The Unmet Need in Immuno-Oncology and New Drug Approvals

Benjamin P. Levy, MD: I think that the biomarker is the first unmet need. We do not have, in my mind—and you could ask medical oncologists, and they may give you different answers—a reliable biomarker yet. I think PD-L1 is good, but I certainly don’t think it’s perfect; it’s not fixed like any EGFR mutation, it’s dynamic. So, I think the search and the work needs to be done to better define patients. We have to remember, even with combination strategies, we are not getting a cure and that’s what we really want. We want rational combination strategies that are safe and that can lead to long-term outcomes.

If you look at the data with CTLA4 antibodies with PD-1 drugs, depending on the data you look at, the response rate can be anywhere from 20% all the way to 70%. It can depend on a small number of patients versus a large number. And so, we still have a lot of work to do in that space. I would say that perhaps another unmet need is how these drugs work in the curative setting, just like targeted therapies.

We have to explore immunotherapy, post lung cancer resections, as adjuvant treatment. And an arm of the ALCHEMIST study, which is a very large study and a national effort, is going to look at this, giving nivolumab for patients after chemotherapy to see if these drugs lead to better cures. And then also in the neoadjuvant setting, giving it before surgical resection. You’ve identified an early stage lung cancer patient. It’s giving these drugs to patients before going to surgery and seeing what the drugs do to the tumor.

So, I’ve said this before: I think we are really in an era that we’ve just began to understand. We are in its infancy in terms of really understanding the fascinating, complex interplay between the immune system and cancer. It is not the targeted therapy paradigm. This is very different. This is a whole different animal. As somebody who has an interest in immunology and somebody who is privy to the data, even someone like myself, I struggle with trying to understand how these are all working together. And I also struggle to understand every single new combination that’s out there and why they’re doing it that way.

I think breakthrough designations are reflective and indicative of how many good trials and good therapies we’re not delivering to patients. And I think it’s very exciting. As I said before, we’ve had nine to 10 immunotherapies, monoclonal antibodies, or targeted therapies approved for lung cancer in the past 5 years. And that phrase of “breakthrough designations” is reflective of a changing paradigm; 6, 7 years ago, we had two or three drugs that we used in lung cancer. It is fascinating to see the groundswell of therapies and all the different treatment options.

I think having these breakthrough designations, and the FDA doing due diligence, but yet getting them accelerated for approval based on the trials, is a win. It’s a win for doctors. It makes doctors’ lives a little harder because there’s more options—but that’s good for patients. Patients who have more options, it’s a win. It’s going to extend survival for these patients. If we look at 10 years from now, or even 5 years from now, we’ll see median survival times for stage IV patients truly moving in the right direction.

The metric for what should be used as an endpoint for a clinical trial is evolving. If this question were to be posed to me 5 years ago, I would have said overall survival. But things are changing, and I think the NCI has logged on to this, too. That response rate may be enough to get drugs approved. The problem is that you’re defining rare genotypes in lung cancer. And you’re not going to have enough patients to do a phase III study. If you have a great drug that works very well in a rare, but susceptible, molecular niche cohort, and response rates are 50% to 70%, you don’t need that phase III trial for overall survival.

We don’t have a ROS1. We have an approval for crizotinib for ROS1, a rare rearrangement in lung cancer that has nothing to do with a phase III trial. It was a response rate that was quite high. We’ll never be able to do a phase III trial in that setting. So, I think the overall survival is still very meaningful. But, we need to consider, for these rare genotypes that we’re picking up on in comprehensive genomic profiling, a response rate as a potential surrogate to get drugs approved.