Efficacy and Safety of Olaparib and Niraparib in Switch Maintenance
Robert L. Coleman, MD: We have benefitted through 3 completed phase III trials and an important phase II randomized trial that demonstrated the efficacy of these compounds—the PARP inhibitors—in women who have had a response to platinum-based therapy. These trials all fit under the concept of switch maintenance, which is the idea of giving a patient a platinum-based regimen and seeing the response at the point of maximal response, which was defined in the trials as somewhere between 4 and 8 cycles of treatment, when the patients were randomized to a PARP inhibitor or placebo. Then, they were monitored for first progression. There’s many other exploratory endpoints that have been studied in these trials, but those were the primary outputs for all of these trials.
The first of these 4 trials was a randomized phase II trial. It was a large one, called Study 19, and it involved a capsule formulation of olaparib. While it was undergoing an evaluation to switch from the many capsules that needed to be taken to a tablet formulation, which was more potent or concentrated, a large study was done where they randomized patients who had platinum sensitivity to olaparib versus placebo. It was almost earth shattering for us. We had never seen, in a randomized trial, the hazard ratios as low as they showed in this particular trial. And, when they went back, retrospectively, and found the patients that had a germline mutation, we saw hazard ratios that were less than 0.2. As I said, we have never seen something like that.
We knew that these drugs should work in BRCA patients, but this was done in all-comers and it was assessed by investigators. We found that this drug seemed to almost double the time to progression in patients who had received it instead of placebo. This data was actually brought to the FDA, and underwent a full ODAC (Oncologic Drugs Advisory Committee) review. And, because of a number of pretty important points, ODAC recommended against approval of it as a maintenance therapy. One of the major reasons that ODAC opined against the drug was that there was an ongoing trial, SOLO-2, happening. SOLO-2 was a phase III trial in a more selected patient population. Only about 4 months later, we actually got the treatment approval of the compound olaparib, as I mentioned earlier, in fourth-line and beyond patients who carried a germline mutation.
So, that story was evolving back in 2014, and we were pretty happy with that. Ongoing, at the time, were 3 other trials. The first to report was the NOVA trial, which looked at another PARP inhibitor, niraparib. This trial was set up in a similar way to Study 19, in that it took all-comers. But it had really embedded 2 trials into 1. In one cohort, they looked specifically at the germline cohort of patients—patients who had a germline mutation of BRCA. The second cohort was everybody else. That included patients who had a somatic mutation, those who had what was called homologous recombination deficiency (HRD) not due to BRCA, and those who were wild-type, negative for everything. So, this was 2 embedded trials. In these trials, they used what’s called a blinded independent central review for radiology. All of the imaging studies were sent out and reviewed externally, and they reported on the results of this at a meeting that occurred in 2016, which was hosted by the European Society of Medical Oncology.
While it was a big news event, many of us who have been watching this field expected this to be a very positive trial, particularly in the patient population who matched what we saw in Study 19 with olaparib. But it was an important advance. It was a phase III trial. It was adequately powered, and it clearly demonstrated that they work and they work big in the patient population that are most vulnerable to these drugs. The hazard ratios were quite small. They were between 0.2 and 0.3. And, for the groups that were less strictly defined, such as those that had the homologous recombination deficiency, a somatic BRCA mutation, or were wild type, they also seem to benefit.
Most surprising, there are 2 interesting factors from that trial. One was that for the patients who were not germline, were not somatic, and were not HRD, they were pure wild type, and this was a large patient population, about half of the cohort in that group still seemed to benefit. That tells us that our tests for assessing homologous recombination deficiency probably still are not perfect, and they still need to be further optimized. In that particular trial, they used an assay that looked at 3 components to determine whether or not homologous recombination deficiency was present. But, it obviously is still not perfect.
The other important finding, which also recapitulated one of the earlier trials I mentioned with rucaparib and somatic and germline mutations, was that if you looked at the somatic mutation in the NOVA trial, and you looked at the germline patients in the NOVA trial, they had overlapping progression-free survival curves. This would make sense because a germline patient likely has a somatic event. And, for a somatic patient without a germline mutation, one would have the same mutation. So, we expected to see that those were the case. But, it confirmed that initial observation that was seen, and it ultimately reported with the 2016 approval of rucaparib.
So, NOVA was a very positive trial for efficacy. I think what we were also very interested in, as was the FDA, what the toxicity was. One of the major toxicities that we were all very concerned about, and it did come to the table when ODAC reviewed Study 19 with olaparib, was, what is the risk of a secondary malignancy, ie, the MDS, AML? Those secondary malignancies were actually very low and continued to be reported as very low across all of the phase III trials that have been done. This reassures is that the particular risk didn’t seem to occur, despite the fact that many of these patients had very prolonged exposure to the drugs. That was good.
The second thing we learned, and this will, I think, emerge when we start talking about all of the phase III trials, was that they found a little bit of a difference and some similarities in terms of the toxicities. So, for NOVA, with niraparib, there was a high degree of gastrointestinal side effects, which is what we would expect to see and have seen with every one of the trials that we’ve done. We saw a slightly higher rate of myelosuppression, predominantly in the form of thrombocytopenia, in this trial. Many people were a little alarmed with this because grade 3/4 thrombocytopenia does cause the need for some type of medical intervention and, if nothing else, closer surveillance.
But, it’s interesting. When you go back and look at the trials that were done for establishing the maximum tolerated dose of the drugs, they all had very similar types of problems with myelosuppression when they received high doses. I think that with NOVA and the niraparib dose, maybe they’re at an upper limit or an upper end of the tolerability of that drug? And, as it turns out, in some of the subsequent reports that have been done on NOVA, about a third of the patients can stay at the dose in which they started, and two-thirds of the patients may require some reduction or interruption. There does need to be some attention paid to the tolerance of these drugs, as patients go forward. So, that trial was filed with the FDA. We were expecting an answer for their approval in June. And, by surprise, on March 27th, we received approval that the drug was approved, not only in the germline patient, but in all-comers. This removes the need to assess for HRD by one of the tests, as I mentioned. And, it then became available for use in patients in the switch maintenance setting.
While that was cooking, SOLO-2, just a couple weeks before, had reported at the SGO (Society of Gynecologic Oncology) meeting. SOLO-2 was a trial that was very similar to Study 19, but was done in the germline patient population. So, NOVA—all-comers; SOLO-2—germline, BRCA. That trial also enrolled patients after a platinum response, but their response was assessed by the investigators. Both NOVA and SOLO-2 looked for progression-free survival. And, as we saw and expected to see in SOLO-2, patients who were randomized to olaparib, the tablet formulation versus placebo, had a dramatic improvement in the delay for progression. This was similar to what we saw in Study 19: a very low hazard ratio. Again, this was quite reassuring in the spectrum of activity that we had never seen until they started to show off these PARP inhibitors. So, we were quite happy to see that. It was confirmatory.
When we looked at the adverse events in the SOLO-2 trial, we saw very similar side effects that we would have seen across all trials of PARP inhibitors. Gastrointestinal toxicity, taste changes, weight loss—these are very common side effects that we see with PARP inhibitors. What we didn’t expect to see, but what was noted, was that there were higher rates of anemia than we might have anticipated. But again, the high degree of anemia, the grade 3/4 rates, were actually relatively consistent across all of the compounds.