I want to talk to you about our journey, our journey of understanding that patients with newly diagnosed advanced ovarian, tubal and peritoneal cancer who respond to platinum-based chemotherapy are still at very high risk. Very. So we should keep them in remission using a very tolerable regimen. That started, published in 2003, GOG-178, which was a taxane. In fact, you remember that that study was stopped prematurely by the data safety monitoring committee because nine months of additional doses of paclitaxel improved progression free survival by seven months but it wasn’t enough.
Then you’re familiar with the evolution that in 2018 we got bevacizumab approved. Again, more tolerable in maintenance based on our study GOG-218, but it was not enough and there is no biomarker.
Then we got a biomarker – BRCA. Our third GOG study, GOG-3004, which you know as SOLO1, published in The New England Journal, evolved the standard and it was wonderful. Then the next study, PRIMA. PRIMA niraparib, again now in all comers, we got that labelled in April of 2020.
Now two years have gone by but questions remain. What is the best patient for maintenance treatment? What is the best agent? What is the best biomarker? So ATHENA is designed to answer some of these questions and at the ASCO 2022 meeting I presented ATHENA-MONO. As you may or may not know, ATHENA has four arms randomised 4:4:1:1. There is an arm which has nivolumab only in it which will not be analysed; it sorts out the relative contribution. But ATHENA-MONO takes rucaparib monotherapy, again in these patients who respond to front-line platinum-based therapy, and compares it to placebo. In a completely separate, appropriately powered, randomised placebo-controlled trial like MONO, COMBO will compare now rucaparib as the control arm to the experimental arm rucaparib nivolumab.
ATHENA-MONO, again which I presented on April 6th 2022 at ASCO and which was published simultaneously in The Journal of Clinical Oncology, had as its primary endpoint investigator assessed progression free survival in the HRD subset which was measured by FoundationOne CDx with an LOH score of 16% or greater. That study ultimately not only hit that endpoint but hit the intent to treat hierarchical step-down endpoint and it was sensational. Many of you probably enrolled to this; I want to share with you at a high level the endpoints.
If a patient is HRD positive and ultimately responds to platinum and gets rucaparib they have a 53% better chance of remaining in remission versus placebo. Again, if a patient receives rucaparib maintenance in this area, in this setting, the median time to recurrence is 28.7 months but if the patient gets placebo or observation it’s 11.3 months. 28.7 versus 11.3 months. You say, ‘Well, what about all comers?’ Well there still the hazard ratio is 48% better, a hazard ratio of 0.52, more than doubling of the progression free survival – 20.2 months on rucaparib versus 9.2 on placebo or observation. Really an amazing result and it was a celebration in Chicago.
We also analysed the blinded independent central review to add confidence. In fact, if you look at the HRD test negative group using this BRCA analysis, the hazard ratio is still 0.60, 6.4 months versus 12.0 months at the median. In other words if you have an HRD test negative patient, if she receives rucaparib maintenance treatment for up to two years she can have a 5.6 month improvement at the median.
This was consistent across all pre-specified prognostic subgroups. The adverse reactions were as has been seen with all PARP inhibitors – GI, nausea, fatigue, bone marrow such as anaemia. Even in the placebo group those patients also had some symptoms.
Now, rucaparib is delivered at four doses, 600mg, 500mg, 400mg, 300mg, and we showed that dose intensity can be maintained. Dose intensity is important with PARP inhibitors and in the intent to treat analysis 88% of the patients were able to be maintained on 80% or greater of the dose. This did not vary based on age or weight, again a very important discovery. We did notice, and confirm again, there are some idiosyncratic increases in liver function tests which resolved predominantly without dose interruption or dose reductions.
This maintenance treatment time used to be considered, ‘Oh, let’s stop treatment. Let’s give her a holiday.’ Well it’s still true.
If the patient receives rucaparib versus placebo in the HRD subset, is assessed by the investigator, there’s a 53% improvement in the risk. This is a proportional improvement in the risk, meaning that the tails of the curves do not come together. Meaning that if a patient responds to platinum-based therapy and has an HRD positive test they can live progression free for more than two years, 28.7 months at the median compared to placebo or observation 11.3 months. That’s dramatic. Even in the intent to treat analysis it is prolonged more than double with a 48% improvement, again a proportional hazard, 9.2 months with observation, 20.2 months in the rucaparib arm in the all comer subset.
We had a pre-specified assessment using a blinded independent central review. This was not analytic but it’s important because it adds confidence. For example, in the HRD test negative, or the HRP, subset, which is what we really care about because this is a high unmet medical need, there was also a 40% reduction in hazard ratio, meaning the hazard ratio is 0.60. Meaning that if a patient is HRD negative or HRP and gets rucaparib she can live a year without progression which is unprecedented because in the placebo it was 6.0 months at the median.
Now, the adverse reactions were not different than has been presented in other studies with the most common being GI, fatigue and bone marrow, predominantly anaemia. Most were low grade. I point out that 92.7% of even the placebo group had treatment emergent adverse events, this can be a symptomatic subset. There were two patients in the rucaparib group, one that had MDS and one that had AML for 0.4%.
Now, rucaparib does have some idiosyncratic low grade transaminitis, ALT, AST, that generally resolves without dose modifications.
Now, it’s all about the patient experience. Remember paclitaxel was too toxic, there was no biomarker for bevacizumab. So we want to maintain the patient symptom free. Symptom free from the medication but also symptom free because recurrence can cause life-threatening symptoms. So we measure that carefully using the FACT Assessment Of Cancer Therapy ovarian module – there’s no detriment. So the patient experience is maintained.
So, ATHENA-MONO showed that first line maintenance treatment significantly improves progression free survival irrespective of HRD status. Even in the HRD test negative or HRP subset it was statistically significant and clinically meaningful.
In fact, I didn’t show you, but measurable patients at the end of platinum-based chemotherapy continue to respond at a very high level and these responders are durable, way more than 50%. No new safety signals. No detriment in patient-reported outcomes. This was a relatively mature study, more mature than the other studies that I mentioned in my introduction, with more than two years of follow-up at the median. In fact, we already, although we don’t want this to happen, have some deaths and that’s encouraging that we’re impacting survival.
I want to thank my international colleagues, this is a GOG/ENGOT collaborative trial. Rebecca Kristeleit is the ENGOT lead; Japan also was a key contributor, NRG Oncology Japan with Keiichi Fujiwara. So thank you for your input. Hopefully this will be practice changing, adding more flexibility and more options in the frontline maintenance treatment. I encourage you to look at the paper which is published simultaneously on June 6th 2022. So long for now.