Treatment Sequencing in Advanced Ovarian Cancer

Kathleen Moore, MD:Despite the excellent responses seen in SOLO-1 and outcomes, we still have patients who are recurring. We hope it’s far fewer than in the past, and certainly even with bevacizumab and the advances there, we still have patients with recurrent disease. And so, the considerations are very similar, but there are some differences to how we think about the subsequent line of therapy.

Certainly, we look at the appropriateness for retreatment with platinum as a key indicator. This is based on how well the patient tolerated platinum last time, how well they responded, and the time period from that platinum-based therapy. The dogma has been that if that interval is greater than 6 months, you would reuse platinum.

What has changed, however, is this dogma that if it’s less than 6 months, you would not reuse a platinum. That has really been challenged among patients who responded beautifully to platinum in the last line of therapy and only recurred when you stopped it, versus those patients who didn’t respond at all.

So 1 difference is that we’re really looking at patients who don’t fail but their tumors fail. Patients whose tumors fail to respond to platinum are really inappropriate for platinum again. Patients whose tumors respond to platinum—the longer they respond the better, but even if they recur in the last 6 months—there’s still a consideration for platinum. And so, that definition is changing, which I think has helped patients get re-exposed to the best drug we have so far in ovarian cancer, and I think that’s helped push out some of the progression-free survivals we’ve seen in the recurrent setting.             

That’s 1 change regarding the definitions and how you categorize patients into these recurrences. Of course, you still consider residual toxicities, and patient decision making, regarding schedules and all of those sort of nuances that are very important to select the next line of therapy. But here, again, knowledge of the patient’s genetic profile. Does she have aBRCAmutation or another mutation that may be a potential benefit for PARP [poly ADP ribose polymerase] inhibitors, such asRAD51C, potentiallyPALB2? We’re thinking about these and are doing studies. Does she have a genetic profile that would make you want to use a PARP inhibitor if she has not seen a PARP inhibitor yet? And have you looked in her tumor? Does she have anything in the tumor? Specifically, a somaticBRCAmutation, but all of those other mutations can also be somatic and again, would make her likely to respond to a PARP inhibitor in the recurrent setting.

The other wildcard is this homologous recombination deficiency [HRD] assay, which doesn’t test for those specific genes. What it looks for is evidence in the tumor that the tumor struggles to fix its DNA. If the tumor struggles to fix its DNA, that is a predictive biomarker of a better response to a PARP inhibitor than someone who does not have evidence of that in their tumor. And so, we have seen several studies. NOVA evaluated niraparib as a maintenance strategy following a response to platinum in the recurrent setting. ARIEL3 studied rucaparib in the same population. And SOLO-2 and Study 19 studied olaparib in that population. Each of them had biomarker programs. They’re all a little different.

But in summary, you can say that patients who had aBRCAmutation, either germline or somatic, did the best if you used a PARP inhibitor following response to platinum-based chemotherapy. And again, this is all PARP inhibitor-naïve patients. Patients who had evidence of homologous recombination deficiency in their tumor did next best. And then, those patients who had neither a mutation ofBRCAor evidence of homologous recombination deficiency did the least best with a PARP inhibitor, but they still benefited from the PARP inhibitor.

So while homologous recombination deficiency assays are helpful, they don’t delineate a population of patients who don’t benefit from PARP inhibition. That’s 1 of the struggles we have. Again, in the second-line, or the platinum-insensitive recurrent setting, when you have a patient sitting in front of you, you have to make a decision about whether you’re going to use bevacizumab with the chemotherapy, which, again, improves responses. That’s incredibly important when you have a symptomatic patient. And then, bevacizumab maintenance. Or are you going to just use chemotherapy and hope for a good response, which happens about 50% of the time, and then hope to put her on a PARP inhibitor? Again, this assumes a PARP-naïve population.

That’s a very different scenario than the case we discussed previously. There isn’t a head-to-head study of bevacizumab as a maintenance strategy in that population, and PARP inhibitors as maintenance. So outside of a patient with aBRCAmutation, you really can’t say which one is better. The studies that evaluated these 2 agents are entirely different populations. I’ve listened to a lot of experts kind devise an algorithm that works for them. In my practice, I don’t want any non-BRCApatient, at this point, to not have access to a PARP inhibitor. As I said, there’s nothing that tells me yet who shouldn’t get one.

This may be the only opportunity in this setting for a patient who did have aBRCAmutation to access a PARP inhibitor. And so, I tend to use a PARP inhibitor here, use platinum-based chemotherapy—whatever doublet makes sense for my patient—try to get them into a good response, and then move to a PARP inhibitor, just trying to avoid the potential that, should she be platinum-resistant, I can’t give her a PARP inhibitor later. Other experts really believe in the overall survival benefit demonstrated with bevacizumab. It’s modest, but it’s there. And so, they’re going to start with bevacizumab. More people probably start with bevacizumab and then decide later what they’re going to do, for which there’s no data, but a lot of people do that.

I think this line of therapy has become very complex. Now we’re using PARP in the frontline, and we’re going to have fewer PARP-naïve patients coming in. So I think this is an area where we’re going to see shifting sands in terms of standards of care.

Transcript edited for clarity.

Case: A 58-Year-Old Female With Heavily Pretreated Recurrent Ovarian Cancer

H & P

• 58-year-old female diagnosed in 2014 with stage IV ovarian cancer
  • Pathology: high-grade serous carcinoma, epithelial ovarian cancer
  • CA-125: 460 U/mL
  • CT with contrast of the pelvis, abdomen, and chest revealed a 4-cm mass in the left ovary and peritoneal carcinomatosis
  • Patient underwent suboptimal debulking surgery; residual disease 1.5 cm
  • Received IV/IP carboplatin/paclitaxel (6 cycles); achieved complete remission
• 2 years later (2016) symptoms returned; CA 125, 255 U/mL; ECOG: 1
  • Received carboplatin/paclitaxel (6 cycles) and bevacizumab; achieved good partial remission; CA 125, 45 U/mL; continued on bevacizumab maintenance
• 1.5-years following second-line therapy (2017), again presented with symptoms; CA 125, 550 U/mL; ECOG: 0
  • Genetic testing;gBRCA1/2wild-type
  • Received carboplatin/gemcitabine (6 cycles); CA 125, 46 U/mL; achieved complete remission
• Currently:
  • CA 125, 620 U/mL
  • CT shows several small masses in the lung left lower lobe (largest is 3 cm)
  • ECOG: 0