In NSCLC, biomarker testing rates fall short

Create PDF

Charna Albert

June 2020—Testing rates for actionable biomarkers in metastatic non-small cell lung cancer patients are below where they should be, and the overlap of PD-L1 expression with genomic targets causes confusion for oncologists and patients, said Geoffrey R. Oxnard, MD, oncologist at Dana-Farber Cancer Institute and associate professor of medicine at Harvard Medical School, in a recent CAP TODAY webinar.

He and co-presenter Lauren Ritterhouse, MD, PhD, associate director of the Center for Integrated Diagnostics, assistant pathologist at Massachusetts General Hospital, and assistant professor at Harvard Medical School, addressed how to optimize molecular testing. The webinar was made possible by a promotional sponsorship from Astra­Zeneca and is at captodayonline.com.

The 2018 CAP/IASCLC/AMP guideline advises testing for EGFR mutations and ALK and ROS1 rearrangements. The 2020 NCCN guidelines recommend, in addition, testing for BRAF mutations and PD-L1 expression. Both of these guidelines say broad sequencing panels are important in capturing a wider range of targetable variants, to include MET, RET, ERBB2, and KRAS, Dr. Oxnard said. Both also say plasma-based testing is an emerging alternative.

“For efficiency in time of testing and in use of tissue, in our practice we’re now using tumor NGS with an in-house panel as our go-to approach” for results on the five key variants and the emerging variants.

Despite the guideline recommendations and the benefits of targeted therapy, Dr. Oxnard said, “current testing rates for actionable biomarkers are just inadequate.”

He reported data presented last year (Gierman H, et al. ASCO annual meeting, 2019, Abs 1585) showing that patients with NSCLC continue to be undergenotyped for the NCCN-recommended genes. The percentage of NSCLC patients tested for EGFR was found to be 54 percent, ALK 51 percent, ROS1 43 percent, BRAF 29 percent, RET 17 percent, MET 15 percent, and ERBB2 11 percent. In the same study, the rate of testing for PD-L1 expression was 48 percent.

Twenty-two percent of patients were tested for all four genes with FDA-approved on-label drugs. Seven percent were tested for all seven genes with associated therapies included in NCCN guidelines.

Other studies have found that for genes with associated targeted therapies, the rate of testing for EGFR was the highest, at 72 percent, followed by the rates of testing for ALK (69 percent) and ROS1 (38 percent) (Kim ES, et al. J Thorac Oncol. 2019;14[3]:338–342). But the rate of comprehensive testing for four major types of alterations—NCCN-recommended and emerging biomarkers at time of study—was eight percent.

Dr. Oxnard

“One of the things we’re struggling with is how we can improve these testing rates,” Dr. Oxnard said, to increase the likelihood that patients will be connected to the oral targeted therapies or, in the case of high PD-L1 expression, to immunotherapies.

But another problem: “These two paths straight overlap,” confusing oncologists and patients, he said. Several studies have shown that up to 70 percent of patients who are EGFR mutation positive also express at least one percent PD-L1. (For ROS1, it’s 100 percent overlap, for ERBB2 it’s 53 percent, for MET and RET it’s 75 percent. Mazieres J, et al. ASCO annual meeting, 2019, Abs 9010.) How to proceed? In the same ASCO presentation by Gierman H, et al., 37 percent of patients with known actionable biomarkers such as EGFR and ALK and no evidence of progression on targeted therapy (n = 84) were reported to have received immune checkpoint inhibitors. For 65 percent of the 37 percent, the test result was available prior to checkpoint inhibitor initiation.

Yet the molecularly targeted subsets of lung cancer do poorly with immunotherapy despite PD-L1 expression. So he sees the role of tumor genotyping as twofold: “It tells which patients to steer toward an oral targeted therapy, but it also tells us which patients are less likely to benefit from immunotherapies.”

“PD-L1 expression can be seen in EGFR, ALK, and RET,” he said, “and yet it’s somewhat of a false-positive in those instances because it’s not then associated with immunotherapy sensitivity.” Eight of the nine most recognized first-line trials of immunotherapy excluded patients with treatment-naive metastatic EGFR-mutation-positive NSCLC, so all of the FDA-approved labels for first-line immunotherapy, as a single agent or in combination, exclude those patients. Thus, the recommended molecular testing is essential, he said, “to steer them toward an effective oral therapy and away from an ineffective immunotherapy that may add toxicity.”

There may be a role for immunotherapy in genotype-defined molecular subtypes of lung cancer, but “it’s not a first-line therapy,” he said. “It’s a later-line approach” when the treatment options are fewer. It’s targeted therapy, he said, not immunotherapy, that improves overall survival rates for these patients, according to the evidence.

He presented a case seen commonly in the clinic. “I see a 48-year-old male who is a former light smoker and presents with lung adenocarcinoma, stage IV NSCLC metastatic to the lungs.” The patient is eager to begin an effective therapy. “When I see him, his pathology is signed out, and testing for PD-L1 has already been completed” (greater than 90 percent positive). “What I struggle with is, do I dive in and treat this patient with immunotherapy, or do I hold off and wait for next-generation sequencing results, which can help clarify the role of PD-L1 in his care?”