Drugs targeting the immune-checkpoint pathways have shown promising activity in non–small cell lung cancer (NSCLC). In a recent article in The New England Journal of Medicine, Garon and colleagues reported the results of the KEYNOTE-001 clinical trial evaluating single-agent pembrolizumab (Keytruda) in patients with advanced NSCLC1; the study is reviewed in this issue of The ASCO Post.
Pembrolizumab is a humanized monoclonal IgG4 antibody directed against programmed cell death protein 1 (PD-1). PD-1 is a co-inhibitory receptor on activated T cells that can promote T-cell exhaustion on binding to its ligand PD-L1, thus inhibiting the cytotoxic T-cell response.2 PD-L1 expression is a common mechanism of immune evasion in NSCLC.3
A Closer Look at KEYNOTE-001
KEYNOTE-001 is a large international multicenter phase I trial evaluating the safety and efficacy of pembrolizumab in patients with advanced melanoma and NSCLC. The lung cohort of the trial enrolled a total of 495 patients with advanced NSCLC; among these patients, 384 had received prior therapies, and 101 patients were treatment-naive.
The adverse effect profile of pembrolizumab in the trial was favorable, with less than 10% of patients experiencing grade 3 or higher toxicities. The most common immune-related adverse event was hypothyroidism (6.9%, any grade), which was reported to be easily managed with thyroid supplementation. Pneumonitis was reported in 18 of the 495 patients (3.6%), and half of them were reported to be grade 3 or higher (including one death, 0.2%).
The overall response rate for the study was 19.4% in the total of 495 patients evaluated (18% in 394 previously treated patients and 24.8% in 101 untreated patients). The median duration of response was 12.5 months in the overall population and nearly 2 years in previously untreated patients.
The trial evaluated the role and the optimal cut point of PD-L1 expression in predicting response to pembrolizumab. Of the 495 patients, 182 patients were enrolled in the training cohort, and 313 patients were enrolled in the validation cohort. The PD-L1 diagnostic assay employed was Dako EnVision FLEX+/HRP polymer kit, using the antibody clone 22C3 to detect PD-L1.
Using receiver-operating-characteristic curve analysis in the training cohort of 182 patients, the PD-L1 cutoff was determined to be membranous PD-L1 expression in at least 50% of tumor cells. Using this cut point, nearly one-quarter of the patients evaluated tested positive for PD-L1 expression.
The objective response rate in PD-L1–positive patients was 45.2%, and the median overall survival was not reached. Response rates were 16.5% in patients who had low expression of PD-L1 (1%–49%) and 10.7% in those who were PD-L1–negative (< 1%).
The results of the KEYNOTE-001 trial further our understanding of and enthusiasm for this exciting class of drugs for the treatment of lung cancer. The most consistent finding with these agents is the durability of responses. This was again demonstrated in this trial, with the median duration of response being over a year in patients responding to pembrolizumab (84.4% of the responders had no disease progression at the time of the data cutoff for the publication).
Potential Predictive Biomarker
Early clinical studies with PD-1/PD-L1 inhibitors suggested PD-L1 expression by immunohistochemistry may predict response to these agents.4-7 However, clinical trials have measured tumor PD-L1 expression using immunohistochemistry with different antibodies, including Dako clone 28-8 (nivolumab [Opdivo]), Spring Bioscience clone SP142 (MPDL3280A), Spring Bioscience clone SP262 (MEDI4736), and, in the current study with pembrolizumab, Dako clone 22C3.
In addition, trials reported thus far had variable definitions of positivity for PD-L1; some assays use tumor expression, whereas others use PD-L1 staining in the stromal immune infiltrates. The percentages of positive cells required to consider the case as positive or negative are also dissimilar.
Moreover, there also appears to be inherent spatial and temporal heterogeneity of PD-L1 expression as a function of interaction and adaptation to the dynamic tumor immune microenvironment. Hence, the reproducibility of PD-L1 assays and their potential as a predictive biomarker have been questioned.
The biomarker evaluation in KEYNOTE-001 trial was well planned and had comprehensive validation in an independent cohort. The findings from this trial suggest the predictive value of PD-L1 expression, with nearly 50% response rates in PD-L1–positive (> 50% tumor cells) patients. However, in patients who were defined as PD-L1–low or –negative (< 50% tumor cells), the responses rates were still higher than expected with cytotoxic chemotherapy (15.2% in the validation cohort overall). These results will be further explored in an ongoing prospective randomized phase III trial of docetaxel vs pembrolizumab (KEYNOTE-10).
Closing Comments
The KEYNOTE-001 trial provides valuable information regarding the role of PD-L1 expression and the efficacy of pembrolizumab in both refractory and treatment-naive patients with NSCLC. The PD-L1 assay used in the trial appears to enrich for patients who could potentially respond to pembrolizumab. However, the drug also has activity in low–PD-L1 patients and is comparable (and perhaps superior) to cytotoxic chemotherapy in NSCLC, especially in the refractory setting.
Thus, the utility of the PD-L1 assay in NSCLC patients, particularly in the second line and beyond, remains uncertain. However, PD-L1 expression in the future could potentially be used to stratify patients in clinical trials and inform decisions regarding combinatorial strategies with other immunotherapies, chemotherapy, or radiation therapy. ■
Disclosure: Dr. Velcheti reported no potential conflicts of interest. Dr. Herbst has been a consultant for Merck.
References
1. Garon EB, Rizvi NA, Hui R, et al: Pembrolizumab for the treatment of non-small-cell lung cancer. N Engl J Med 372:2018-2028, 2015.
2. Schalper KA, Venur VA, Velcheti V: Programmed death-1/programmed death-1 ligand axis as a therapeutic target in oncology: Current insights. J Receptor Ligand Channel Res 8:1-7, 2015.
3. Velcheti V, Schalper KA, Carvajal DE, et al: Programmed death ligand-1 expression in non-small cell lung cancer. Lab Invest 94:107-116, 2014.
4. Taube JM, Klein A, Brahmer JR, et al: Association of PD-1, PD-1 ligands, and other features of the tumor immune microenvironment with response to anti-PD-1 therapy. Clin Cancer Res 20:5064-5074, 2014.
5. Topalian SL, Sznol M, McDermott DF, et al: Survival, durable tumor remission, and long-term safety in patients with advanced melanoma receiving nivolumab. J Clin Oncol 32:1020-1030, 2014.
6. Herbst RS, Gordon MS, Fine GD, et al: A study of MPDL3280A, an engineered PD-L1 antibody in patients with locally advanced or metastatic tumors. 2013 ASCO Annual Meeting. Abstract 3000.
7. Herbst RS, Soria JC, Kowanetz M, et al: Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature 515:563-567, 2014.
Dr. Velcheti is Assistant Professor, Department of Hematology and Oncology, Cleveland Clinic. Dr. Herbst is Ensign Professor of Medicine, Professor of Pharmacology, Chief of Medical Oncology, Director, Thoracic Oncology Research Program, Associate Director for Translational Research, Yale Comprehensive Cancer Center, Yale School of Medicine.
