I just presented the results of the ATALANTE trial. ATALANTE was an ENGOT trial assessing the combination of atezolizumab, which is an anti-PD-L1 antibody plus bevacizumab and carboplatin-based chemotherapy in platin-sensitive relapse of ovarian cancer. This was an international randomised study that included more than 600 patients. We had actually two co-primary endpoints that were the PFS, progression free survival, in both the ITT population and the PD-L1 positive population. So this was the first randomised trial of immunotherapy plus antiangiogenics in this setting of platinum sensitive relapsed disease.

What was the methodology of the study?

It was a randomised phase III study and we stratified patients according to the PD-L1 status with a threshold of 1%, the chemotherapy arms and the platinum-free interval: 6-12 and more than 12. We performed the PFS analysis and actually during the study we found that an unexpectedly low rate of patients had PD-L1 positive samples. All patients had a biopsy before entering the study. So we decided to increase the enrolment from 405 to 614 patients in order to have power enough to conclude. 

The study went very well and we had data cut-off by the end of last year. I just presented the first results at this ESMO meeting.

What have you found so far?

Unfortunately the trial is negative as we couldn’t find a significant difference in between arms in both the ITT and the PD-L1 positive populations. We performed a lot of subgroup analyses to identify a subset of patients in whom a small or large benefit could be found but it was always negative. We also presented the TSST results with no difference and the overall survival, although not mature, showed maybe a trend in favour of the experimental arm. But since the methodology did not allow to test overall survival if the primary endpoint was negative we cannot conclude and we won’t be ever able to conclude either with maturity of the data. 

Are you able to speculate on why the results occurred the way they did?

I don’t really know, perhaps immunotherapy just doesn’t work. This is not my feeling; my feeling is that it works but we are still unable to identify those patients in whom we can find a benefit. My thinking is that immunohistochemistry is not enough and hopefully we will be able to do a lot of translational research because we have biopsy samples for almost all patients. So we have plenty of material and now it’s time to think and try to identify maybe a molecular signature or something that might predict the efficacy of immunotherapy.

What can we do with these findings? What can this data be used for in the future?

I don’t really know when we can do that and if immunotherapy will be, again, on the ground ever in ovarian cancer. We look forward very much on having the results of the ATHENA combo that will be presented hopefully next year. It’s a different study but it’s the last phase III study that is currently running with IO drugs in ovarian cancer. So far, the JAVELIN 100 and 200 were negative, the IMagyn050 was negative, now ATALANTE is negative, that’s a lot. 

What is promising in ovarian cancer at the moment?

We had many different results from maybe ADCs that are coming, especially folate receptor. I think that the whole landscape of treatment is going to change. Interesting data also on the overall survival with SOLO1 and PAOLA-1, so it has strengthened a lot the efficacy of PARP inhibitors which is one side of the coin and immunotherapy is the other side of the coin. But so far PARP inhibitors are much more in the light, of course.