The METEOR study was a phase III randomised to look at the activity of cabozantinib compared with everolimus in previously treated patients with TKI. Patients were clear cell histology patients who had failed one or more than one TKI before entering the study. They could have received PD-1 antibody, could have received cytokines, they could also have previously treated brain mets which is quite different from other studies. They were randomised between cabozantinib and everolimus under a one by one ratio. We had stratification factors based on prognostic factors and based on whether they received one or more than one TKI. The primary endpoint of the study was progression free survival in the first 375 patients enrolled and this was reported in Vienna some months ago and published in The New England Journal of Medicine. It showed clearly that this drug, cabozantinib, improves PFS, response rate and a good trend in overall survival compared with everolimus.

That’s what we recently published and what I have done in this meeting is to present an update of the study and present some subgroup analysis. So the first update we have done is to look at the progression free survival and response rate in the overall patient population, so not only the 375 but 658 patients. Interestingly the progression free survival is very similar to the one we had previously reported, 7.4 months versus 3.9 months. The response rate was 17% by independent review and 24% which is very consistent with what we previously reported at ESMO.

Then we looked at some subgroups, some subgroups of interest. We looked at series group, good, intermediate and poor risk. I would say that cabozantinib is very active in all subgroups with probably a little more activity in the good and intermediate risk group compared with everolimus. We looked at to zero and one which was to PS which was accepted in the study. Numerically the PFS is higher in PS0 patients than in PS1 but benefit over everolimus is very similar in both subgroups of patients.

Then we looked at some sites of disease of interest. One of the sites is bone mets because cabozantinib seems to have a specific on bones. Patients with bone mets, either alone or with visceral mets, have a very strong improvement in PFS compared with everolimus. The hazard ratio is very interesting, 0.22, with bone and so very impressive here.

We looked then at prior therapy so we looked at TKIs received, sunitinib, pazopanib especially. The PFS with sunitinib was 9.1 months, very impressive, which was reported at ESMO. It’s a little numerically lower with pazopanib but still very significantly better than everolimus in pazopanib treated patients. We also looked at one versus more than one TKI. Once again a very similar PFS, still in favour of cabozantinib. The last subgroup we looked at was patients who had received prior immunotherapy, PD-1 or PD-L1; not a very large group but still 32 patients. In this subgroup of patients the hazard ratio was 0.22, very, very strong in favour of cabozantinib which certainly gives an idea that this drug will be active in patients who have received, as an example, nivolumab which is becoming a standard of care in kidney cancer.

So overall a very nice subgroup analysis, all of them favour cabozantinib. Some interesting data on bone mets on prior immunotherapy.