2016 Forecast Series—The Future of Immuno-Oncology with Patrick Ott, MD, PhD, Clinical Director of the Melanoma Center and the Center for Immuno-Oncology, Dana-Farber Cancer Institute
Dr. Patrick Ott is also an Assistant Professor of Medicine at Harvard Medical School.
OBR: Retrospectively, what stands out in your mind as highlights in immuno-oncology from 2015?
PO: The combination of checkpoint inhibitors ipilimumab and nivolumab leading to remarkable anti-tumor activity in melanoma and other cancers; combination of PD-1 inhibition with other agents that counteract immune-suppression in the tumor (for example IDO inhibitors); the increasing recognition of tumor mutations/neoantigens as important targets for immunotherapy including immune checkpoint blockade illustrated by the success of PD-1 inhibitors in tumors with mismatch-repair deficiency and many other preclinical and clinical studies.
Then there’s also the issue of why everybody isn’t responding to these therapies. One reason may be that not every tumor has an immune response already ongoing, and if there is no immune response in the tumor, then maybe the checkpoint inhibition may not necessarily work, or at least there is something else needed to trigger that. There are different names for this like “how to make a cold tumor hot” or “how to convert a non-inflamed tumor to an inflamed one.” Researchers have been looking into the mechanism of why some tumors may be inflamed and others not and are searching for therapeutics to attack this problem.
OBR: How does combination therapy tie in? Can or should we combine immunotherapy with chemotherapy?
PO: I am not a huge proponent of simply combining checkpoint inhibition with chemotherapy without a strong rationale. It’s done in many clinical trials using checkpoint inhibition, and it may work, but I think there’s more scientific potential in doing combinations with other immune-modulating therapies – adding a vaccine rather than chemo, for example. There are so many other different ways to stimulate the immune system that I don’t think chemotherapy would necessarily be my number one modality in terms of prioritizing combination approaches.
Many combination immunotherapy studies are done now because they can be done relatively easily from a regulatory and intellectual property standpoint, not necessarily because they’re scientifically the best thing to do. It is sometimes easier to do a clinical trial where you add immune therapy to an already FDA approved chemotherapy rather than combining with an agent that may not be FDA approved or for example be “owned” by another company. There are a lot of hurdles that prevent those trials from happening. So, there may be some role for combining immunotherapy and chemotherapy, but I don’t think it’s the most exciting avenue forward.
OBR: Bristol Myers’ Opdivo and Merck’s Keytruda have continued to dominate the I-O landscape in the past few years. Will we see these two knocked off their pedestals anytime soon, and what kind of 'next generation' therapies could take their place? Are there other products or will better selection of patients drive clinical success in the future?
PO: I don’t think [Opdivo or Keytruda] will be knocked off their pedestals because we know that the clinical data on PD-1 inhibitors are fairly similar in how they work and what they target. I wouldn’t rule it out, of course, but it’s unlikely that another agent will take their place just yet. The efficacy of PD-1 or PD-L1 inhibition as single agents is so far unmatched in many cancers, arguing that this approach may be the basis to build other therapies on, at least for the tumors where we see this single agent activity.
OBR: You don’t see big differentiation between the PD-1 inhibitors?
PO: At least 7 PD-1/PD-L1 inhibitors are in development (some of them already approved), but there’s not a big difference in how they perform, at least in the broad sense.
In general, PD-1 inhibitors are unlikely to go away any time soon. They have just arrived and are here to stay. The reason why we are thinking of PD-1 inhibition as a backbone for other therapies is that they are so effective and well tolerated. The future may involve a combination that has PD-1 inhibition in there, but in terms of next generation, it’s unlikely that PD-1 will fall by the wayside and be replaced by something else. There are no data suggesting that right now.
OBR: Do you think that patient selection with respect to genomics and susceptibility will produce higher response rates?
PO: Yes. I believe that patient selection in combination with other agents will get us there. Biomarkers are a fast-evolving field. We know that PD-L1 positive patients probably respond better to PD-1 inhibition, but this biomarker is complex and likely just the beginning of more sophisticated predictive markers or sets of markers that could better select patients that will respond to immunotherapy, and potentially it’s just the beginning. Other predictive biomarkers will enrich patient populations further and lead to higher response rates.
OBR: So we’re still optimizing how to use the drugs and understanding which patients need a combination or not?
PO: There will be other checkpoint inhibitors eventually, targeting molecules such as TIM-3 and LAG-3, for example. We will be profiling tumors for expression of those and other markers and potentially tailor treatment accordingly. There are many antibodies in development.
Then, of course, we could add more to the IPI + NIVO regimen, but this treatment is already fairly toxic. Nevertheless, it is certainly conceivable that researchers may find a triple combination that’s also tolerated. For example, the IDO inhibitors that I mentioned earlier haven’t added much toxicity to PD-1 inhibition, so maybe they could be a partner in a “triple” therapy.
OBR: Where are we in understanding how to implement or use a PD-L1 biomarker? Are we sure that it has clinical utility?
PO: Pembrolizumab was approved for lung cancer with a companion PD-L1 test, which is based on its clinical development incorporating careful analysis of PD-L1 expression and its association with response in the studies that lead to this approval. That being said, these markers are suboptimal: there’s no standard test, and it’s difficult to test, with a lot more complexity compared to a genomic marker. If you test for a mutation, that’s fairly cut and dry, but with PD-L1 expression, there are many factors that can give you different results. In the lung field early on, PD-L1 was probably a bit overemphasized and some clinicians and investigators were saying, “If this patient doesn’t have PD-L1, why would we even consider treating with a PD-1 inhibitor?” This “oversimplified approach” could actually lead us in the wrong direction. If you just focus on the patients that are positive for biomarkers that are not very well validated, many patients may unnecessarily lose out on treatments that could be beneficial for them.
OBR: One of the biggest cancer stories last year was Jimmy Carter’s successful treatment with Keytruda for melanoma that had metastasized to his brain and liver. What impact, if any, does a story like that have on prescribers and patients in trying these new therapies?
PO: The Jimmy Carter story is a very good example of patients hearing about something and thinking this is what they need, but I don’t think that actually impacts oncologists too much. For one thing, it’s way too early to tell how much he really benefits, and secondly, he received other treatments as well. He had resection and radiation in addition to the immunotherapy. Obviously I don’t know the whole case there, but I don’t think it’s a good example of saying, for example, Keytruda works and not Opdivo. I’m not sure that would be the right conclusion.
OBR: We also saw the immune checkpoint inhibitors move into blood cancer treatment this past year, and there was a lot of news about novel combination therapies. What do you think the future holds for these products in blood cancers?
PO: I’m not an expert in hematologic malignancies, but I would not necessarily look at them differently. We know they can work in lymphoma, and it works remarkably well in Hodgkin’s lymphoma. PD-1 inhibitors have received a special label by the FDA and may be approved for that Hodgkin’s lymphoma soon. I would look at them as malignancies that can certainly respond to immunotherapy. Some of the acute leukemias may be a bit tougher, though, and myeloma for example hasn’t really responded yet to immune checkpoint inhibition.
OBR: So, it’s too early to make broad claims, but you see promise?
PO: There’s certainly a role for immunotherapy in some hematologic malignancies, for example Hodgkin’s lymphoma, as I mentioned earlier. A bone marrow transplant itself which is used for the treatment of many hematologic malignancies is thought to work through an immune mechanism. So, scientifically and biologically, there’s a lot of reason to believe that immune therapy may work, even for leukemia.
OBR: At ASCO’s plenary, it was noted that a higher dose of Keytruda could potentially cost $1 million a year in a clinical trial, and that combining Opdivo and Yervoy to treat melanoma would be roughly $300K. The high price of cancer drugs was certainly a contentious issue and professional groups and physicians got involved in the whole value discussion. Are the costs of these new therapies truly equal to their benefit? What are some other downsides — other than cost?
PO: I think cost is a very important issue. It’s also very complex. The fact that there’s a lot of money to be made for industry is certainly driving the field, but also I think not necessarily the right things are always done. Too much money is actually not helpful, and I think that has implications on different levels.
For example, why do we need to have 6 or 7 different PD-1 or PD-L1 inhibitors being developed simultaneously? This redundancy is potentially a huge cost driver and not necessarily development in the right direction.
It goes back to your earlier question – will Keytruda and Opdivo fall from their pedestals? In a more resource-constrained environment, we would try to move on from there and not develop 5 other PD-1 inhibitors, but rather focus on additional targets, rational combinations, and novel approaches. A more collaborative approach with less redundancy of drugs and trials would be desirable. However I do realize that there are certain realities – and competition in industry as well as the potential for financial profit are absolutely driving forces that will ultimately be good for patients. On a society and health care economics level, we do have to be mindful though about the costs so that the system doesn’t break.