Optimal Role of PD-L1 Still Unclear in NSCLC

Tracey L. Evans, MD

As a biomarker, PD-L1 expression has proven to be a factor in selecting anti—PD-1/PD-L1 immunotherapy treatment for patients with non–small cell lung cancer (NSCLC). Yet, uncertainty remains regarding whether or not such treatments should be in used in patients who are PD-L1–negative.

OncLive: What is the current understanding of whether PD-L1—negative patients will benefit from a checkpoint inhibitor?

Using the current tests and multiple PD-1 inhibitors approved, how can oncologists decide which one is best for which patient?

Evans sat down with OncLive for an interview during the 2016 OncLive State of the Science Summit on Advanced Non—Small Cell Lung Cancer, which took place on September 17 in Philadelphia, where she shared her perspective on the role of PD-L1 testing in lung cancer. She also discussed the role of liquid biopsies and other assays that could possibly be on the horizon.Evans: We just don’t understand everything about the pathway. We also don’t understand PD-L1 staining; it’s dynamic. Patients can test positive for PD-L1 on certain biopsies and then not on other biopsies, and then there may be some heterogeneity of the tumor. It just shows that the PD-1/PD-L1 pathway is active, even in patients where the tumors don’t seem to express the marker.I don’t think we know that. The problem is that the tests used have been linked to [specific drugs]. There are different versions of PD-L1 testing; one was used with the approval of pembrolizumab (Keytruda) that requires demonstration of PD-L1 positivity in order to use the drug.

That is because of the way the studies were done. The studies were done in positive patients, and that’s where the benefit was seen. There was a marker, but a different antibody, used to test for PD-L1 in the patients on the studies of nivolumab (Opdivo). Within the squamous subset, PD-L1 positivity was not shown to be predictive of outcomes. It was predictive of outcomes in the nonsquamous subset.

In the future, will PD-L1 testing continue to have a role, or will something else become the standard?

However, even in the PD-L1—negative patients, the benefit of nivolumab was the same as for chemotherapy. You do not need to determine the PD-L1 status for patients in order to use nivolumab. It’s approved for all-comers in the second-line setting. For pembrolizumab, it is only approved for patients who demonstrate PD-L1 positivity. Whether one drug is better than the other for patients who are positive, that is not yet known.There absolutely will be a role, and that is because of the first-line studies where they have compared both nivolumab and pembrolizumab to first-line chemotherapy in patients with advanced NSCLC. The pembrolizumab trial we know was shown to be positive, and the patients in that trial had to show PD-L1 positivity of at least 50%.

What role will liquid biopsies play in the management of NSCLC?

The nivolumab study did not show an improved PFS benefit for patients treated with nivolumab over chemotherapy in the first-line setting. However, the cutoff for that, using their biomarker, was 5%. In selecting which patients are appropriate for first-line PD-1 inhibition, the PD-L1 marker will be very important. Will there be another marker coming along that might be a better predictor? I would hope so, but we don’t have one yet.Liquid biopsies come into play for a couple of reasons. Biomarker testing in patients with nonsquamous NSCLC is necessary to determine if patients have an EGFR mutation, an ALK translocation, or if they have another one of the rare mutations or translocations that is actionable. However, because of the positioning of lung tumors, it can be difficult to get enough material to run these critically important tests.

Sometimes you can find positivity for a biomarker by doing a blood-based assay where circulating tumor DNA can be shed into the bloodstream and can be analyzed by using these liquid biopsies. If a patient does not have enough material for tissue-based testing and you find one of these biomarkers on the blood-based assay, then it is appropriate to treat these patients based on the results.

This has been best demonstrated in patients who are known to be EGFR-mutant and develop resistance to a first-line EGFR TKI. It is in that setting that we know demonstration of the T790M mutation can predict for response to osimertinib (Tagrisso), so you have to biopsy these patients to show that.

Are there any ongoing biomarker-specific studies that you are interested in seeing the results of?

However, if you can show a T790M mutation on a liquid biopsy—on a plasma-based assay—then you can use that to treat patients with osimertinib. The results are similar to what we see if we show T790M on the tissue-based assay.We have only had the first-line PD-1 inhibitor studies reported by press release with their primary outcomes, so I’m absolutely very interested in learning the details of those studies. How were those patients selected? What were their response rates? Those are the ones I am most looking forward to.

“PD-L1 testing selects for an enhanced patient population who are more likely to respond to PD-1/PD-L1 inhibitors,” said Tracey L. Evans, MD, an associate professor of Clinical Medicine at the University of Pennsylvania, Abramson Cancer Center. “However, even patients who test negative for PD-L1 may still have a response, though the response rates are lower, and positivity for PD-L1 does not guarantee a response to PD-1/PD-L1 inhibitors. It’s not as straightforward of a biomarker as an EGFR mutation, for example.”