The Role of Molecular Testing in Lung Cancer
David Spigel, MD: When we look at if there certain groups of patients who benefit over others, it really doesn’t pan out. In other words, everybody who has unresectable stage III cancer, whether it’s adeno or squamous—and I’ll get into this in a little more detail, but even if there are molecular findings, such as EGFR mutations or PD-L1, we use PD-L1 expression—across the board everybody seemed to benefit. There’s not a gender or an age or, as I mentioned, a histology subset that does better or worse.
Let’s talk about EGFR in PD-L1 because this has got the most potential from PACIFIC recently since approval. Let’s talk about EGFR first. To get on the study, it wasn’t a requirement to know the EGFR status of patients, so it wasn’t part of the primary design to say, “Well, patients are stratified for EGFR mutation status.” As it turns out, there were very few patients with known EGFR positivity. In fact, the numbers end up being about 24 vs 43 patients in terms of placebo, durvalumab exposed, placebo exposed when you look at the overall survival [OS] analysis.
Those patients who were EGFR positive, in that limited sample, if you look at the wide confidence interval, it crosses 1 on the forest plot. What we know about that is that we don’t have enough information to say whether a patient benefits from durvalumab or not. Actually, if you look at the statistical analysis of the forest plot, the median of that confidence interval was to the left of 1, suggesting benefit for durvalumab. But because it crosses 1, all we can say is the confidence interval is wide and we don’t have enough patients. Now, because immunotherapy has not proven itself in EGFR-mutated patients, some controversy has come up about whether you should use durvalumab in patients with EGFR mutations. And this is still a healthy debate.
Personally, I look at this trial as an intent-to-treat study. In other words, we treat all patients who went on this study as we would our patients we’d see in the community. Because EGFR testing was not required for this trial, I don’t think it’s required in standard-of-care treatment. For my patients who have unresectable stage III, I don’t routinely test for EGFR. But if I did, if I did have that result, I still think durvalumab is the treatment for them because it’s proven to offer a survival advantage. And we don’t know with certainty that patients with EGFR mutations would not benefit. I don’t want to withhold an opportunity for those patients.
Now, for PD-L1 expression, the same can really be said. PD-L1 expression was not a requirement for getting on the study. It was a post hoc analysis required by European regulatory authorities. In fact, in Europe, durvalumab cannot be used unless there is PD-L1 expression. That was probably an overreach of regulatory authorities because the study was not designed for that cutoff to be used in a primary analysis.
Here’s what we know. If you look at patients who have greater than 1% PD-L1 expression, in the forest plot the advantage for both PFS [progression-free survival] and OS is in favor of durvalumab. We say anybody who has expression on that analysis seem to benefit for durvalumab. If you look at the patients who are 0—and there were about 225 patients in total who would be scored as 0—the confidence interval crosses 1. If you look at the median of that confidence interval, it’s to the right of 1, suggesting that patients with no expression of PD-L1 do not benefit from durvalumab. That’s why European regulatory authorities do not allow durvalumab for those patients.
However, if you look at progression-free survival, the opposite could be said. That is, it crosses 1, but the benefit appears to be in favor of the 0s. This is like EGFR: We simply don’t know. With all the problems of PD-L1 testing, the heterogeneity and the tumor sample of testing and the tumor sample you have, and the uncertainties about this post hoc analysis, when you have a therapy that substantially affects 4-year overall survival, progression-free survival, I’m not going to deny a patient access to that because I have an IHC [immunohistochemistry] in front of me that says their PD-L1 is 0. I just don’t think that’s fair, and I don’t think there’s enough information there to deny a patient a chance of...
To that end, I think durvalumab is the standard of care for patients with unresectable stage III cancer after chemoradiation. It’s given for 1 year. Is that enough? Is that too much? We don’t know. That’s just how that study was designed.
Transcript edited for clarity.
