Markman Proposes Shift in Ovarian Cancer Clinical Trial Paradigm

Maurie Markman, MD, discusses the need to change the paradigm of clinical trials and FDA approvals in ovarian cancer.

Maurie Markman, MD

Historically, FDA approvals of oncologic agents have occurred as a result of positive phase III trial results. However, Maurie Markman, MD, proposes that in this era of precision medicine, phase III trials with the endpoint of overall survival (OS) should not be the be-all and end-all in respect to drug approvals—particularly in ovarian cancer.

“There is a risk that we might discard potentially valuable drugs because of this irrational endpoint,” says Markman, president of medicine and science, Cancer Treatment Centers of America, and editor-in-chief of OncologyLive. Instead, he suggests allowing other endpoints and earlier-phase or basket studies as the basis for future regulatory decisions.

OncLive: Can you discuss your presentation on the treatment paradigm of ovarian cancer?

In a lecture during the 35th annual CFS®, Markman discussed the need to change the paradigm of clinical trials and FDA approvals for ovarian cancer.Markman: My talk at this meeting was to basically challenge the existing paradigm in clinical trials in the United States and worldwide in ovarian cancer. There has been so much progress with new drugs and new strategies, and as a result, phase III randomized trials looking at OS as a primary endpoint is simply no longer rational.

The reason for that is if you conduct a trial with “regimen A” versus “regimen B,” patients progress, then go off of the trial, that would be progression-free survival (PFS) or disease-free survival as an endpoint—which was what the trial endpoints have to be now. However, if you looked at OS as an endpoint, that would assume that what happens to a patient after they go on that trial is going to be the same on either arm of the trial—and that is a completely irrational endpoint.

We are not talking about patients who might live 2 or 3 months after they finish a trial; we have patients who are living 1 to 3 years or longer after they finish a trial. Therefore, holding that drug or combination—which a patient would have received 2 or 3 years ago—hostage to a survival endpoint when all of the things that might have happened to the patient population during the interim is not controlled simply makes no sense. Stating that doesn't mean that an individual drug or strategy can't improve OS—that is not what I am saying. If we had this wonderful new drug or strategy and we do a trial and hit a “grand slam” with survival, that would be wonderful. That is not the point. The point is that we could see an enormously positive outcome.

The other very important issue that I want to comment on is not just the issue of PFS in phase III trials, but in fact the new paradigm that we have, which is basically precision cancer medicine. When you have a situation where you are dealing with a relatively uncommon cancer, such as ovarian cancer, the likelihood of completing a phase III randomized trial is low. Therefore, we have to come up with other strategies to evaluate efficacy. We don't have to do clinical trials, but is it the only way to move a drug forward into the non-research domain with phase III randomized trial in the era of precision cancer medicine?

Are there any steps being taken toward this?

We need to come up with a novel approach. That could be looking at high objective response rates (ORRs), or time to disease progression compared with a well-categorized historical control. It could potentially be having patients’ own natural history of disease characterize their time to progression. For example, [they could go] on this novel therapy and the time to progression was 50% to 100%. There are a variety of ways we could look at this, but the point is that we need to come up with strategies that are acceptable to the regulatory agencies, insurers, and ultimately, the patients. There is no question that what I am describing and advocating for is happening. The extraordinary poster child for that is the approval of a checkpoint inhibitor [pembrolizumab (Keytruda)] based on robust phase II trial data with microsatellite instability-high (MSI-H) tumors, and I applaud the FDA for this. What they have done is extraordinary important for patients.

Other than OS, what other endpoints might be worth exploring?

Obviously, it is my hope that we see many more examples of this in ovarian cancer. For example, MSI-H occurs in 2% of patients with ovarian cancer. It would take 20, 30, or 40 years to do a trial only in ovarian cancer where one looked at this. The response rates in [the FDA approval] were more than 40%, with many of these responses lasting more than 6 months or a year. If a patient has ovarian cancer and their tumor is MSI-H, they now have the potential to benefit from this therapy because of the FDA’s action. The FDA needs to be encouraged for these kinds of events. You can take ORRs, as they are measurable, and the duration of responses in combination with that. One can make a very valid argument that the tumor shrank, but how valuable is that? The duration of response as well as percentage of response could be a composite.

What about a situation where you didn't have easily measurable disease? This is common in ovarian cancer. There, you might not be able to use measurable response, but you could use time to disease progression or time to stable disease. These are some objectively measured endpoints, but then you will want to combine it with some other parameter.

Again, it could be the patient’s own time to progression on their prior therapy. You can add to that the toxicities, of course, symptoms of the patients—if they subsided. Symptom improvement is critically important, but sometimes the measurement of symptoms is difficult. We talk about quality of life (QoL) and that is very important; however, the QoL related to the improvement of the symptoms versus the QoL related to the toxicities of the therapies, sometimes get mixed. Therefore, in a setting where one tries to be objective and measure, measurement is hard. It is not a question of importance of symptoms or QoL, it’s a question of measuring objectively.

Is there anything happening in ovarian cancer that is resembling this?

My concern is not rigorous research or measurable outcomes; it is the classic reliance on phase III randomized trials. These have done as well in the past, but we must come up with other ways to measure benefit so we can take the enormous knowledge that we are gaining and convert that into therapies of benefit. There are a lot of trials ongoing. We are seeing basket trials, which are one of the very exciting strategies that are ongoing. These include different drugs for varying tumor types. For example, ASCO's TAPUR trial is an extraordinary trial; in fact, it is a paradigm-changing trial. A patient with ovarian cancer could be on that trial if they had a particular mutation and they would be included in that basket. If it turns out that a particular target with a particular drug meets satisfaction, hopefully, there will be an approval and ovarian cancer will potentially be included in that. This is similar to what happened with MSI-H and checkpoint inhibitors.

We have to figure out this trial issue, we have to learn how to do trials, gather information, and decide whether these strategies are of value or not. There are opportunities today in terms of understanding the biology; there are hundreds of drugs out there and we have to test them and find their impact.