Charles Drake, MD, PhD
Finding a robust positive predictive biomarker for immunomodulatory drugs is difficult, says Charles Drake, MD, PhD, as the immune tumor microenvironment is so complex.
In an interview at the PER® 1st Annual International Congress on Oncology Pathology™, Drake, director of Genitourinary Oncology at NewYork-Presbyterian/Columbia University Medical Center and associate director for Clinical Research at the Herbert Irving Comprehensive Cancer Center, shared his insight on the promise and challenges with immunomodulatory drugs, and touched on immune-activating agents in prostate cancer.
OncLive: What therapies do you characterize as immunomodulatory?
That is a great question. Currently, when people talk about immunomodulatory therapies, most people think primarily of drugs that modulate immune checkpoints. Particularly, the antibodies that block the interaction between PD-1 and PD-L1. So, that is a series of 5 monoclonal antibodies that are FDA approved. The other immunomodulatory antibody blocks CTLA-4, that is called ipilimumab (Yervoy), and that is FDA approved in melanoma at the current time, but is being tested in combinations in other tumor types.
In addition to PD-1 and CTLA-4, there is a whole series of immune checkpoints on T cells. There are probably about of 10 to 20 important ones that are being tested in various phase I and phase II trials.
So, I think I would say, when we talk about immunomodulatory drugs, we are mostly talking about those PD-1/PD-L1 and CTLA-4 antibodies.
What are the key factors to look for when selecting patients for immunomodulatory therapies?
A potentially perfect biomarker would be very discriminatory, that is, it very carefully assesses patients who would or would not respond. [Biomarkers] are judged by their positive predictive value and their negative predictive value. Either one would be okay, but if you have a really perfect negative predictor, you could select patients who have no chance of responding—that would be interesting.
I personally think it is going to be very hard to have a very robust positive predictive marker because the immune microenvironment is complicated, it has multiple cell types including T cells, macrophages, other cells types. Also, the tumor has heterogeneity without it—that is one of the challenges in developing a biomarker. But really, we are looking for something with good discriminatory power for our patients.
It has been suggested that immune-activating agents in prostate cancer would be more effective if given earlier. As a member of the SITC guidelines panel for immunotherapy in prostate cancer, what are your thoughts?
In prostate cancer, there are some retrospective data from the drug sipuleucel-T (Provenge), which is a vaccine against prostatic acid phosphatase. These are retrospective data that looked at patients based on their PSA level. And it turns out if you look at patients with the lowest levels of PSA—that is the earliest disease—it really looks like, retrospectively, the clinical benefit is higher in those patients than in patients that have a higher PSA or are further down in the disease process. You have to remember that those are retrospective data, so those data really need to be repeated in a prospective fashion, moving forward.
There is similar data, though, in other tumor types with other kinds of immunotherapy. There is a general sense that using these kinds of drugs earlier in the disease process might be helpful. But, I have to say with the checkpoint antibodies we have all seen really beautiful responses in fairly late-stage patients with large tumor burdens. While it would be helpful to use patients with lower tumor burdens, it is possible that at least some patients with a lot of tumor burden can respond.