Ramez N. Eskander, MD: What we know about PARP inhibitors is not only do we think about them as a maintenance strategy, but many of us think about ovarian cancers in remission as microscopic tumor sites that reside in the peritoneal cavity. So, maintenance really has some treatment component to it. Additional studies that were done looked at efficacy in a prespecified patient population, in a cohort of patients who emerged from Study 10, a rucaparib study, as well as a cohort of patients who were adopted from ARIEL2 Part 1. These were patients with BRCA mutations, somatic or germline.
In that cohort of patients for efficacy data, they saw an overall response rate of 54%. In addition to that, 61% had greater than or equal to 3 prior lines of therapy. So you have a heavily pretreated patient population with an objective response rate above 50%, which is really exciting data for us. As we discussed previously, with each disease recurrence the magnitude of benefit declines. If we have patients who have had 3 or more prior lines of treatment with an overall response rate of greater than 50%, it suggests efficacy of this regimen even in that population.
There are additional data important to highlight when we talk about efficacy of rucaparib. The ARIEL3 study, which was in the platinum-sensitive recurrent setting, allowed patients who had a partial response with what we call bulky disease, or disease left over that was greater than 2 cm after completion of their cytotoxic chemotherapy. Even in that population, we saw a meaningful response rate, suggesting that the PARP inhibitor is effective therapeutically even for those patients who have measurable disease that remains after the cytotoxic combination chemotherapy is completed.
Loss of heterozygosity [LOH] is one of the markers that we would consider reflective of genomic instability. The theory behind the assays that we do to evaluate for what we call homologous recombination deficiency, which LOH, or loss of heterozygosity, is a component of, when we are doing these assays, is that we are trying to determine whether we’re going to identify a vulnerability in the tumor that helps identify patients who may benefit from a PARP inhibitor such as rucaparib. Again, ARIEL2 Part 1 specifically incorporated an LOH assessment in that population with a cutoff of 14%. That was revised in ARIEL3.
The relevance of this is that patients who have an elevated LOH score or who have homologous recombination deficiency appear to benefit or have a greater magnitude of benefit over what we would call a biomarker-negative population. Patients who are BRCA mutated by definition are homologous recombination deficient. So the LOH score doesn’t really apply to a patient who has a known germline or somatic BRCA mutation.
But there is a caveat here, and that caveat stems from the fact that this is not a perfect assay. LOH is not predictive of response with a sufficient sensitivity to define patients who would benefit only by that mode of assessment. There are patients who are BRCA wild-type and who are LOH low or homologous recombination proficient who will still benefit and have a meaningful response. In fact, in the ARIEL3 study, I think among that cohort 30% were responders to rucaparib versus 5% in the placebo arm. So you can see that it’s not perfect in predicting response, because patients who are LOH low and BRCA wild-type may still respond to PARP inhibition. That ties to the FDA approval of niraparib in the frontline setting for all-comers, which is based on the data that were released after the PRIMA trial.
Transcript edited for clarity.
Case: A 71-Year Old Woman With High-Risk Ovarian Cancer
Treatment and Follow-Up