Tom Herzog, in front line we did PRIMA (NCT02655016) and PAOLA (NCT02477644), which went beyond 

 and it was great. Now we’re also looking in patients beyond 

 in recurrent disease. We have the treatment indications: for rucaparib, it’s germline and somatic; for olaparib, two priors; three priors germline; and then there’s HRD, three priors for niraparib. There are opportunities for combinations in nonmolecularly signature groups. Tell us about those novel combinations, such as the combination of dostarlimab and PI3K [phosphoinositide 3-kinase], as well.

Niraparib plus dostarlimab, TSR-042 as I used to know it, is called MOONSTONE (NCT03955471), a phase 2 study. Drug combinations is really where we’re going, and you see that in front line, too. The idea of a checkpoint inhibitor with a PARP is that the PARP may increase tumor mutational burden and make a checkpoint inhibitor more effective, increasing tumor-infiltrating lymphocytes. You were talking about making a choice for platinum-sensitive disease, but maybe we don’t. If we look at AVANOVA (NCT02354131), which came out with niraparib plus bevacizumab, versus niraparib monotherapy, presented by Mansoor Raza Mirza, MD. The study showed the PFS [progression-free survival] was 11.9 versus 5.5 months, which is a significant improvement. There are a lot of things here in terms of combinations. Dr Konstantinopoulos presented, at ESMO [European Society for Medical Oncology] this year, olaparib plus alpelisib, which is a PI3 kinase AKT inhibitor, and it showed some really good data with relatively small numbers. We have combinations of VEGF with TKIs, so there are a number of studies going on.

In phase 1 and phase 2 of the TOPACIO (NCT02657889) trial, looking at niraparib plus pembrolizumab in platinum-resistant disease, the overall response of combined was 18%. These are platinum-resistant patients and the stable disease rate was very high at 33%.

I think there are a lot of combinations out there that are really appealing, and it points to the need to place these patients in clinical trials, because we have more combinations than we do patients. It’s really important that we put patients in clinical trials and figure out the very best combination for our patients.

 Thank you. We talked about the NCCN [National Comprehensive Cancer Network] Guidelines for front line, which preferred choice in recurrent disease is a clinical trial, to your point. Every time you enroll in a clinical trial, you are following the NCCN Guidelines at the highest level.

Videos

Novel Combination Strategies in Recurrent Ovarian Cancer

That’s a great summary. There are 6 indications of PARP inhibitors in recurrent disease: 3 for treatment and 3 for platinum-sensitive maintenance. There are 2 indications of bevacizumab: platinum-resistant and platinum-sensitive disease. And although there’s only 5 chemotherapy approved backbones, there’s really 6. Three are platinum-sensitive, carboplatin/gemcitabine, carboplatin/pegylated liposomal doxorubicin, carboplatin]paclitaxel. In resistant disease, liposomal doxorubicin, weekly paclitaxel, and topotecan. There’s a lot of flexibility, and I think the key message is that they do need these targeted therapies.

Let’s talk about what’s new; clearly these patients in recurrent disease are not being cured. We have some single-agent studies, some of which were presented here at the Society of Gynecologic Oncology [SGO annual meeting]. Katie Moore, tell us about some of these single-agent, interesting, novel therapies, some of which are antibody drug conjugates [ADCs].

 These are all relatively small studies but a glimpse of what’s to come in the next few years in terms of larger potentially randomized studies. Debra Richardson, MD, presented on a novel antibody drug conjugate, XMT-1536; it doesn’t really have a name yet. This is an antibody drug conjugate that targets NaPi2b, a sodium phosphate transport. This has been a target in some other ADCs that actually had a nice signal in the past.

This is an interesting target in that it’s a lineage marker for ovarian cells. Where you have NaPi2b expression is anything that’s originated from the ovary or fallopian tube, so it’s really overexpressed only on ovarian cancers and not on normal tissues. This is an ideal target for these antibody drug targets, which you can think of as Trojan horses, where they get to the gate by finding that target; they get let into the gate, and then they release this payload of very potent cytotoxins, totally dissimilar from any of the normal chemotherapy we would use.

In this case it’s an auristatin chemotherapy. It’s a pretty novel antibody drug conjugate and has a very high drug antibody ratio, so it really delivers a punch.

They reported on their phase 1 data with 59 patients; 37 of them were patients with epithelial ovarian cancer. They’re at a very high dose range with no DLTs [dose-limiting toxicities]. The partial response rate thus far among women who received a reasonable dose of the drug, excluding those at dose level 1, is 33%.

This is an interesting signal in very heavily pretreated patients who are in phase 1 centers, and it’s moving its way into larger multisite phase 2 studies as we speak, something that folks should be aware of coming down the path for them to have access to. We’ve definitely had it open here for a while and have seen responses, and those are encouraging. So that’s an antibody drug conjugate.

 I’m going to stop you right there because you, Tom Herzog, and I, we’re co-associate directors of the GOG [Gynecologic Oncology Group]. And these studies with that ADC targeting NaPi2b are open within the GOG. And Sharyn Lewin has a reputation of being at one of the earliest sites to open our studies, and certainly Tom Krivak is at a high enrolling center. So don’t forget these opportunities in that study and in the one you’re going to tell us about next, and to operationalize these enrollment situations through the GOG partners mechanism. Go ahead, tell us about the next one.

 Of the other things that were presented, probably the one that’s most likely to move forward falls into the realm of novel ways to target angiogenesis, and potentially come at those patients who have seen bevacizumab maybe several times and bevacizumab stopped working. Can you still target the angiogenesis pathway effectively? 

There’s a pathway called the NOTCH signaling pathway that has 5 ligands. The most critical, or the ones that we’re targeting right now, are DLL4, NOTCH1, and NOTCH3. There are 2 related abstracts that were presented at SGO, one run by Robert Coleman, MD, and presented by Katelyn Handley, [MD], looked at a pure Notch inhibitor called demcizumab, which is a DLL4 antibody, with weekly paclitaxel, in a phase 1b population of recurrent ovarian cancer. They reported a response rate that was respectable. It was about 22%. So it was there, but not overwhelming. And it had some of the expected toxicities, which are hypertension, and notably, pulmonary hypertension. When you only target NOTCH, you lead to this vessel sprouting in the pulmonary vasculature. That’s one of the tricky things about targeting NOTCH. If you add in VEGF inhibition, you can decrease that vessel sprouting and hopefully avoid that complication, and hopefully also increase your efficacy.

That’s the drug navicixizumab, presented by me. It was navicixizumab plus weekly paclitaxel in platinum-resistant ovarian cancer. It was another phase 1b study. About 70% of the patients who enrolled in the study—all of them had epithelial ovarian cancer—had seen prior bevacizumab, which is unique to this study. And we had an overall response rate of 43%. Again, they were very heavily pretreated. This agent is moving into a randomized phase 2 trial through the GOG partners mechanism, and we’re pretty excited to see it moving forward.

: Thank you. I was really looking forward to that, very concise.

Recurrent Ovarian Cancer: Novel Agents Under Investigation

Katie Moore, I saved bevacizumab for you because you did a lot of these PARP inhibitor studies. Is there still a role for bevacizumab in recurrent ovarian cancer? And if so, tell us about that opportunity.

 I absolutely think there’s still a role for bevacizumab in recurrent disease. We know, from randomized phase 3 data, that bevacizumab is still as effective even if you’ve had prior bevacizumab. Even for patients who recurred during prior bevacizumab, if they are put back on it with chemotherapy and bevacizumab to follow, the benefit remains. So bevacizumab after bevacizumab has been demonstrated. Now, I think the new indications will inform some of the opportunities for bevacizumab, but quite a few more women will be eligible, thankfully, for PARP inhibitor in the front line. You all just nicely reviewed SOLO-2, ARIEL3, and ENGOT-OV16/NOVA, which were all done in PARP-naive women. 

Now we are going to have a lot of these patients who are PARP pretreated and may have progressed on a PARP; we don’t have data yet on the efficacy of PARP after PARP. We do have that data for bevacizumab, and a lot of them will not have received bevacizumab in the front line.

The role of bevacizumab actually comes roaring back now in this line of therapy, and the nice thing is that it’s been so well studied. The original study GOG-0218 and OCEANS limited us to carboplatin/paclitaxel and carboplatin/gemcitabine, which are both nice backbones. Now we have data out of the AGO [German Gynecological Oncology Group] with carboplatin and pegylated liposomal doxorubicin, which is my favorite regimen, just because the schedule is so nice for patients, showing that bevacizumab added to that is not only safe but is at least equivalent to carboplatin/gemcitabine plus bevacizumab and maybe a smidgen better.

There are now 3 backbones in a platinum-sensitive population in which you can use bevacizumab. I think the role of bevacizumab in this line of therapy probably went down a little bit with the PARP approval. But now that PARPs have moved to the front line, I think bevacizumab is actually going to have a resurgence here.

My goal because, like you all, I like maintenance and I like targeted therapy, is to have every patient have an opportunity to get bevacizumab and PARP, and maybe even together. And how do I do that? I work under the umbrella of the labels because that informs reimbursement.

There are, Tom Herzog, some rare instances though where I’ll see a patient, PARP and bevacizumab naive, who has a platinum-sensitive relapse, even though bevacizumab works after bevacizumab. How do you make that decision in the platinum-sensitive relapse, in the targeted-therapy naive, bevacizumab versus PARP?

 I think that is difficult. This is a quickly changing landscape too because while we see 40%, 50% of the patients getting bevacizumab in the front line, there’s a decent chance they may not receive bevacizumab. There’s going to be less of a chance in the near future where they won’t have received a PARP. So that could reverse things a little bit, and I may be more inclined then to go with bevacizumab if they haven’t had bevacizumab. But also, I’ve always looked at whether they have pleural effusion, ascites, anything like that because bevacizumab does a terrific job on those patients. I’m always anxious to put those patients on.

You then turn the corners on that, if they’re naive. You also have that window of opportunity hypothesis, right? That’s the only time that you can use a PARP on these patients because if you go down the line of treatment, you won’t be able to do that. And so I look at that opportunity as well, and I look at what that might mean in terms of a lost opportunity. Because I can always bring bevacizumab in. We know from the AURELIA data how wonderful bevacizumab works in the platinum-resistant setting as well. You always have the opportunity for bevacizumab, you don’t always have the opportunity for PARP, especially if they don’t have a germline or somatic mutation.

Yes, Sharon and Tom Krivak, does that resonate with you that if you have a platinum-sensitive patient who’s responding, really that may be the only opportunity for PARP unless they have a molecular signature? So you should capitalize on it because they might be an exceptional responder.

:I love bevacizumab and platinum-sensitive recurrence, though sometimes I get off the highway and venture into the weeds. Like Katie, I like carboplatin/Doxil q4 every 4 weeks and bevacizumab q2 every 2 weeks. You increase the response rate and try to get a complete response and then switch them over to a PARP at that point. I then try to see how long they’ll last on the PARP again. I try to do HRD [homologous recombination deficiency] testing at times to let me see how long that that PARP duration may be. I’ve done that quite commonly. I love carboplatin/Doxil q4 every  weeks and bevacizumab q2 every 2 weeks, try to get a complete response or get the best response after 6 to 8 courses, and then switch them over to PARP.

Thank you for that. And it’s not just the schedule, it’s the adverse effect profile because you have less alopecia and less neuropathy.

Bevacizumab’s Role in Recurrent Ovarian Cancer

I’ve been easy on you, Sharyn, but I’m not going to be easy on you now. There are these three indications for maintenance treatment after responding to second-line or beyond platinum. I’m not going to go through the SOLO2, Study 19, but they are not too dissimilar to NOVA and ARIEL3. The first PARP inhibitor that became approved in treatment was fourth line, and it was only germline. As Tom said, we have a regimen line of therapy earlier in rucaparib, two priors not three. And as Tom Krivak said, we now have beyond 

, we have an HRD [homologous recombination deficiency] molecular signature.

We have three treatment indications—I get it the labels are different a little—and we have three identical labels in platinum-sensitive maintenance treatment. So the tough question for you is, in the setting of overlapping labels, which PARP?

: That’s a great question. If you look at the hazard ratios on these trials, these PARPs are pretty similar when it comes to efficacy. It really boils down to safety profiles and tolerability. From personal experience, which PARP I’d use the most on trial or off trial, I do have my favorites, but the safety profile is what it boils down to.

Tom Krivak, how about familiarity? Does familiarity play into that, or is it more safety?

 You get familiar and comfortable with the PARP you feel is the safest. We’ve never had three drugs very similar come on in competition, so it’s been hard for us to figure out how to use it. I think it’s nice having choices. I agree completely with Sharyn in that it’s adverse effects. If somebody has really bad nausea/vomiting on olaparib, you can consider changing over to rucaparib or niraparib. If some patients are not as driven and may not want to take their medications as often, you have once-a-day dosing with niraparib.

I work in a large health system, and I think at some point it’s going to be cost-driven. We’re going to get two PARPs, and we have to decide which ones because we’re going to go back. I hope it never comes to one option because I completely agree with Sharyn in that they’re all not equal. The science of medicine is understanding these studies, and the art of medicine is taking each of these studies and applying them to the right patient so we get the patient to be on the drug for the longest period. That is the key.

Recurrent Ovarian Cancer: Choosing a PARP Inhibitor

I’m going to transition now to recurrent disease. As Dr Moore said, these patients do develop recurrence, and there are a number of opportunities. We have six PARP inhibitor indications in recurrence. I’m going to ask our panelists to go through each of the PARP inhibitors in no particular order, and to talk about the maintenance opportunity after platinum-sensitive response, as well as treatment. I’m going to start with Tom Krivak, and maybe you could talk about niraparib and what the indications are for niraparib in recurrent ovarian cancer.

, there was a study of niraparib with two cohorts, a germline cohort and a nongermline cohort. This study was looking at platinum-sensitive recurrent ovarian cancer with platinum-based chemotherapy and then PARP maintenance. We saw a significant benefit with a hazard ratio of 0.27 in the

For the subgroup analysis going into the nongermline-mutated patients, they broke it down and looked at 

 somatic mutations, HRD [homologous recombination deficiency], and HR [homologous recombination] proficiency. When looking at the 

 somatic mutation patients, there are very similar outcomes with the platinum-based chemotherapy followed by PARP maintenance with niraparib.

In breaking down the patients who had HRD, they looked at the somatic patients with HRD, and then broke out the HRD, excluding the somatic patients. There was an improved PFS [progression-free survival] and a hazard ratio of 0.45. Going to the HR-proficient patients, there was an improvement of about three months in that group.

It was a nice trial in that we saw a really nice prolonged response in our 

-mutated patients, responding to platinum-sensitive recurrence. As we’re trying to build on our biomarkers, this was a nice trial because of the exploratory analysis of HRD, the Myriad Genetics, Inc test. I think we could debate the absolute benefit of three months in the HR-proficient group—what I used to call the triple-negative group, negative

 germline, and HRD-negative—which is one of the most complex patients to treat due to quick recurrence.

I think that built on the OCEANS trial, which looked at bevacizumab. Now we’re looking at this using ENGOT-OV16/NOVA and niraparib for platinum-sensitive recurrent ovarian cancer. 

The beauty of that, as Katie Moore said, is that we had the integrated biomarker with platinum sensitivity. We underemphasized the HRD testing, but HRD testing is much more valuable in frontline if you decide to do it. But NOVA was the first platinum-sensitive maintenance opportunity, and then tell us about QUADRA.

 NOVA expanded the indication. It went from 

 mutations to all-comers. NOVA was well designed because of safety nets to look at certain baskets of patients. QUADRA, this is Katie’s trial, and it’s odd for me to be talking about Katie’s trial because she’s done so many good trials, had heavily pretreated patients, both HRD and 

 patients. It screened about 700 patients and enrolled about 463. Patients who had a 

 mutation, regardless of platinum-sensitivity, in the fourth-line could receive niraparib as treatment, so going from maintenance to treatment. If a patient was HRD-positive, that patient had to have platinum sensitivity. The study had some very nice numbers with respect to who responded and who didn’t respond. The platinum-sensitive 

 patients were probably around 40%, and platinum-resistant patients were right around 25% to 30%. When looking at the HRD folks, it was around 25% to 30%. To me, the study extended out how we can use niraparib and changed the indication going from a maintenance drug, even though I consider maintenance to be treatment maintenance in most of these cases, to actual treatment.

Katie’s done a ton of good trials, and this was another great trial to add to the options for our patients with recurrent ovarian cancer.

The beauty of QUADRA, and thank you Katie Moore for doing that, is that it showed us the 3 predictors of PARP activity: platinum sensitivity, molecular signature, and number of lines of therapy. We’ve leveraged those in PRIMA and PAOLA, so that was great. We also have rucaparib, Tom Herzog, so tell us about rucaparib and what that offers to our patients.

We have two indications for rucaparib currently. Unlike the other PARPs, it’s approved for recurrent disease for those patients who have had three- lines of prior therapy. It’s one line earlier than four lines and beyond, it’s three lines and beyond. Those, of course, are patients who have mutations.

And then it has an indication based on the phase 3 ARIEL3 data, where Robert Coleman, MD, presented this at ESMO [the European Society for Medical Oncology annual meeting] in 2017, and then it became approved in 2018. Jonathan Ledermann, MD, and Robert Coleman were co-PIs [principal investigators] on this particular trial. The randomized phase 3 trial was similar to NOVA, but there were some differences. But what was remarkable to me was how close the control groups were in these phase 3 trials. These were high-grade serous or endometrioid, greater than or equal to two prior lines of platinum, and had to have a CR [complete response] or a PR [partial response] to the penultimate platinum that they had right there.

They also had to normalize their CA-125 [cancer antigen 125], which is interesting. They didn’t have a size limit on their residual disease, which was a little bit different. They had 564 patients randomized 2:1 either to the rucaparib or to the placebo. They also had the step-down statistical analysis, first looking at their 

 mutation group, with a hazard ratio of 0.23. Then they went to the HRD group, with a hazard ratio of 0.32. Their HRD was measured by Foundations Medicine, Inc, LOH [loss of heterozygosity] greater than 16%. They then went to the intent-to-treat population, with a hazard ratio of 0.36 as you get out there.

The medians change, and you can see the same effect where if we look at the LOH low population, it was still positive, but the difference in the medians was 6.7 versus 5.4 months. So very much what we’ve seen in these first-line trials, with a different PARP, but very interesting safety profile, pretty much in line with what we’ve seen with rucaparib from the Study 10 and ARIEL2, which formed their treatment label.

PARP Inhibition’s Role in Recurrent Ovarian Cancer

 The next study that we’re going to hear about is the second strategy, adding an anti-VEGF to a checkpoint inhibitor. Tell us about your study, Katie Moore: IMagyn050.

IMagyn050 is a randomized phase 3 study, 2 arms. It enrolled all histologies except for mucinous, so very different from what we’ve just discussed. All of those studies—PRIMA, PAOLA-1, VELIA—are limited to high-grade serous and some high-grade endometrioid. But this allowed clear cells as well. Biomarker stratification was PD-L1, and that was an important difference between our study and JAVELIN OVARIAN 100. We are stratifying there. Then we randomized to paclitaxel, carboplatin, placebo, and placebo maintenance with bevacizumab. So everyone gets bevacizumab throughout, and then placebo, placebo maintenance, or paclitaxel, carboplatin, bevacizumab, atezolizumab; and then atezolizumab-bevacizumab maintenance. It’s a dual primary end point of progression-free and overall survival. It finished accrual in April 2019, and we are in final stages of data cleaning.

 We’re excited that it should be the next large study. Hopefully it’ll be a breakthrough.

 Sharyn Lewin, you’re an advocate of this trial, and you enrolled in your practice called FIRST. Tell us what FIRST is.

FIRST is presently open and enrolling patients. It’s a very interesting study for women with stage III and IV epithelial ovarian cancer. It’s looking at standard of care, which is paclitaxel and carboplatin IV [intravenous], interestingly with or without the bevacizumab and looking at the role of VEGF. Patients are randomized to either the dostarlimab or placebo in cycles 2 through 6 and with or without Avastin, based on your prior decision. Then there is a maintenance phase looking at either niraparib, dostarlimab, or placebo. It’s a complicated schema looking at the different formations of that. It’s just looking at the combination of dostarlimab, bevacizumab, and niraparib versus placebo.

 We’re excited, I enrolled in that too, and we’ll have to see what happens. Then I have a study called ATHENA. ATHENA is a pure maintenance trial that distinguishes itself from these other trials in the PRIMA theme where we focused on the maintenance opportunity. It is just a simple question: Should we add a checkpoint inhibitor to a PARP inhibitor? 400 patients get maintenance PARP, à la PRIMA, and that PARP inhibitor is rucaparib. Then 400 patients get nivolumab plus rucaparib, and integrated into that is a placebo. The first end point to report will be the PRIMA opportunity, just PARP versus placebo. Over time we’ll do the primary end point, which is rucaparib versus rucaparib-NIVO [nivolumab]. There is a small nivolumab arm of 100 patients. It’s 400 rucaparib, 400 rucaparib-nivolumab, 10% or 100 nivolumab, 10% placebo. This trial has finished enrollment within the United States; it’s winding down. This will probably be, at least in the rucaparib-versus-placebo arm, the next study to report after IMagyn050 in the frontline setting.

Let’s go ahead and end there. That was a comprehensive conversation about frontline treatment, really leveraging the opportunity of GOG-0218 and bevacizumab approved in June 2018, and then ultimately the opportunity for SOLO-1, PARP inhibitor and maintenance therapy, PRIMA, PARP inhibitor beyond 

, and then ultimately the combination PAOLA-1.

I want to quote my good friend Tom Krivak, who says, “It’s a maintenance world”—the most important thing is that maintenance is used. I’m going to quote my good friend Katie Moore, who says, “I just don’t understand why people would think that these patients with advanced ovarian cancer are cured.” There’s a synergy there that if they’re not cured, you have the opportunity to keep them in response. All these studies did not show a decrement in quality of life.

Frontline Ovarian Cancer Management: Ongoing Clinical Trials

 I want to finish our conversation by pivoting away from the NCCN [National Comprehensive Cancer Network] Guidelines and talking about what’s new and evolving. A paper that Jonathan Ledermann and I presented at SGO [Society of Gynecologic Oncology Annual Meeting on Women’s Cancer] is a paper called JAVELIN OVARIAN 100. I think all of you are aware that there are 3 established ways to make checkpoint inhibitors better. One is added to chemotherapy. And JAVELIN OVARIAN 100 addresses that. I’m going to go next to you, Katie Moore. The second way is to add an anti-VEGF to a checkpoint inhibitor. The third way is to add a CTLA4, another checkpoint inhibitor. We don’t have a lot of data on that, although NRG Oncology’s NIVO [nivolumab] vs NIVO-IPI [nivolumab-ipilimumab] study was just published.

JAVELIN OVARIAN 100 was presented at SGO. Quite frankly, it was a study that to me was very disappointing because, as you know, we can add chemotherapy to checkpoint inhibitor in triple-negative breast cancer, in lung cancer, in lots of opportunities. We just decided to do it in ovarian cancer. I get it; It was a ready, shoot, aim approach, rather than aiming before we took on this massive endeavor. But there were 998 patients on carboplatin-paclitaxel, and then we added avelumab to it, either avelumab in the maintenance phase—because that’s really what we believe in, maintenance—or with chemotherapy and in maintenance.

The study was stopped by the Data Safety Monitoring Board. In the maintenance arm, the hazard ratio, when you add checkpoint inhibitor and maintenance, avelumab did not overlap 1, but it was on the wrong side. It was a hazard ratio of 1.43, with a 95% confidence interval of 1.05 and 1.95. So it suggested—didn’t test the hypothesis but suggested—that maintenance checkpoint inhibitor was harmful after response to chemotherapy. That was surprising. When it was added to chemotherapy and maintenance, there was a hazard ratio of 1.14. We took a deep dive into the biomarkers, and some of the biomarkers were presented here, PD-L1 expression specifically, and there was no real signal. And PD-L1 expression measurements are complicated. You can either measure the immune cells or the tumor cells. About a quarter of the immune cells were positive, about a quarter of the tumor cells were positive, and there was a lot of overlap in the immune cells and the tumor cells. And 9% were negative. So it wasn’t helpful. We’re doing RNA-seq [RNA-sequencing] and other biomarkers. We’ll probably have those for you at the IGCS [International Gynecologic Cancer Society], Tom Krivak, where you’re the program chair. So that was disappointing.

Considerations for the JAVELIN OVARIAN 100 Study

Sharyn, let’s assume you use bevacizumab [Avastin] as the label says. If you have a wild type patient and they respond, then you give the PAOLA [PAOLA-1 (NCT02477644)] regimen or bevacizumab alone. Then it goes back to the same F subnote that says HRD [homologous recombination deficiency] might be helpful. I think you like that. Do you agree with that? Do you like that?

 Tom summed it up very nicely. There are definitely different decision points in ovarian cancer. One decision does dictate the other. Whether to use Avastin is irrespective of biomarkers in some people’s mind. But I do believe in the biomarker. It’s not a perfect test. It would be nice to actually see the score if you are using myChoice HRD. But I do think those data influence my personal recommendations for maintenance, once I have that information.

Thank you for that. If you have bevacizumab started and you have a 

 mutation, then interestingly, Katie Moore, the category 1 recommendation is not to stop the bevacizumab. Category 1 is bevacizumab plus olaparib. What do you think? If you have bevacizumab going and you have a PR [partial response] or CR [complete response], should stopping the bevacizumab be considered, or should you just add olaparib onto, it which is what the NCCN [National Comprehensive Cancer Network] recommends?

 That’s another fantastic question we don’t have clinical trial data for. But that was one of the big problems with the olaparib indication in the first place, because SOLO-1 (NCT01844986) did not allow bevacizumab. This was a big problem in many parts of Europe when we ran the study, where it is the standard of care in frontline. People would finish their six cycles with bevacizumab and then stop the bevacizumab so they could go on to olaparib. Does that really make sense? This allows you to use both. Quite frankly, I like it. At the SGO [Society of Gynecologic Oncology] meeting, as you know, Dr Ignace Vergote presented a purely hypothesis-generating—so I’m not at all falling all the sword that this is the thing—modeled comparison of PAOLA-1 and SOLO-1.

The propensity-weighted analysis pulled the patients with 

-associated cancers out of PAOLA-1, and then matched them on key prognostic endpoints. They did a weighting so that the patients looked similar in two groups. Then they did this series of comparisons. They did a comparison of olaparib versus olaparib plus bevacizumab. What’s the benefit of the bevacizumab? The hazard ratio was 0.71. There did, by this mathematical experiment, appear to be at least an additive—and probably only an additive—benefit to both. But that’s OK. It was a little bit better. It would be nice to prove this in a real prospective clinical trial, but I doubt we ever do that. The hazard ratio is 0.71.

To some earlier points that people brought up, only in patients with 

 mutations, they compared olaparib to bevacizumab. The hazard ratio in favor of olaparib was 0.48. There’s still a group of people who think bevacizumab and olaparib are similar, but they’re not in a 

-associated cancer. Then they compared bevacizumab versus placebo in these two populations because there’s a group of people who believe that these patients are cured and don’t need any maintenance therapy. The hazard ratio was 0.65 in favor of the bevacizumab. They did a whole bunch of these analyses, but the most interesting of them is this olaparib/bevacizumab versus olaparib that would suggest an additive benefit. You’re certainly not harming people, I don’t believe. You definitely have to watch the safety profile because you’re adding another active agent with toxicities. But if you’re going to start bevacizumab, honestly, I would keep it going and add the olaparib in that population.

Thank you for that. Tom Herzog, interestingly, in the NCCN guidelines, if you start bevacizumab and a patient has a 

 mutation, they give you the opportunity to add niraparib to bevacizumab. Obviously, that would not be prospectively tested in an informative way. But there was a study that looked at that here at the SGO. I think that it’s called OVARIO (NCT03326193). Is that right Tom?

 Yes, the OVARIO trial. Let me present the OVARIO trial, and then I wanted to ask Katie a question. Let me ask you now, since we’re talking. What are you going to do for your next patient, germline or somatic-mutated, whom you would normally follow SOLO-1 for? Would you add bevacizumab?

 To our earlier point, the decision for bevacizumab is made much earlier, often before I know the germline status of the patient. I’ve made the decision to choose bevacizumab because they have a lot of ascites or bulky disease. That would be my practice, and I’m trying to get a faster, deeper response because I’m trying to get them out of trouble. I get the germline information and in the past, I would have to stop the bevacizumab at the end of chemotherapy, which we all know from a PFS [progression-free survival] standpoint is not how the drug is supposed to benefit. Certainly, I got the response benefit, so now I would leave them on bevacizumab.

 The answer is yes. If you had previously put them on bevacizumab, you would leave them on.

 Got it. Great. Let me just cover the OVARIO trial, because this is really interesting in terms of the NCCN basically saying that you can substitute these. We do have some data on niraparib with bevacizumab in advanced ovarian cancer looking at Stage III/IV. Remember, there are data that were presented by Dr Melissa Hardesty. It’s a straight phase 2 trial with 105 patients. The primary endpoint is a landmark analysis of how many are progression-free at 18 months. We have the preliminary data from six months, which was really impressive: about 90% of patients were progression-free. The 12-month data came out and the actual presentation is 75% progression-free. A little bit higher in those who had HRD, 88%, versus those who were HR proficient, where it goes down to 66%. The discontinuation rate was 25%, which is not a lot different than what we saw in PAOLA-1, which was about 20%.

Ovarian Cancer: Maintenance Following Frontline Bevacizumab

Let’s pivot to the NCCN [National Comprehensive Cancer Network] guidelines. NCCN guidelines inform reimbursement and are certainly educational on how we treat our patients. As you know, the NCCN guidelines were updated on March 11, 2020. If you think we can’t build consensus, although I think we can, it takes the NCCN 127 pages to tell us how to treat epithelial ovarian cancer. That’s impressive. Page 11 is the most informative to our conversation here about frontline therapy. It’s entitled, “Epithelial Ovarian Cancer, Fallopian Tube Cancer, Primary Peritoneal Cancer, and What to Do Frontline.” They start with this bifurcation of no bevacizumab use during primary therapy or bevacizumab use during primary therapy. They don’t tell us when to do it, there’s no preference. They don’t say, “This is when you should use bevacizumab, this is when you shouldn’t use bevacizumab.” I thought that’s what the NCCN was supposed to do, tell us how to treat our patients. It doesn’t do that. That’s why I have Tom Krivak because I can ask Tom Krivak, since he presented GOG-0218 [Gynecologic Oncology Group 0218 study], when do we use bevacizumab? When should we or when shouldn’t we?

 You’re absolutely correct. When you look at the NCCN guidelines, you’d like to see who gets bevacizumab and who doesn’t. It’s nebulous, so everybody can say, “I’m right and you’re wrong, look at the NCCN.” How do I use bevacizumab? I use it by the label. Surgical debulking is important, but if you have stage III/IV ovarian cancer, you’re going to get bevacizumab up front and then tailor with maintenance therapy.

Obviously, that’s not what the NCCN says. They have a very nice algorithm where they look at upfront use versus use in the maintenance phase; look at 

 wild type or not, and germline status; and then go through complete response versus partial response and how you could potentially use bevacizumab—as you said, Brad—as the building block to add a PARP inhibitor to. Or do you use PARP alone? To me, what I like about the NCCN guidelines is that they say you can use bevacizumab however you want to, basically, in upfront ovarian cancer. We really need to be testing these folks, and we should stratify based off of testing. It’s taking all the data that we’ve discussed and allowing us to say that for certain tumors we’re going to use bevacizumab alone for that patient. For other patients, they’re going to get a combination therapy. For other patients, they’re going to get single-agent PARP inhibition. They’ve done a nice job of covering all their bases for how we have our biases with therapy.

It creates four cells. You have the bifurcation of bevacizumab—yes, no—and then each one of those cells, it has germline or somatic 

Unknown is in the no-mutation subgroup. All of a sudden you have four cells, and we can work through them. If you don’t have bevacizumab, which is at least 50% or about 50% of the patients, and you have a 

 mutation, you need the SOLO-1 (NCT01844986) or PRIMA (NCT02655016) regimen. NCCN guidelines give you both of those options. If you get no bevacizumab and have a 

 mutation, SOLO-1 or PRIMA are category 1. Pretty easy. If you do not receive bevacizumab and you are wild type or unknown, it gives you the option to observe or use PRIMA, which is niraparib. Then there’s a footnote, Footnote F, and Footnote F says that in the absence of a 

 mutation, HRD [homologous recombination deficiency] status may provide information on the magnitude of the benefit of the PARP inhibitor. But it’s a category 2B recommendation. We’ll see more as the labels become clear and as the reimbursement becomes clear. But they also don’t tell us if we should use the HRD test, yes or no. They don’t tell us when to use it. They just say that if you do use it, it’s a 2B recommendation and it might be helpful. Do you have any comments on that, without bevacizumab? Do you agree with that? We’ve gone through the assay, but do you like this, Tom Krivak? Tom Herzog? How do you work through that? Is this informative or not?

 It’s been a month or so since I’ve read them, but I thought you summarized them well and I know that it was category 2B in terms of the HRD. I think one of the other things—if I didn’t misread them—was that it also included stage II? Quite interesting, right?

 Because these studies are all stage III/IV. But I think, Brad, you’ve taught us that one decision affects the other decision, right? If you make that first decision, your real first decision is, are you going to operate right away or are you going to give chemotherapy and then operate? If you’re going to operate right away, that takes you to your next decision, if you’re going to use neoadjuvant therapy. Your next decision is bevacizumab or no bevacizumab. Now, with all these new data from PRIMA and PAOLA-1 (NCT02477644), how are we going to incorporate that? Now you’ve really got to make some decisions. Are you going to add to the bevacizumab? Are some people going to, when they get their HRD information back, stop bevacizumab even if the patient is responding well, and then start a PARP inhibitor as maintenance?

It’s going to be very interesting to see what people do with the information because it’s also going to be temporal in terms of when you’re getting this information and when you test. Obviously, the sooner, the better, so that you can make the best decision for the patient. I found the NCCN guidelines to be interesting, and nebulous was a good way of wording it.

Ovarian Cancer: Updates to NCCN Guidelines

Tom Krivak, I’m going to call you out on a contradiction. You said, “I’m a biomarker guy. But if the biomarker is negative, I’m going to use niraparib anyway.” I don’t get it.

 It’s predictive and prognostic. It’s not perfect. 

 is a great biomarker for risk-reducing surgery, but do all 

 patients respond to platinum-based chemotherapy? Is that 100%? I don’t think so.

 It’s a biomarker, there’s no doubt about it. I did read the abstract by Dr Elizabeth Swisher, and here is how I looked at it: you have veliparib, which is not an approved PARP for ovarian cancer, and you have 2 groups that crossed unity. You’re comparing some things. I’d love to see the slides of the presentation. I’m biased. I believe in biomarkers to help me stratify and develop my plan for sequencing, because I don’t think we know how to sequence. That’s what we’re talking about. How do we sequence chemotherapy? Doing this for 25 years as a fellow, it seems like the best way to sequence chemotherapy is when you get in a large group and they tell you that you have to do this. You go through your algorithms. But to me, I’m looking at HRD [homologous recombination deficiency] as a biomarker to help stratify, to help me develop a sequence.

 You’re seeing the difficulty that the experts have right now, because we don’t have labels yet. We have a lot of new data. We have very different interpretations of these subgroups. As I sit in these different panels, I see that. I believe any information is going to be helpful. I saw the subgroup analysis where there’s benefit; therefore, I do want to use that. For me, the more information, the better, and I think these tests will evolve. They’re not perfect, but they’ll continue to get better. There are a number of tests in the works, and we’ll see if we can do better in terms of being not only prognostic but also predictive as we move forward. I like these subgroup analyses. It’s information we’re otherwise not going to get. On the other hand, I do believe we have to be careful in terms of regulatory status, because if the intent-to-treat population and the primary population are positive, it gets a little dangerous to deny subgroups benefit and label. I have concerns about how these subgroups are being used, not only clinically but also from a regulatory standpoint. The quick answer from me, Brad, is yes. I’m going to continue to use HRD. I’m hoping for a better test. It’s great to have options, but this is an evolving dynamic.

What’s most provocative is in PAOLA, if you don’t have an HRD signature, the hazard ratio is 1.00. Those 2 therapies together have toxicity, financial and otherwise. I think you’re on thin ice—I would never criticize you, but you’re on thin ice—if you say, “I’ve got a biomarker that says the hazard ratio is 1.00, and I’m going to do both drugs together anyway.” If the label prevents you from using both drugs because of that reason, then it’s going to be simple. I like flexibility too, but I’m really not sure that’s the thing to do.

Sharyn, you and I talk all the time. You like the assay, and so do I. What’s the best treatment for a patient who doesn’t have HRD and doesn’t have somatic 

? What’s the best maintenance treatment for that patient? Tom Krivak says I’ve got to use maintenance no matter what.

 That’s a great question. When you really look at these papers and think about it, if a patient has HRP [homologous recombination proficiency], that will be someone who I observe or who I consider giving bevacizumab to. But for me right now, with the data we have, I would really not push a PARP in that particular patient.

 That’s completely reasonable. That’s how I look at it because I do agree, with HRP I give 2 drugs together versus sequencing, and you may get more long-term longevity with having sequencing. I would like to comment on the development of the HRD test because I did read the abstract. How they developed it at Myriad Genetics, Inc was to capture your 

 patients. In the abstract at SGO [Society of Gynecologic Oncology], if you use 33 as a cutoff, you captured 96% of the 

 patients. When you use 42 as a cutoff, you captured 92%. The data I looked at when they developed the 42 cutoff were capturing about 95% of 

 patients and wanted to make sure that HRD was capturing patients who look like that. When you look at the Foundation Medicine, Inc test, the test was looked at and cutoffs were made based on response. You had the Myriad test that was developed looking at 

 and then plugged into studies looking at responses, whereas Foundation was developed and a cutoff was based on responses, not necessarily looking at the percentage of patients they missed who had a 

 mutation. The tests were developed a little differently. That’s my understanding of how they did it and how Myriad has made some of its cutoffs to try to capture those 

Treating Ovarian Cancer Based on HRD Status 

In PRIMA the primary end point included HRD [homologous recombination deficiency]. In PAOLA it was not even an end point, but because it helps inform decisions, it was reported. Tell us, Katie Moore, about this HRD assay that they used and what it means. We’ll get into whether we should use it. If we’re using it in all comers, if we believe that PRIMA is an all-comer opportunity, then maybe we don’t need to test it. But if we believe that in PAOLA it didn’t really work in patients who didn’t have the molecular signature, then it’s mandatory. Which is it?

 Let’s just start without the assay. What’s HRD? Homologous recombination is the high-fidelity way that all our cells fix their double-strand DNA breaks. That function, in a lot of tumors—epithelial ovarian cancer is 1 of them—has some Achilles’ heels. You can lose this homologous recombination function through point mutations, most commonly in 

, through epigenetic mechanisms, or through loss of other genes. There are increasingly identified ways that a tumor can lose its capability to complete homologous recombination and become homologous recombination deficient. The clinical biomarker for homologous recombination deficiency is platinum sensitivity. That gives you your answer of who is homologous recombination deficient, but you’d like to know that up front and use it as a predictive biomarker for selection of DNA-damaging assets, such as PARP inhibitors.

There are a number of companies that have tried to commercialize this. Two are on the market. One looks at loss of heterozygosity [LOH] alone as a percentage. It’s basically the percentage of the chromosome that has lost heterozygosity between the 2 chromosomes, and if it’s 16% or higher, it’s considered LOH positive. Then the assay that was used in these 2 studies, in PRIMA and PAOLA, uses 3 components—LOH not measured in the same way, telomeric allelic instability, and large state transitions—and combines them. They have a cut point of 42 in these 2 studies. That’s based on some science that is hand waving. You dichotomized a function of DNA by a point score to say these women qualify and these women don’t. That’s what was used in these 2 studies and is the test that’s been most widely studied among the most patients with ovarian cancer.

Whether you send it to Foundation Medicine, Inc and get LOH or send it to Myriad Genetics, Inc and get the 3 scores, it also gives you the somatic 

This is a tumor test, and there is useful information: HRD and somatic 

. Do we need this for everybody, or is it not a very good test and you’re going to use PARPs anyway?

 The truth is, we don’t have a good answer. It is a test. I think it’s a reasonable test. I have a lot of problems with it, though, and I’ll just go on record as saying that. When you talk about PRIMA, which was what Tom Herzog discussed, the analytical end points were HRD and ITT [intent to treat], and they met both end points. It was designed as a registration-enabling study, so they had their premeeting with the FDA. It should get labels in both settings, in my opinion. PAOLA was not designed as a registration study. It was an investigator-initiated trial that was very well done, and I’m really proud of Isabelle Ray-Coquard for doing it. I’m jealous that they can do these things in Europe. They had a great response and have filed it without the usual premeetings and things that go into designing a registration trial. Is that how you would have designed the trial if you were going through registration approval? I don’t know; probably not. Now we’ve stuck this assay in there as a secondary analysis, and it’s not hypothesis tested. It wasn’t 1 of the prespecified analytical end points. To make a registration decision based on that makes no sense to me. But the concession is that they hadn’t talked to the FDA before, so that just may be what they did. I think it should be an ITT indication. I don’t know if I’m going to get my way.

But are you going to do the test? These are great points, but are you going to do the test? I don’t mean to make it complicated.

 Why would I? This is what I believe. I’m going to reference 1 of the studies that came out of SGO [Society for Gynecologic Oncology Annual Meeting] that I know we’re not going to talk about. VELIA was another big phase 3 PARP study. They have an analysis that Liz Swisher presented, where they looked at HRD as well as a score of 33 to capture more people, but then they also looked at it across the continuum. They looked at the predictive capability of the actual score—1, 2, 3, 100—and they found that it’s not predictive of a response to veliparib. It’s prognostic. It’s prognostic to identify a new group of patients who are probably in trouble and for whom we need to develop better trials, but it wasn’t predictive. The magnitude of benefit of veliparib was the same above or below 33, above or below 42. These cut points are arbitrary. I don’t like to accuse it of not being a biomarker, I just want a good biomarker. If you make me pick 42, well, someone who’s score is 40 doesn’t get a PARP, and I think that’s dangerous. You’re leaving people behind.

Ovarian Cancer: What is the Role of HRD Testing? 

This really is the issue, Tom Krivak. You said you believe in maintenance therapy, and so do I. I think all of us do. I think that’s the evolution of our field. But now the question is, “What maintenance therapy?” I understand, cross-trial comparisons are inappropriate, but clinical decisions are appropriate and made every day. As we look at this, PAOLA-1 is provocative because of the active control. I always say if you can’t beat something, compete against nothing. Many of these PARP inhibitor trials competed against nothing. Well, this year, there was a competition against a very active—in fact, now an FDA-approved—control.

As we look at this, this is a key question. I don’t know what you’re going to say, and I understand that cross-trial comparisons are not appropriate. But when you make clinical decisions, do you look at the hazard ratios of PAOLA-1 and PRIMA—PARP alone versus PARP–bevacizumab [Avastin]—or do you look at the medians? For example, Sharyn said in the HRD [homologous recombination deficiency] non-

 group, which is the example patient we just talked about, the median is 28 months. Or do you say that you can’t really use that? Certainly, the elephant on the table is this sequential therapy. Maybe we need to look at bevacizumab and then PARP, or PARP and then bevacizumab, or the combination. We’ll never have that information. How do you incorporate this information that Sharyn so articulately reviewed with us?

 That was great, the way Sharyn and Tom and Katie and everybody reviewed the data. The way I look at it is that we have options. I like testing for HRD. I know we don’t have any FDA labels, but given the fact that we have the HR [homologous recombination]–proficient group showing no benefit with dual therapy, when I see the HR-proficient group, I want to use single-agent niraparib or single-agent Avastin. In the HRD-positive group, the patients who have homologous recombination deficiency, that’s a tough question. Should you go with both? Could you look at the absolute benefit versus hazard ratios? To me, they both work. Which is going to work better? I don’t think we’ll ever get that trial. If I have a patient who had really bad thrombocytopenia and issues with her platelets, I may be a little less hesitant to put her on niraparib and favor Avastin with olaparib. I’ve heard some people say, “Just give olaparib alone. That’s where you’re going to get your maximum benefit instead of olaparib-bevacizumab.” But the trial was olaparib-bevacizumab. I would use that. If somebody had a massive PE [pulmonary embolism] and doesn’t want to be on bevacizumab, then I would use single-agent niraparib.

I struggle regarding the hazard ratio/absolute benefit. Katie had a very elegant study with SOLO-1, and it was olaparib versus nothing. When you look at the hazard ratios, I always wonder what the hazard ratio would be in PAOLA-1 if there was no active treatment. Should the 

-mutated patients receive olaparib with bevacizumab? That’s the struggle that I have. But across the board, for all 3 or 4 subsets—what I just talked about—I’m going to give them something based on adverse effects and based on their molecular marker. Or at least offer them something and they can make that decision. But that’s how I look at how these 3 studies give us options that are all complementary. Really, when you look at all 4 studies, they solidify that we need to be using maintenance therapy for these patients.

That’s the key point. Remember, Tom Herzog, you talked about PRIMA very nicely, but it was a very high-risk population. It was the population that we feel the most comfortable using bevacizumab in. There was a paper at the SGO [Society of Gynecologic Oncology] meeting that went through some of the subpopulations of PAOLA. Do you have that on your mind?

 It didn’t really teach us a whole lot in terms of what we already knew, but it certainly reinforced our understandings. I could throw it out to the panel, for those of you who may not have read it. Who do you think did best? They looked at whether you had primary cytoreduction versus interval cytoreduction and if you had residual disease or didn’t have residual disease. Then they looked at what those hazard ratios were within the PAOLA subgroups. What they found was that if you had the primary cytoreduction, your hazard ratio was 0.52 versus 0.66. But with no residual disease, you got it all the way down to 0.47, and that was a big difference compared with 0.61. It’s a pretty big difference there, approximately 14%. With residual disease, there wasn’t a difference. It was really no residual disease and primary cytoreduction: those patients did quite a bit better. But the treatment effect, other than with residual disease, was across all subgroups.

Thank you for that. And again thank you, Katie Moore, for leading the meeting and for getting these out in the public domain.

Ovarian Cancer: Factors in Deciding on Maintenance Therapy 

The other question is, why don’t you use both? Why don’t you use bevacizumab [Avastin] and PARP inhibition together? We had that published in the 

, and the approval is in process. Sharyn, tell us about the study that we call PAOLA-1. There’s a 1 after it just like SOLO-1, because it’s a first-line study. Go ahead.

 PAOLA-1 was 1 of the groundbreaking discussions that started at ESMO [European Society for Medical Oncology Congress]. This was a trial that enrolled women with stage III and stage IV epithelial ovarian cancer. Patients were enrolled regardless of surgical outcome, which was a little different from PRIMA. Women who were eligible for the study all had to have either a CR [complete response] or a PR [partial response] following carboplatin, paclitaxel, and bevacizumab therapy. This was a European trial. It was enrolled throughout 8 different countries. Bevacizumab was utilized in the frontline setting with chemotherapy and then as maintenance. These women were then randomized, if they had a CR or PR, to either bevacizumab alone or olaparib and bevacizumab. The bevacizumab arm was continued for 15 months and the olaparib was continued for up to 24 months, or until there was some type of unacceptable toxicity or progression.

The primary end point on this study was time from randomization until investigator-assessed disease progression or death. Certain subgroup analyses included PFS [progression-free survival] by BICR [blinded independent central review] and PFS2 [progression after the next line of therapy]. There are also secondary end points, and tumor was analyzed for HRD [homologous recombination deficiency] via myChoice CDx. In total, over 800 patients were enrolled. Baseline characteristics were well balanced between the groups. It’s interesting that about 30% of patients had stage IV disease. Most of the patients were actually NED [no evidence of disease], going to either a complete cytoreduction or having a CR after frontline therapy. It was very interesting.

The duration of investigator-assessed PFS was significantly longer in the olaparib and bevacizumab arm than in the bevacizumab alone arm, with a median of 22 months versus 16.6 months. That led to a hazard ratio of 0.59. There was essentially a 41% risk reduction of progression or death if receiving the combination of olaparib and bevacizumab versus bevacizumab alone.

Looking at the various subgroups, for patients who had HRD-positive 

-mutated tumors, median PFS was amazingly 37 months with the combination of olaparib and bevacizumab versus only 17 months in the bevacizumab-alone arm. That was a hazard ration of 0.33. If you’re looking at patients who had HRD positivity without a tissue 

 mutation, the median PFS was 28 months in the combination arm versus 16 months in the Avastin-alone arm. That was a hazard ratio of 0.47. Interestingly enough, in the patients who were negative for HRD or had an unknown status, we saw that the median PFS was the same between the 2 groups: 16 months in both arms, and the hazard ratio was 0.92.

We did not see any new safety signals. As I mentioned, the trial did meet its primary end point. It was very impressive to see the response of the 2 drugs together—bevacizumab and olaparib—in these patients.

 PAOLA-1: Combination Maintenance Therapy in Ovarian Cancer 

Let’s just assume, Katie, that you saw a patient who comes into your office and says, “I responded to my frontline chemotherapy. I had a serious advanced ovarian cancer, and my doctor did this HRD [homologous recombination deficiency] test; here it is.” And she says, “My doctor is not very passionate about recommending that I get PARP maintenance, and I don’t have a 

Would you say—with the same passion you had for SOLO-1—to this patient that this is a no-brainer. You need PARP maintenance, you have a molecular marker, as Tom alluded to. This would be more than a doubling of the median progression-free survival. The median for an HRD-positive patient, excluding 

, has a hazard ratio of 0.5—Tom, what you said—going from 8.2 to 19.6 months. It is more than doubling of the progression-free survival. What would you tell that patient, Katie? I’m going to ask you, Tom Krivak and Sharyn Lewin, the same question.

 You asked me about my passion. I do believe that PARP inhibitors are the standard of care with shared decision-making for patients who have 

-associated cancers. For patients without 

-associated cancers, they are a standard of care, but there’s a spectrum in there. This would be on the stronger edge of evidence. The HRD group was 1 of the analytical subgroups of PRIMA, and it met its end point. I believe that, and we’ll get into the debate later. I would strongly counsel this patient regarding the potential benefit of use of niraparib, and then the known risks and benefits. I would especially talk about what the gains would be in terms of progression-free survival and what that means, without an overall survival end point that we won’t have for quite a while; though in PRIMA, we may get it much earlier than on some of the other studies. Then it would be shared decision-making between the patient and me.

I think this is a little different in that there are options for both. Some patients just want to do surveillance. I don’t know if I’d beg someone to start maintenance, which I do with 

-associated cancers. But I would strongly recommend this. I think it makes a lot of sense. We know these patients are going to recur. We just have to stop talking about all these cured high-stage patients. They don’t exist. I really wish they did, but they don’t. Try to prolong response as long as possible. I just don’t know why you wouldn’t.

I like this idea of shared decision making. But let’s face it, when you give the patient the option of taking a pill that might have some adverse effects versus going on vacation—trust me, you can go on vacation and take a PARP inhibitor too—they have a tendency to take the easy way out. It’s just human nature, right? Every patient thinks they’re cured. You say they don’t exist. Every patient thinks they’re the exception, and I don’t know why the doctor thinks they’re the exception, but there are many care providers who do.

 If you present observation as an equivalent option—and it’s listed like that in the NCCN [National Comprehensive Cancer Network] Guidelines, which is mind-boggling—then they’ll think, “Oh, I could do something or not do something? I’m going to not do something.” We invoke the counseling.

That’s what I’m trying to discuss here. It’s the strength of the recommendation in everything we do. Whether it’s a Gardasil 9 vaccine or PARP maintenance, it’s the strength of the recommendation. Sharyn, how strong would you recommend this? HRD in hand, a patient who didn’t receive bevacizumab and responded to frontline therapy comes to you and says, “Dr Lewin, you have a great reputation. Do I need maintenance niraparib?”

 Honestly, I agree with everything that Katie said. I want to take a minute to highlight the importance of frontline germline testing, which unfortunately not 100% of women with ovarian cancer receive. There were data at the SGO [Society of Gynecologic Oncology] Annual Meeting showing about 60% do. I don’t know why it’s not quite 100%. But I agree, it really is shared decision-making. I am a big biomarker believer. I know the inherent problems with the subgroup analyses. But I think if the patient does have HRD—and of course if she has a 

 somatic or germline mutation—I’m going to strongly encourage PARP maintenance.

Tom Krivak, how strong is the recommendation? HRD in hand.

 We talk about financial conflicts. What I can say is that I have a biomarker conflict too, in that I believe in HRD and I believe in maintenance. When I give talks, I automatically say that I graduated fellowship in 2002. I thought Maurie Markman’s trial of 12 monthly cycles versus 3 maintenance paclitaxel was great, and I’ve used maintenance therapy ever since. It’s important because what’s really moved the needle has not been the up-front 6 cycles of whatever therapy we give: the maintenance therapy has moved the needle. To me, if somebody has an HRD-positive tumor, I would strongly recommend maintenance, whether it’s single-agent niraparib or a combination therapy. But I would have to say part of my bias is that when I see folks with advanced-stage ovarian cancer, I’m going to recommend some form of maintenance therapy, and I don’t believe in active surveillance in that group.

People want to know what your conflict of interest is. Trust me, your conflict of interest is not because someone paid you to go to work, whether it be a drug company or your hospital. Your conflict of interest is that you have seen patients benefit and stay alive and live better with maintenance therapy. I get that. In my mind, it’s evidence and evidence based. For the audience, I’ve got 4 panelists here and myself to offer—with a strong recommendation—a maintenance PARP inhibitor in 

mutated disease. With an HRD molecular signature, you can maybe soften it a little. Your head doesn’t have to explode, to quote Katie Moore, but my head would explode if someone says, “Well, I don’t know.”

Practical Applications of PRIMA in Ovarian Cancer Management

mutations, maybe two-thirds or three-quarters of those are germline. We want to go beyond 

, and clearly PARP inhibitors work beyond 

: we’ve shown that in the recurrent setting. There’s now a third study that was also published in the 

by Antonio González-Martín called PRIMA. Tom, tell us about PRIMA and what the opportunity is here beyond 

 It is also known as GOG-3012, right? We can’t forget that. This compared maintenance niraparib with placebo, and niraparib was given as maintenance for 36 months in this trial after patients completed primary chemotherapy. Eligible patients had stage III with gross residual disease after primary surgery or had stage IV disease, and this trial included patients who had neoadjuvant chemotherapy as well. Both serous and endometrioid histologies were included in this trial, importantly, and to be eligible, you had to have a response, either partial [PR] or complete [CR], to previous chemotherapy.

Brad, you’ve shown this with your weights-and-plates work in the past in terms of impact, especially on thrombocytopenia, for those patients who had a platelet count less than 150,000 permm3 or who weighed less than 77 kg. You actually reduced the dose. That happened about two-thirds of the way through this trial. That’s important in terms of interpreting some of the toxicity data.

The patients were randomized 2 to 1. There were 733 patients who were randomized in total. Importantly, they used hierarchical step-down in terms of the statistics. They actually did stratification based on whether you received neoadjuvant chemotherapy or not, what your HR [homologous recombination] status was, and whether you had CR or PR. Those were the 3 stratification factors. The primary end point in this was homologous recombination deficiency [HRD], and we can get into discussing what exactly that is and what that means. But the primary end point was the HRD population. If that was positive, then they were able to go on to the other populations, and the second population they looked at was the entire intent-to-treat population. That’s important to understand. The primary end point that they used within these groups or cohorts was progression-free survival [PFS] as determined by BICR: they had radiological blinded independent central review to determine if you actually had progression or not.

In their first assessment for the group that had HRD, they found a significant improvement in the median. The PFS was 21.9 months for the group that received niraparib and was 10.4 months for the group that had placebo. The hazard ratio was very impressive, 0.43, so there was a 57% reduction in the hazard of either relapse or death. If you look at it in the intent-to-treat population, it was still impressive: 13.8 versus 8.2 on the medians, with a hazard ratio of 0.62. That’s in all comers who were in this study. The phase 3 patients had to have residual disease, so they were a fairly difficult group. When you’re looking at the medians, you have to be cognizant of what the patient population was.

Then there were a number of subgroup analyses that were looked at as well, getting back to what Katie just talked about. If you look at the 

 group, you’re down into the 0.3 range. If you break it out—Brad, you had data at the SGO [Society of Gynecologic Oncology] Annual Meeting this year looking at 

1 showed a slightly higher hazard ratio of 0.39. It was a little lower, 0.35, for our 

 and look at just the HRD, the hazard ratio is still impressive, 0.5. That’s interesting. If you look at your HR [homologous recombination] proficient patients, you still had a positive hazard ratio. It did not cross 1, and the hazard ratio was 0.68, that’s very impressive in terms of what you’re seeing with this data. Now, the medians were not very different. The difference was about 2.7 months, so you’d have to include that in how important you think that difference is.

I don’t think there were any significant new safety signals from this trial. We continued to see anemia, nausea, and thrombocytopenia—the big 3—largely as grade 1 and 2 events. But we still saw 28% nausea that was grade 3 or higher. In terms of thrombocytopenia, the amount definitely dropped down a little, but we still saw a significant amount of grade 3 or higher events that were present at about 29%. We still saw some of those safety signals that were the same.

What does this mean, Brad? It means you have a patient population where you could argue you could use a PARP in any of these subpopulations. I’d love to hear about these subgroup analyses that really aren’t powered and what they mean. I think people have different interpretations of what that means and how they’re going to apply it clinically. But I do think many people are going to look at these data and say, why wouldn’t I use it on everyone?

I wanted a comprehensive answer, and that was comprehensive. The impact of this study deserves a comprehensive assessment as you just provided. I’m not going to debate yet—we will—whether HRD testing should be performed.

PRIMA: Frontline Niraparib Maintenance Therapy in Ovarian Cancer 

-associated disease. We have Katie Moore here, who was the first author published in the 

 related to the approval in December 2018. Tell us about that study and what that has meant to our practice in the frontline setting.

 As you know, I love to talk about SOLO-1. SOLO-1 was a randomized phase 3 study that enrolled only women with 

-associated cancers, predominantly germline mutations in 

. They had to have advanced cancer, an attempted cytoreductive surgery, and at the end of platinum-based chemotherapy, complete or partial response to chemotherapy. At this point they were randomized 2 to 1 to receive olaparib tablets versus placebo until progression or toxicity or the 2-year mark. The primary end point was investigator-assessed progression-free survival [PFS].

That took a long time to reach, but it finally resulted and demonstrated a ratio of 0.3, which translated to a difference in the median progression-free survival—counted from the end of chemotherapy, importantly&mdash;of 13.8 months in the control arm and not yet reached in the arm that was treated with olaparib, but it is somewhere around 47 to 49 months. The PFS2 [PFS after 2 lines of therapy] was positive as well. There have been several subsequent exploratory analyses in all the different surgical and residual disease subgroups, demonstrating equal magnitude of benefit regardless of how you analyze it. That appropriately led to a change in the standard of care for these women.

, I define that as what reasonable physicians would do in a similar setting. It doesn’t mean it’s the only standard of care. Tom Krivak, is this the only standard of care if you have a 

, but I know we use it a lot. Given how the question is phrased, I would say that you must have a reason not to give it. To me, when you look at those data, they’re just so impressive that I completely agree. Sometimes there’s a slow adoption, but to not be on this therapy with the germline mutation as well as somatic mutation, I get nervous. I think if you would show patients some of these Kaplan-Meier curves when they say, “I don’t think I really need it,” and you ask if they want to be on the top line or bottom line, there is such separation that most patients would say, “I don’t understand why you aren’t making me take this.”

Thomas C. Krivak, MD

Articles