Jared Weiss, MD
The FDA approvals of the PD-1 inhibitors pembrolizumab (Keytruda) and nivolumab (Opdivo), as well as the PD-L1 inhibitor atezolizumab (Tecentriq), have delivered exciting new options in non–small cell lung cancer (NSCLC).
Despite these advances, a majority of patients with NSCLC do not respond to these agents, and there is not a reliable biomarker to ensure the right patients are receiving the treatments, says Jared Weiss, MD.
In an interview with OncLive
, Weiss, assistant professor, Thoracic Oncology Program, UNC-Chapel Hill, UNC Lineberger Comprehensive Cancer Center, offers his expert insight into the actual impact of the immunotherapy advances in lung cancer.
OncLive: What are the latest developments with immunotherapy advances in lung cancer?
Certainly, the PD-1s and PD-L1s have dramatically improved care and we’re waiting to see exactly what their total use will be, for example, into the adjuvant and neoadjuvant setting and perhaps into the stage III setting.
We’re going to see the CTLA-4 combination next. I am excited to see that for the hope that it can impact patients in the shorter realm. From there, though, there are at least a dozen targets that make a lot of sense to explore. I think it’s just going to take time to see what hits amongst those.
How reliable is PD-L1 as a biomarker?
In short, PD-L1 is a very problematic biomarker. You have 4 different assays developed by 2 different companies under 4 different sets of staining conditions, read out in 4 different ways. It’s no surprise that it's a little bit of a mess at the end of the day. There have been attempts to harmonize these. The Blueprint Project presented some preliminary data very recently looking at this. All the assays, other than SP142, seem to stain relatively similarly. SP142 did seem to stain fewer tumor cells and fewer immune cells, as well.
One thing that people have tried is to look at each test the way another test was supposed to be scored. For example, can we improve test A by also looking at immune cells as was done in test B? It turns out that doing that worked horribly. Each test works best the way it was developed and harmonization was poorer when you try to look at them in alternative ways. When you use them, they should be used as designed. Together as a group, I think that they show some positive predictive value but also fairly impractical negative predictive value, since they don't tell you who should not use these drugs.
Many ask what do we do with this in the real world? In the real world, I think if you're looking to use a therapy that has a biomarker indication you get that biomarker and you use it as indicated. What those of us in the research world should do is stop trying to torture PD-L1 into confessing something and instead move on to more useful measures.
In colorectal cancer, I think that mismatch repair seems to be a much better biomarker. mRNA expression signatures, particularly centering around interferon gamma and what I might call the ripples of gamma, seem to have legitimate promise. There is some promise for mutational load as being something of a predicator, although if we could refine that a little bit, down to mutations that lead to recognizable neoantigens, we might do yet better.
What is the “big picture” of where we are at with immunotherapy?
Immunotherapy has captured the human imagination for more than 100 years. In more recent history, high-dose IL-2 has achieved something that looks like cure for a minority of myeloma patients. With the newer therapies, such as the PD-1s and PD-L1s alone and with combinations with CTLA-4 and perhaps other agents, we're seeing more durable control than we've seen in the past. It’s really a revolution to be counting in years instead of months for some patients, but we’re still not curing the overwhelming majority of patients.
Cure or something that looks like cure is a very small minority of the patients we’re treating. What we have right now is not a true revolution or dramatic change in care. What we have is a bigger step in the evolution of cancer care than we’ve historically had in the last decade in incremental gains. I know lots of people who are alive and having fewer toxicities and thriving more for a time because of these advances, but what’s next is curing more people.
I think that if you look at the evolution of science, there are a number of agents that can be legitimately called high promise. There is hope that we might be achieving cure or very durable control without substantial toxicity for a much larger group of patients.