We know that MSI metastatic colorectal cancers are a little bit particular because the prognosis and the chemo sensibility is a little bit less compared to the microsatellite stable disease. It’s also a rationale to use immuno-oncology and anti-PD1 because MSI has a lot of mutation, a lot of infiltration with lymphocytes. It’s clear that it’s a very particular tumour with a lot of neoantigens and a very high mutation tumour burden, that’s the reason. The story began now five years ago with the first paper by the John Hopkins in the US with the incredible New England Journal of Medicine and after we have a lot of data. We have a lot of trials with pembrolizumab, with nivolumab, with nivolumab and ipilimumab, with good efficacy in MSI high but refractory patients. It was necessary to have a phase III to have labelling in first line, to have labelling in Europe and KEYNOTE-177 is this trial.

What was the trial design?

The design of the trial was very simple, it was a randomisation, one to one, pembrolizumab 200mg every three weeks versus chemo – FOLFOX or FOLFIRI plus or minus cetuximab and bevacizumab. Two years of pembrolizumab therapy, for chemo, like pembrolizumab, and we stopped in the case of toxicity or in the case of progressive disease. The primary endpoint was a dual primary endpoint, it was progression free survival and overall survival by a blind independent review committee for the CT scanner. It was a variation by RECIST 1.1.

What were the results?

The results are three good news for the patients. The first one is the progression free survival is better in the pembrolizumab arm versus the chemo arm, it’s 16.5 months in the pembrolizumab arm versus 8.2 months in the chemo arm. It’s a hazard ratio of 0.6 and a very significant p-value. It’s really rare to see such a large difference between the two curves of PFS in metastatic colorectal cancer. This good PFS is also after 24 months of follow-up 48% of patients in the pembrolizumab arm are without progression and can stop the therapy, in the chemo arm it’s 90% and for the majority of patients in the chemo arm, patients that continue on other therapies except in the case of surgery of metastasis. It is the first good news for the patients.

The second good news is the toxicity profile because the pembrolizumab is 200mg and it is well tolerated. The toxicities, we have 22% of grade 3/4 versus 66% for chemo and it’s really a huge difference. With pembrolizumab it is immune toxicity, with chemo it’s the classical with nausea and vomiting, neutropenia, alopecia, normal toxicity. But with pembrolizumab it’s pretty manageable, we don’t have toxic deaths. It was really the classical toxicity, easily manageable.

The third good news for the patients is the fact that the convenience of the therapy is very good for pembrolizumab because it is a one hour injection every three weeks compared to chemo which is really more complex with a half day in hospital, a pump of 5FU and all the possible toxicities the day of the therapy with the chemo and the targeted therapy.

What are you concluding from these results?

The main conclusion is the fact that pembrolizumab provides a clinically meaningful and statistically significant improvement in progression free survival in MSI high colorectal cancer and should be the new standard of care for these patients.

Is the final overall survival analysis still to come?

The overall survival was the other primary endpoint of this trial. The steering committee asked us to wait, it’s too early, not mature. It is planned if 180 events occurred or one year after the PFS analysis. Probably it is for next year and it’s important to underline the fact that in this trial it’s a crossover for pembrolizumab in the chemo arm and patients with progressive disease in the chemo arm will receive pembrolizumab or other immuno-oncology if it’s out of trial. At this time we know that approximately 55% of the patients in the chemo arm received anti-PD1 or anti-PD-L1 and for the interpretation of the overall survival it will be important to take this point into consideration.