of all the PD-L1 positive tumour cells, lymphocytes and macrophages to the total number of cells that you encountered times 100.

Patients were eligible who had squamous cell carcinoma of the oral cavity, oropharynx, larynx or hypopharynx whose recurrent metastatic disease was not amenable to curative local therapy. They had to have good performance status; tissue had to be available for PD-L1 testing and if the cancer had arisen in the oropharynx they had to have known p16 status because, as we just heard from Dr Mehanna, those patients with HPV associated cancer have a better prognosis. Patients were stratified by their tumour’s expression of PD-L1, their p16 status and their ECOG performance status, and they were randomised one to one to one to pembrolizumab monotherapy, to a novel regimen of pembrolizumab together with a platinum chemotherapy, 5-fluorouracil, followed after six cycles by pembrolizumab monotherapy or to the EXTREME regimen given for six cycles followed by cetuximab monotherapy.

What we found was that in patients who had elevated biomarker CPS 20 using pembrolizumab alone as first line therapy was superior to the EXTREME regimen with an improvement in median survival, a hazard ratio of 0.61 that was statistically significant. Median overall survival improved from 10.7 to 14.9 months and the benefit was durable – at two years survival was 22% with the EXTREME regimen and 38% for those patients who had started with pembrolizumab monotherapy.

Similarly in patients who were biomarker enriched with having a CPS of 1 or higher pembrolizumab was superior to the EXTREME regimen extending median survival from 10.3 to 12.3 months. Again the effect was durable so at two years 19% of EXTREME treated patients and 30% of pembrolizumab treated patients were alive. The hazard ratio here was 0.78 and it was statistically significant.

Turning now to the comparison of pembrolizumab with chemotherapy, so the regimen of pembrolizumab chemotherapy compared to the EXTREME regimen, this time in the total population without biomarker selection, the novel regimen was also superior to the EXTREME regimen extending median survival from 10.7 to 13 months, again this benefit persisted at two years with 19% of EXTREME treated patients and 29% of patients treated with the novel regimen alive. The hazard ratio was 0.77 and this was statistically significant.

This is an interim analysis that was pre-planned at a certain duration of follow-up. There are further analyses still to be forthcoming including whether or not pembrolizumab monotherapy is superior to the EXTREME regimen in a biomarker unselected patient population. At this interim analysis we know that pembrolizumab was non-inferior to the EXTREME regimen.

Neither pembrolizumab nor pembrolizumab plus chemotherapy improved response rate over the EXTREME regimen but when patients responded they had a longer duration of response. Pembrolizumab alone had a better safety profile than the EXTREME regimen and pembrolizumab plus chemotherapy had a comparable safety profile to the EXTREME regimen.

So, pembrolizumab alone and pembrolizumab given with platinum and 5-fluorouracil should represent new standards of care for the first line treatment of recurrent metastatic head and neck carcinoma. Immune checkpoint monotherapy with pembrolizumab allows patients to live longer and has a better safety profile than the previous standard of care for those patients whose tumours express PD-L1. Adding the immune checkpoint inhibitor pembrolizumab to platinum-based chemotherapy has a comparable response rate and allows patients to live longer compared with the standard of care. Pembrolizumab can be safely added to platinum-based chemotherapy. There are further analyses of biomarker and clinical predictors that will be forthcoming from this study and these may eventually optimally guide the choice of whether to administer pembrolizumab alone or in the novel combination.

Obviously this work would not have been possible without the participation of the 882 brave patients who enrolled and we thank them and their families.