The ANITA trial was a randomized phase 3 trial that explored the incorporation of atezolizumab to standard therapy with platinum-based chemotherapy followed by niraparib in platinum-sensitive recurrent ovarian cancer. Unfortunately, the study didn’t find any significant efficacy of adding a desolation map, but we have many exploratory objectives, including biomarker research...
The ANITA trial was a randomized phase 3 trial that explored the incorporation of atezolizumab to standard therapy with platinum-based chemotherapy followed by niraparib in platinum-sensitive recurrent ovarian cancer. Unfortunately, the study didn’t find any significant efficacy of adding a desolation map, but we have many exploratory objectives, including biomarker research. One of the exploratory objectives was to figure out if a baseline plasma proteins could predict somehow the benefit or even the prognosis in this study. And what we performed was a multiplexing immune assay of 74 proteins covering essentially proteins related to immune modulation, angiogenesis, as well as apoptosis. We have done a univariate analysis first, then a second step was a multivariate COX regression analysis looking for an overall survival modeling using a tenfold cross-validation. What we found is that there are four proteins associated with worse prognosis in terms of progression-free survival and overall survival that were VISTA, CD31, VEGFA and also UPAR. And on top of that LOX1 was also associated with the worst overall survival. Then we integrate these proteins along with other clinical factors in our multivariable COX regression model. We need a 10-fold cross-validation. That means that only those markers that were consistently predictive of prognosis in at least six-fold were selected for our final overall survival model. Our final overall survival model was highly predictive of overall survival with a C index of 0.69 and especially it was very predictive at six months with an AUC of 0.87. It is remarkable that we have identified VISTA as the most relevant prognostic factor. And indeed, when we have analyzed the quartiles of the expression and we have classified by quartiles and correlated that expression with the outcome of the patients, we saw that the patients with the highest expression of VISTA in the fourth quartile had a much worse prognosis compared with the others. Indeed, at six, 12 and 24 months, there was a huge difference with overall survival so in summary I think that we can conclude that there are two important proteins associated with the worst prognosis in this population one is VISTA the other is VEGFA I think that this information could be used for future designs of research in this setting. And we are now in the process of validating our OS modeling as well as these findings in other independent cohorts. But it is remarkable that in this academic study, they were able to have access to the sample of the patients and to provide this important information for the community.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
