DR LOVE: I can’t talk to you and not ask you about the niraparib study. I mean, to my naïve eyes, this was world-shaking. And I don’t know — to me the data created a new world for ovarian cancer treatment, in my humble estimation. Can you just briefly talk about what they found? Of course, the incredible thing was the benefit that they saw with this PARP inhibitor not only in people who were BRCA germline-negative, but in people who were HRD-negative. So can you kind of summarize the paper?

DR BIRRER: Yes. It’s of course out now in The New England Journal of Medicine, which tells you how high quality it is.

But this study has been going on for a while. It’s taking patients who have essentially platinum-sensitive recurrence, and they get treated. The tumor shrinks, disappears, and then they go on a maintenance phase with this PARP inhibitor, which, in the Phase II trial and in the Phase I trial, there was all the evidence that we could tell that this was going to be a very active PARP inhibitor. So the biochemistry is very clear. Now I think we have clinical data to support that.

But what was astonishing about the data was just the impact. If you look at the prolongation of PFS — actually, before I say that let me just define it a little bit more. They did take a big effort to try to divide patients down into their genomic makeup. And, as you alluded to, there were patients who had germline mutation of BRCA1 and 2. There were patients who had wild-type germline BRCA1 and 2 but had the gene mutated in the tumor. We call that somatic.

And then they used this tool to identify a whole other group of patients who had neither of those. The BRCA1 and 2 were completely normal but, based on this tool, had what looked like HRD. So you can see there are buckets of patients. And when you looked at the impact of the PARP inhibitor in this trial compared to control, the hazard ratios — and this is a measure of how beneficial it is — were just very impressive. The hazard ratios were down into the 0.2, 0.25 range, basically pointing out that patients who got this drug had a huge benefit versus patients who didn’t.

Now the question is, who benefits the most? And I think this is the interesting issue, which is clearly, germline patients benefited. No doubt. We pretty much expected that. The somatic mutations behaved like the germline. This is now the second data that tells us this. The ARIEL studies using rucaparib had also revealed this. And this is very important because we now know, if a patient has a mutation in the tumor but not in their germline, they’re still going to benefit.

But the third part was really interesting, which is when they took this HRD group that was measured by their tool, they benefited less than the other groups, but they still benefited. But if you took out all of those patients and you were left with the final group, normal BRCA status, no HRD based on the tool, they even benefited. In fact, it was a hazard ratio of 0.58 with a p-value of 0.02. Now you could argue a little bit about the p-value, but they’re still benefiting.

So I think it’s a great study. It also tells me a second thing, which is although we think we’re really smart and we understand the science and the genomics, the truth is we don’t really quite get it yet. Their HRD tool probably is not as accurate as — well, we know it’s not as accurate as it should be. And that’s partly because we still don’t quite understand what causes homologous recombination defects in these tumors.

DR LOVE: Well, it’s funny because we were talking earlier about the fact that 25% of the patients in the survey never had BRCA germline testing. But of course there are family implications to that. But beyond that, I was wondering whether, if you’re going to use a PARP inhibitor anyhow, whether that’s going to change?

DR BIRRER: I’m sorry. Just repeat that.

DR LOVE: Whether or not, if it turns out as you are guessing or thinking might happen, that you’re going to use PARP inhibitors regardless of HRD or BRCA, yes, you might want to know BRCA for other reasons, but therapeutically maybe you won’t need to use it.

DR BIRRER: Yes. Well, I mean it’s possible. I’d be disappointed if that happened, but it may in certain circumstances. Because I think, as you point out, family issues are going to be incredibly important for that patient population. But the second issue is, even though we might give PARP to everybody we still know HRD status. Certainly BRCA1 and 2 mutation is a very strong prognostic factor. And that’s something I like to tell my patients, say, “Lookit. You’ve got this mutation. It’s not great news for your family, but we know how to handle this. But also recognize, your tumor is going to be exquisitely sensitive to chemotherapy.”

A lot of BRCA tumors — and the patients undergo 6, 7, 8, 9 cycles of different things with complete responses because these tumors are on the edge of dying, because of their DNA repair abnormality. You just need to push them over.

DR LOVE: Yes. And in a weird way, I was kind of thinking maybe it’s going to evolve the way PD-1 is and checkpoint inhibitors in lung cancer. And maybe we’ll end up using BRCA and HRD in ovarian cancer to figure out clinically when to use it. This was in platinum-sensitive recurrence?

DR BIRRER: Correct.

DR LOVE: So would you like to use it in platinum-resistant disease?

DR BIRRER: Yes.

DR LOVE: Would you like to use it up front?

DR BIRRER: Yes. That’s the million-dollar question.

And my guess is rucaparib will be approved, also. And then the question becomes which one do you use and when do you use it? I think moving these things all the way up front is a flawed approach. And the reason I say that — I’ve been a big critic of this — if you try to combine PARPs with chemo, it’s extremely difficult. They are very myelosuppressive. Not surprising. So when this has been tried, what happens is you have to dose reduce the platinums and everything else you’re giving to try to get the PARP inhibitor in. So that’s a problem.

In addition, if you’re treating at day zero and your patients are living, median, 5 years or 6 years, then I would ask, “Do we need to worry about MDS and leukemia if we have that kind of follow-up?” Because, remember, the trials to date have been later in the course. And so I’m not so sure we have the safety profile that far out. I think up front, I’m not a big fan, but it begs the question because you could bring them in as maintenance up front.

DR LOVE: That’s what you would think normally.

DR BIRRER: Yes.

DR LOVE: When I said “up front,” I meant as maintenance, really.

DR BIRRER: Yes, not in combination.

DR LOVE: Yes.

DR BIRRER: And I think what’s going to happen right now, if I see it, is that olaparib is at 3 lines or greater. Rucaparib, in front of the FDA, is going to be for somatic mutations, 2 lines or greater. So in fact, they’ve kind of got a little edge there. But NOVA1 is going to be, in the platinum-sensitive queue for maintenance, in front of all of them. So I think the natural reflex would be, if, let’s say, they all get approved, based upon the NOVA1 data, I think you’re going to see a lot of people using it in that context.

Now, do you use it on — perhaps you might segregate out the somatics and the germlines and pick them, because they get the best, the most benefit, and then bring it in later on, on the wild-types, later in the course of disease, because you could give it anytime. You might see some partitioning like that.

But the other question is do we know what causes resistance? Answer is pretty much no. We do know there are reversion mutations in BRCA1 and 2, but it begs the question that if you treat a patient early in her course, they get a response, they do well, but then the tumor comes back. Two years later, they’re still alive and they’re getting treated. Could you re-treat with PARP inhibitors? Don’t know the answer to that.

Metastatic wild-type nonsquamous non-small cell lung cancer (NSCLC)

Metastatic colorectal cancer (mCRC)

Metastatic epithelial ovarian cancer (OC)