Markman Highlights Key Areas of Investigation in Oncology

Oncology Live®, Vol. 23/No. 12, Volume 23, Issue 12

In Partnership With:

Partner | Cancer Centers | <b>City of Hope</b>

Maurie Markman, MD, offers expert insights on the hurdles facing community oncologists, the uptake of biomarkers in the clinic, and an overview of what to look forward to in his area of expertise: gynecologic oncology.

In an interview with OncologyLive®, editor-in-chief Maurie Markman, MD, offers expert insights on the hurdles facing community oncologists, the uptake of biomarkers in the clinic, and an overview of what to look forward to in his area of expertise: gynecologic oncology.

What are some of the hurdles in the translation of molecular testing strategies into community settings?

The question of the use of genomic testing for targeted therapeutics is a critically important one in oncology. There is no question that there are major advances occurring weekly and monthly in the cancer sphere in many diseases, [but] that is a problem. The community oncologist who is taking care of patients with many different types of tumors is simply overwhelmed. Every oncologist wants the best for his or her patients, but with the number of new genetic tests for unique markers that they [must] understand across dozens of different diseases, it makes this [almost] impossible. And there is no universal decision support mechanism available to them to help them and, as a result, they fall behind.

For example, in the gynecologic cancer space, [the oncologist] looks at probably the best, most important single marker, which is BRCA, because we know the value of this testing. A recent study showed that 40% or fewer of patients, looking at electronic health records, have had the test. The recommendations from ASCO [the American Society of Clinical Oncology], the Society of Gynecologic Oncology, and the National Cancer Institute are very clear, but it is not being done—not because the doctors do not want to do the best for their patients, but because they are simply overwhelmed. And we need to figure out how to do something about this.

Lung cancer is another good example. We have progressed so far, yet we still have doctors who are trying to appropriately figure out the best tests to order for their patients. And we now have multiple molecular tests, multiple different drugs with recommendations for frontline and second line, and multiple immune-therapeutic agents in 1 disease.

How can clinicians who take care of multiple different tumor types, including lung cancer, stay up to date on all the things that are going on in lung cancer? It is a very simple issue. We have got to figure out how to develop decision support mechanisms for doctors in the community.

What strategies should be further explored in ovarian cancer and beyond?

One is we have got to come up with decision support in the oncology community to figure out how to use these incredibly important biomarkers and targeted therapy. This is true across the board. The fact that data have come out from several sources, looking at electronic medical record use, that less than 40% of patients with ovarian cancer will absolutely need BRCA testing and not have BRCA testing is a huge problem. We need to figure that out.

The second thing we need is we [must] figure out how to do these comparative effectiveness trials. We have [several] drugs that have been approved and a number of strategies approved, and we are seeing more of them. [These are agents] that are approved by the FDA in the first line, second line, third line. The treating oncologist and their patients are going to ask, “What’s best?” and there will not be an answer because we have not done those kind of [comparative] trials. We have got to come up with a strategy.

If the patient asks what will work best for me and the oncologist says, “I don’t know,” where are we? And I would argue we’re still in the dark. But we can answer those questions. We don’t need 1000 patient studies [to] get a handle on [how these agents compare]. If they’re relatively the same, that’s great. You can pick based on toxicity. You can pick on the basis of efficacy. You can pick on the basis of convenience. But we won’t know any of that because we will not have done those comparative trials.

The problem in the absence of comparative trials is that 1 trial may be an entirely different patient population that were compared with another. Beyond that is figuring out decision support for oncologists and figuring out how we’re going to do these comparative effectiveness trials.

Since the introduction of immune checkpoint inhibitors into clinical practice a decade ago, we have been hearing about the need for predictive biomarkers in response to treatment. We do not seem to have moved past PD-L1 as a biomarker, which often has been described as “imperfect.” Is there any progress on this front in terms of getting closer to more specifically defining the patient populations who would benefit?

It is clear we need biomarkers. For a number of diseases and drugs, immunotherapy is a classic example of that. And obviously we hoped we would be able to find them relatively easily, but this has not been the case. We had PD-L1 [expression as a marker] in some tumor types and it is actually predictive, not as a yes or no, but more as an expression of seeing the higher the probability of response. This is not perfect, because then you [must] ask, “Where is the cutoff?” as opposed to “yes or no?”

We also have the [tumor] mutational burden in tumors that has been used and FDA approved for several agents. Obviously, again, individuals who say that’s not perfect—nothing is perfect. One of the issues here is that we are looking for a universal marker in all tumor types. It is likely going to be the case that we will find some markers and tumor types that are more predictive than others.

There may be universal relevance for immunotherapy across the tumors. [Microsatellite instability–high] MSI-H is a wonderful example of a tumor marker that was discovered serendipitously in retrospect. It should have worked because now you have all these neoantigens that were unknown until the people who examined the question found it to be relevant.

Hopefully, we will find those biomarkers as we move forward. It may take a different understanding of pathways. You are often looking now for single abnormalities. It may be a pathway that we were looking at as opposed to a particular biomarker, which only makes things more complicated.

What are some of the areas of focus for the future of ovarian cancer?

With ovarian cancer, clearly bevacizumab [Avastin] remains, although it is a biosimilar, an incredible standard care option. And, of course, we have PARP [inhibitors]. What has been disappointing is the use of checkpoint inhibitors in ovarian cancer. There are specific indications including a small percentage of patients who have MSI-high [disease] who would be eligible for a checkpoint inhibitor or patients with high tumor mutational burdens. Again, it is a small percentage of patients with ovarian cancer, where an immunotherapeutic agent would be relevant, but in terms of using this in the standard of care with chemotherapy or to subsidy chemotherapy, [several] trials have shown no benef it for the addition of a checkpoint inhibitor.

I would have to say that though I am always optimistic. As one of my former mentors said, “You’re an oncologist; you have to be optimistic.” I am always optimistic. The data [we have] so far, whichever inhibitors of ovarian cancer, have not been positive. They are not positive to call this standard of care or to get FDA approval, except in the settings I noted.

It may very well be we simply that we do not have the right biomarkers. It also may be that we don’t have the right drugs. And there are other drugs that might work in this setting that need to be tested to be examined or it could be combination of checkpoint inhibitors. I am optimistic that there is a major role for immunotherapy in ovarian cancer. We just have not figured out what that is yet and I look forward to 2023, and 2024 being able to sort this out.