Practical Use of PD-L1 Testing in Stage IV NSCLC

Alexander Drilon, MD:For this case, we have a 58-year-old female who is a former smoker and who presents with metastatic lung adenocarcinoma involving the liver. She had, on imaging, a right upper lobe mass and, also, as I mentioned, evolvement of the liver parenchyma. And molecular profiling only forEGFR,ALKandROS1returned negative for an actionable driver. PD-L1 staining was performed, but it was below the 50% cutoff. She had an ECOG performance status of 1. I absolutely agree that everyone with a diagnosis of advanced non—small cell lung cancer should get PD-L1 testing upfront whenever tissue is available. We now know that based on the results of the KEYNOTE-024 trial that patients who have a PD-L1 positive D rate of 50% or greater, based on the assay used in that trial, do have improved outcomes if they receive an immune checkpoint inhibitor as a single agent—in this case, pembrolizumab—compared to platinum doublet chemotherapy.

I’ll share my experience at our institution. What we do is the pathologists have reflex testing for any sample that’s non—small cell lung cancer that’s sent to the pathology department. And, beyond doing the histologic analyses, the first pass is to do PD-L1 staining and immunize the chemistry forEGFRandALK, which are rapid tests. Then the rest of the sample is sent off for more comprehensive profiling, next-generation sequencing on our internal assay.

Our situation is unique, obviously, compared to what other people experience in the community in terms of resources. But, PD-L1 testing is something that can be done in a community setting and local pathologists are familiar with, a caveat that there are a variety of different assays depending on the drug that you’re looking at. So, it’s between atezolizumab, nivolumab, and pembrolizumab. There are different cutoffs and different ways of measuring PD-L1, also different antibodies. To be very specific, though, if you’re looking for a test for PD-L1, given the fact that pembrolizumab did gain approval in the first-line setting, if that assay is at your disposal, then I would use that assay.

I will say one last thing. There have been groups that looked at the concordance between the various assays. At least some of the assays track very nicely in terms of scoring by a pathologist, with the exception maybe of the atezolizumab assay, which is scored a little differently, and based on the inclusion of both tumor cells and immune-infiltrating cells. It’s very important to recognize the population of patients that was tested in the KEYNOTE trial. So, to be very specific, these were patients who had PD-L1 positive tumors, which based on the assay for pembrolizumab was a cutoff of 50% or greater. And, in addition, these are patients who are wild-type forEGFRorALK—those whose tumors didn’t harbor a sensitizingEGFRmutation, or a current gene rearrangement involvingALK, likeEMLforALK.

If I have a patient who meets those criteria—EGFRorALK, wild-type, and 50% or greater—I would absolutely give pembrolizumab based on the results of that trial. In patients who do not meet that criteria, either they’re positive forEGFRorALK, or if they’re wild-type, they fall below the 50%, then I would give them standard first-line therapy with a platinum doublet, with or without a biologic.

We don’t have a lot of data on what to do after first-line pembrolizumab, but the obvious answer is to default to the standard of care. For patients who have a durable response to pembrolizumab as a first-line agent, I would then follow-up after progression with platinum-doublet therapy, again, with or without an antiangiogenic agent. What to do, thereafter, we’ll figure out as the data rolls in. But, right now, that would be the algorithm I would follow.

Case Scenario 2:

SR is a 58 year-old female, former smoker 20 PPY (stopped smoking 5 years ago), presents with cough and SOB. She experienced an unintended 10 lb weight loss over a 3-month period.

• Tissue analysis identified adenocarcinoma, EGFR, ALK, ROS1 negative.
• PD-L1 expression was < 50%.
• Patient has stage IV disease with metastasis to the liver.
• A chest CT scan confirmed the primary mass in the upper lobe of the right lung, with several pleural metastases and malignant pleural effusion.
• MRI of the brain was negative for intracranial metastases.
• Patient has not complained of blood in the sputum.
• Patient has history of well controlled HTN.
• Her performance status was 1 at diagnosis.

Patient was started on carboplatin/paclitaxel/bevacizumab for 4 cycles. Patient showed partial response and was continued on bevacizumab until progression.