Patient Selection Vital in Ensuring Improved Response to PD-1, PD-L1 Inhibitors in NSCLC
A LATE AFTERNOON
extended education session on day 1 of the 2017 American Society of Clinical Oncology Annual Meeting in Chicago became a discussion on state-of-the-art uses for immunotherapy in the management of non–small cell lung cancer (NSCLC). Oncologists shared their experiences with managing toxicities from immunotherapy and discussed the role of immunotherapy in specific patient populations.
Edward B. Garon, MD, director of the Thoracic Oncology Program and associate professor of medicine, David Geffen School of Medicine at University of California Los Angeles, spoke about using checkpoint inhibitors, primarily programmed death-1 (PD-1) inhibitors, in first-line therapy and sequencing these agents.
Garon discussed results from Keynote-024,1
a phase 3 randomized controlled trial that compared pembrolizumab as frontline therapy with platinum-based chemotherapy in patients diagnosed with programmed death ligand-1 (PD-L1)–expressing advanced NSCLC (50% PD-L1 expression threshold). The primary study endpoint was progression-free survival (PFS). Despite patients who crossed over from pembrolizumab to chemotherapy, overall survival (OS) improved.
which had a similar study design, compared first-line single-agent nivolumab versus chemotherapy. Crossover was allowed, and PFS was the primary endpoint. The main difference was that the PD-L1 expression cut-off was set at 5%, which encompasses a much broader patient population. Nivolumab, however, failed in the frontline setting compared with chemotherapy. “OS was not different in the 2 arms,” Garon said.
He lined up a series of differences to explain the differential results:
Failure of the randomized trial mechanism. The data showed that women were less likely to be on the nivolumab arm, as were patients with more than 50% PD-L1.
Difference in efficacy
Difference in the patient cohort:
a. The most common clinical difference was the difference in radiotherapy. The Checkmate trial, Garon said, used a higher dose of radiation. “However, a single-institution study has shown that patients who may have received prior radiation may have done better.”
b. Use of PD-L1 expression. Selection of 50% cutoff for PD-L1 expression was used to select patients for pembrolizumab, as opposed to 5% for nivolumab.
Garon also highlighted a tumor’s mutation burden: “[The] higher the mutation burden, [the] greater the benefit,” he said. The take-home messages from his presentation were:
Patients with staining in at least 50% of their tumor cells should be eligible for frontline pembrolizumab.
Those with staining in less than half of their tumor cells should receive standard chemotherapy as frontline treatment.
Adding nonselected therapy for each group remains a topic of debate
Melissa Lynne Johnson, MD, associate director of lung cancer research, Sarah Cannon Research Institute, addressed the conundrum of choosing a suitable immunotherapy agent. She explained that PD-1 and PD-L1 inhibitors are monoclonal antibodies. The interesting thing is that different PD-1 inhibitors bind different faces of the PD-L1 protein. “They also block different protein-protein interactions,” which might result in differences in patient responses based on which drug is administered. Variability also arises from immunoglobulin G isotypes and antibody-dependent cell-mediated cytotoxicity (ADCC). “Avelumab is the only immunotherapy drug that has retained its ADCC function compared with nivolumab, durvalumab, pembrolizumab, and atezolizumab,” Johnson said.
The PD-L1 assay used to assess protein expression also adds to the variability. Patient adverse events (AEs) associated with the treatment have varied, Johnson said: they have hovered around 76% for nivolumab, much higher than pembrolizumab (67.5%), atezolizumab (65%), durvalumab (60.6%), and avelumab (67%).
Compared with docetaxel for nonsquamous NSCLC, survival data in the second-line setting show that the median OS is 12.2 months for nivolumab, 10.4 months with pembrolizumab compared with 8.5 months, and 13.8 months for atezolizumab compared with 9.6 months. In the first-line setting, however, nivolumab has lagged behind pembrolizumab, Johnson showed. The median PFS for nivolumab is 4.2 months and 5.9 months for chemotherapy. Pembrolizumab has a median PFS of 10.3 months. The median OS for nivolumab is 14.4 months, and it has not yet been reached for pembrolizumab.
Johnson emphasized that cost versus convenience is another question that both physicians and patients are concerned with. Nivolumab is administered every 2 weeks (both 240 and 3 mg/kg doses) and costs about $21,990 for a 6-week treatment, while pembrolizumab is administered once in 3 weeks (both 240 and 2 mg/kg doses) and costs about $21,662 over 6 weeks. She proposed evaluating these agents over a long treatment interval to lower costs and the inconvenience to patients of frequent administration.
Johnson summarized her presentation by saying that in the first-line setting, understanding the role of the tumor microenvironment might help understand the differences in patient response, as will identifying additional biomarkers. “Until then, [the] dosing schedule and cost will continue to play a significant part in oncologists’ decision making.”