David R. Gandara, M.D., discusses the latest immunotherapy advances in lung cancer, as well as the next steps toward optimizing the use of these agents.
David R. Gandara, M.D.
While immunotherapy agents continue to move into the treatment landscape for patients with non-small cell lung cancer (NSCLC), there are still questions that are left to be answered, says David R. Gandara, M.D.
In an interview with CURE, Gandara, director of the Thoracic Oncology Program, professor at UC Davis Comprehensive Cancer Center, and a 2017 Giant of Cancer Care in Lung Cancer, discussed the latest immunotherapy developments in lung cancer and the next steps toward optimizing the benefit of these agents.The results of the MYSTIC trial have been much anticipated. This is a phase 3 trial in advanced-stage NSCLC, randomizing patients to standard-of-care platinum chemotherapy versus either Imfinzi (durvalumab) alone or Imfinzi plus tremelimumab, a CTLA-4 inhibitor.
In a press release from AstraZeneca, it was reported that the primary endpoint of progression-free survival (PFS) was not met in either of the experimental arms. This is only a press release, so no data were reported in terms of the magnitude of benefit; it was just reported that the primary endpoint was not met.
This could still play out in multiple ways. The study will not be presented until the 2017 ESMO Annual Congress in early September. It is possible that, with further follow-up, there could be improved overall survival even though the PFS was not improved. It could also be that the magnitude of benefit is different between the two experimental arms, or that the degree of PD-L1 expression, which was at the 1 percent or greater level, could differentiate patients who benefitted versus those who did not. I want to hear the complete data set from this trial. We have to reconcile it against the other studies.
If we put MYSTIC together with the negative results of the CheckMate-026 Opdivo (nivolumab) trial, published by Dr David Carbone and colleagues in the New England Journal of Medicine, it means that trials of checkpoint immunotherapy are not always going to results in a big advance. It’s not guaranteed that it’s going to hit a home run in every setting, and we’re still very naïve in our ability to design trials that will carve out the patients who will benefit the most.
Those negative results also emphasize how impressive the data are from the other first-line trial, the KEYNOTE-021 Keytruda (pembrolizumab) trial with a similar design in advanced-stage NSCLC, which demonstrated improved PFS with patients with a high PD-L1 expression of 50 percent or greater compared with platinum-based chemotherapy. Even though crossover was built into this study, it also showed improved overall survival (OS). This means that the patients on the chemotherapy arm who crossed over to Keytruda and had a benefit did not interfere with the ability to show improved overall OS.
Lastly, the recent FDA accelerated approval of Keytruda plus platinum-based chemotherapy in all comers without any PD-L1 expression requirement, is in contrast to all of these studies. In view two out of three first line immunotherapy trials being negative in advanced NSCLC, I am concerned that the accelerated approval may not pan out in the phase 3 trial in that setting, which is the KEYNOTE-189 study. The data on whether to test for PD-L1 and who to treat based on the results are confusing right now. Both patients and practicing oncologists are going to be confused until the remainder of these studies are reported. How to design clinical trials is an ongoing debate. At one point, 10 years ago, I analyzed the last 24 randomized clinical trials that were done trying to add a targeted agent of many drug classes to chemotherapy versus chemotherapy alone. Of those trials, only two were positive in unselected patient populations. The first was the Avastin (bevacizumab) trial from ECOG and the second trial was the partially biomarker-driven FLEX trial of chemotherapy with or without Erbitux (cetuximab).
If you have a targeted drug, it is important to define the population who is most likely to benefit and perhaps design a biomarker-strategy that incorporates only those patients. Therefore, the drug and its companion diagnostic are more likely to have a positive result.
That being said, many of the trials being done today want to throw the drug at every patient with NSCLC, regardless of biomarker status. My own impression is that the checkpoint immunotherapeutics are targeted drugs. Just like for EGFR mutation or ALK rearrangement, we have to find the appropriate biomarkers.
Right now, it’s PD-L1. In the future, it may be an immune signature, tumor mutational burden, or a combination of things; however, until we find better biomarkers for these drugs, we will likely see some positive trials and negative trials without a rhyme or reason why they are so.
For example, Opdivo is a very similar drug to Keytruda, so why would one of the first-line trials be positive and the other negative? Are there differences in patient populations, the biomarker cutoff that was used for the trials, or even the assay that was used? We don’t know for sure.PD-L1 as a biomarker has had ups and downs. There is an old analogy about a new drug saying that when it first comes out, everyone thinks it’s fantastic but, after a while, it loses its luster. Two years ago, most people were turned off by PD-L1. We have swung full circle with data of Keytruda as a first-line therapy, showing not only an increase in PFS but also OS, despite crossover. This has led to where PD-L1 tested with the Dako IHC 22C3 assay has become the standard of care for first line therapy. Now, with some of the data coming in with chemotherapy plus Keytruda, or even the MYSTIC trial, we are still grasping for what is the best biomarker for the checkpoint immunotherapeutics.My own opinion is that single-agent Keytruda [is appropriate] for patients with PD-L1 expression greater than 50 percent—it’s important to define that population. It means that, in the United States, there is a reason to test every patient. Also, if they’re going to be tested, it should be done with the Dako IHC 22C3 assay. If the results are greater than 50 percent in non-squamous cell stage 4 disease then, in my opinion, they should receive Keytruda.
There are many caveats. As I mentioned earlier, the recent accelerated approval by the FDA of Keytruda with platinum-based chemotherapy with no requirement for PD-L1 expression has left us in a quandary. It’s difficult. You can imagine a patient could be told, “We’re going to test your tumor for PD-L1. If it’s greater than 50 percent, we’re going to give you a drug called Keytruda. If it’s not, we’re still going to give you Keytruda but also throw chemotherapy together with it.” Hopefully, as more results come in we will reach some clarity on these points.