Advanced Ovarian Cancer - Episode 14

Case 4: I-O Therapy for Platinum-Resistant Recurrent Disease

May 23, 2019


Robert L. Coleman, MD:What other potential options do we have? You talked about clinical trials. There’s a lot of interest around immunotherapy. I feel like we’re still trying to find the right secret sauce for that, but we know we’ve had a couple of missteps. What do you think about immunotherapy in this?

David O’Malley, MD:I think immunotherapy is interesting. We haven’t figured out what patients are going to benefit and how to give it. JAVELIN Ovarian 200 was just reported on at our annual meeting for the Society of Gynecologic Oncology in Honolulu, Hawaii. We haven’t seen the paper, but we have seen the presentation. Unfortunately, it did not meet its primary end point. So they were not allowed to have other therapy for their platinum-resistant disease, but they were allowed to have 1 to 3 priors for platinum-sensitive disease. And then it was a 3-arm trial: avelumab, PLD [pegylated liposomal doxorubicin] plus avelumab, or PLD, 1-to-1-to-1 randomization.

Interestingly, when we look at that patient population, the primary outcome was not met, but there were some very interesting findings when we looked at the PD-L1 [programmed death-ligand 1]—positive population. The overall response rate was about 18%, 18.5% when we looked at the combination of avelumab plus PLD. That’s compared with only 3.4% in PD-L1–negative population.

Robert L. Coleman, MD:Before we go on, I wanted to ask our pathology expert a question. What are we assessing with this? Because it seems as if there are many different assays. For lung cancer, they got this; and each drug company has its own assay.

Gregory Riedlinger, MD, PhD:Dave hit the nail on the head. Obviously, there is so much excitement about immunotherapy. But how you actually define who’s going to respond is a big issue. Certainly with the MSI [microsatellite instability] population and certain DNA polymerases, it seems that you get high responses associated with tumor mutation burden. But there’s tremendous interest, or maybe not so much tremendous interest, because with the pharmaceutical companies it may not be in their best interest to necessarily narrow down the population.

But with the PD-L1, there are just some general things to keep in mind: There are a number of different clinical antibodies that are being used. And so for whatever study you look at, you want to look at what specific antibody was used there. And then there are different assessments and different cutoffs. As you touched on with lung, sometimes they’re looking only at the tumor cells. The positive tumor cells and the cutoff might be greater than 50%. And then in other studies, they’re looking specifically at the tumor cells and the immune cells.

In the JAVELIN study, they had their cutoffs at 1%, greater than 1% in tumor cells, or greater than 5% in the infiltrating immune cells. But even with that, where is this biopsy coming from? Was the staining performed on a lymph node versus a primary site? I think there are a lot of caveats to the interpretation of this. But there does generally seem to be some type of benefit, obviously, in the PD-L1—positive population, which seems to make sense with the way the drugs should work.

David O’Malley, MD:In the way they define it, about 55%, 57% of patients were defined as PD-L1—so about half or just over half by their definition. We don’t really know what the definition in gynecologic cancers really means.

Robert L. Coleman, MD:Yeah, it’s all over the place because of what he just mentioned. In KEYNOTE-100, on the ovary subpopulation—also a very large group of patients—your greater-than-half, 50% score was about a quarter of patients. The less-than-1% was about a third of patients. There are all these patients in the middle, so it’s like wherever you draw the line is going to give you whatever that proportion is. It’s interesting. Some of the very earliest data we have in platinum-resistant patients—maybe this is the wrong population to study—there are responders in expressers and nonexpressers. We had that data right away. So we’re getting a larger denominator, but I’m not so sure that we’re actually increasing the numerator by all that much with all this intervention that we’re doing, in terms of the efficacy. At the Dana-Farber Cancer Institute, how do you use immune checkpoints?

Susana M. Campos, MD, MPH:We don’t, actually. When you look at all the checkpoint inhibitors that are out there, of course we’re looking at response rates anywhere from 5% to, at best, 20%. And so we don’t necessarily test for PD-L1. If we choose to give someone a checkpoint inhibitor, it’s usually in the context of a clinical trial, which kind of brings me to an old point. We’re looking at immunotherapy in conjunction with good old-fashioned chemotherapy. But are the new, more novel options combining a PARP [poly (ADP ribose) polymerase] inhibitor with immunotherapy? What’s the rationale? Is the rationale only in aBRCA-mutation carrier, versus, for example, the TOPACIO study? Granted, that was a small study. It was 25% overall objective response.

Robert L. Coleman, MD:It was kind of interesting, right?

Susana M. Campos, MD, MPH:It was kind of interesting. What’s the idea behind that?

Robert L. Coleman, MD:You said “kind of.”

Susana M. Campos, MD, MPH:It’s a start.

David O’Malley, MD:But it also brings us into a whole different conversation with how well PARPs in this patient population...

Robert L. Coleman, MD:I was just going to ask you...

David O’Malley, MD:I want to look at this because I think there are some interesting data out of JAVELIN that we should discuss. Across our immunotherapy trials, as a single agent, it’s amazing—5% plus or minus 5%, as low as 0%, up to 10%, but nothing beyond that. And we saw the same thing in JAVELIN. Avelumab alone was 3.7%. PLD—this comes to this platinum-resistant population. I quote my patients as 15% to 20%. In JAVELIN Ovarian 200, it was a 4.2% objective response rate.

Robert L. Coleman, MD:By BICR [blinded independent central review]. It was by BICR.

David O’Malley, MD:By BICR, plus or minus.

Robert L. Coleman, MD:All the data you are quoting is by investigator assessment. But you’re right. It is lower than expected.

David O’Malley, MD:Correct. And when we use the combination, it did go up to about 13%. That’s in the entire population, not the PD-L1. When we’re looking at the differences between this—again, when we look at the avelumab arm in all comers—it really looks like the natural history of the disease.

Robert L. Coleman, MD:Yeah, right.

David O’Malley, MD:Sixty percent fell off at the first assessment, which was at 2 months. Another almost 30% fell off at the second assessment. So by itself, it is probably not where we should be going right now. Now, all checkpoint inhibitors are not created the same, so maybe there are some other data out there.

Transcript edited for clarity.