The KEYNOTE-427 study is a clinical study of frontline pembrolizumab therapy for renal cell carcinoma patients and there was a cohort A for clear cell histology and cohort B for non-clear cell histology. The clinical endpoints have already been presented previously at ASCO: in the cohort A population the overall response rate 37%, 3% complete response rate. In this year’s ASCO this data was analysed by looking at the depth of response of patients in this cohort and associating different stratifications of the depth of response with overall survival. So patients that had primary refractory disease with progression, so greater than 0% change in tumour metrics by RECIST analysis, 0-30% regression, 30-60% regression, 60-80% or 80-100% regression and then the complete responders as well and assessing these different subsets of the patients versus overall survival.

As you begin to break down a 110 patient cohort into these different subgroups the numbers do become fairly small but the trend of the data suggests that patients with deep partial responses fare extremely well and their overall survival corresponds closely to the complete responders. There was a separate report at ASCO on the cohort B analysed in the same way and it again showed the same trend, that depth of response does seem to associate with overall survival.

Perhaps a very intuitive endpoint but, as clinicians know, many of our patients that fare extremely well do not achieve a complete radiographic response. So trying to provide some data to comment on the very good partial responders and the natural history, the trajectory of survival for these patients, if you don’t achieve a complete response can you not still fare extremely well? Perhaps overall survival will rival the complete responders that we hope are the patients that have the chance to perhaps be truly cured of their disease.

These drugs remain fairly new in the field and so we don’t have extremely long-term perspective of 10-20 years of follow-up. But this sort of analysis of depth of response stratification will carry forward. As the data matures in this cohort it will be used in other similar cohorts to begin looking at patients that do not achieve a complete radiographic response but what is the natural history of these different subsets.

So food for thought and hypothesis generating for future studies in the field.