The basic rationale for the PROfound trial was to validate early observation that patients with metastatic castration resistant prostate cancers whose tumours harbour DNA response defects are likely to respond to PARP inhibition. That was a concept that was based on a fair amount of preclinical data, biological data, Stand Up to Cancer data, suggesting that a good percentage of patients will harbour these types of gene alterations. Then early data from clinical trials and, as it relates to olaparib, from the TOPARP trial. So that was the basic rationale. Obviously AR targeted therapy is one of the standards of care in this disease and the intent was to evaluate whether olaparib will be better than second line AR targeted treatment in the context of metastatic castration resistant prostate cancer.
Basically the clinical trial was designed, it’s a phase III clinical trial. At the time when we designed the trial the most well-characterised cohorts were patients with BRCA1, 2 or ATM gene alterations. So the trial pre-selected for patients who have the genomic alteration that might make their tumours likely to respond to PARP inhibition. There was a panel of 15 genes, not just the BRCA1 or 2 or ATM. Because the other genes we were not sure how well they were likely to respond to the PARP inhibition we created two cohorts. The patients were assigned to the cohort based on their genomic profile and the tissue was centrally reviewed using a Foundation Medicine assay, next gen sequencing assay. Then it was a random assignment, open label, two to one olaparib versus physician’s choice of abiraterone plus prednisone or enzalutamide. Then as patients were treated they were imaged periodically and if they developed disease progression by scans, based on a blinded independent central review of imaging, then they were allowed to crossover if they were on the control arm to provide them the access to the treatment. The primary endpoint was radiographic progression free survival in cohort A, which is the patients with the BRCA1, 2 and ATM and essentially there was a doubling of the median progression free survival with a 66% reduction in the hazard ratio of progression or death that was highly statistically significant in favour of the olaparib. Similar trends were seen in the overall study population, cohorts A and B. The trends in cohort B were positive also but the magnitude was not as big as the magnitude we saw in cohort A.
Then we have some data on gene by gene analyses that would suggest, in fact, some genes actually tend to predict for response to PARP inhibition. Obviously all of these are exploratory at this moment.
Other endpoints included objective response rate which also was significantly higher. It was 33% compared to 2% in cohort A, this measurable disease response rate. Time to pain progression, in fact the median has not been achieved in the olaparib arm whereas it was under a year in the control arm. The trend was very significant with a very significant p-value. Then obviously overall survival is very important, however, the data is not yet mature at this moment. But despite the fact that over 80% in both cohorts that were on the control arm crossed over to the olaparib arm, we actually saw very early separation of the survival curves and trends that are looking very promising.
What about the frequency of the genetic mutations?
I would refer, actually, the audience to the poster discussion session where yesterday the data was presented. Basically, the odds of having those types of mutations are in the high 20% depending on where the tissue is coming from. If it’s the primary it’s in the high 20%, slightly higher, I believe 30-31% in the metastatic disease tissue.
What is your take home message?
From the perspective of a physician who is practising in GU oncology, if in fact the drugs will be approved at the end of the day, this provides an additional option for patients with this disease. Clearly we know that giving abiraterone after enzalutamide or enzalutamide after abiraterone the response rates are not high and the odds of response are extremely low. So within this space this is actually a promising treatment with a clinically meaningful impact. Obviously patients have other choices – chemotherapy and other things – but this would be something that would be quite… what is the word I want to us… desired by patients because, again, you’re preselecting and not giving a random treatment. You’re preselecting for the odds of response and it’s an oral therapy that’s generally well tolerated. Obviously there are side effects there but generally well tolerated low grade side effects. I would say, from my end, this is likely to become the standard of care in patients who have progressed on frontline AR targeted agents.
With the assumption that genomic testing for prostate cancer becomes widespread?
Correct, absolutely. I think that the more we are doing genomic testing the more we are learning that a good number of patients might be eligible for targeted therapy. Having said that, at the end of the day you’d have to have the proof that that targeted therapy is going to work. So aside from clinical trials and providing access to patients to that, the positive thing about this which I think is very exciting is now targeted therapy can become part of standard of care in this disease.