Targeting PD-1/PD-L1 in Non-small Cell Lung Cancer: An Update from ESMO 2016

Scientific Reviewer: Frances A. Shepherd, MD, FRCPC

Frances A. Shepherd, MD, FRCPC
Professor, Department of Medicine,
University of Toronto
Scott Taylor Chair in Lung Cancer Research,
Princess Margaret Cancer Centre



The clinical-trial experience with immuno-oncology agents has heralded a new era of hope for patients with non-small cell lung cancer (NSCLC).1 The most compelling evidence has been demonstrated with agents that inhibit the PD-1 (nivolumab, pembrolizumab) and PD-L1 (atezolizumab) pathways. Used as monotherapy, these agents demonstrated encouraging response rates and meaningful survival times (in terms of both progression-free survival [PFS] and overall survival [OS]), in chemotherapy-naïve and, even more importantly, in heavily treated patients in the early phase 1 and 2 trials.3-12 These encouraging results led rapidly to the initiation of randomized trials comparing nivolumab, pembrolizumab and atezolizumab to single-agent docetaxel after platinum-based chemotherapy. All studies reported superior response and survival outcomes for patients in the immunotherapy arms.10-12

Given the encouraging results in the phase 3 second-line studies, and in earlier-phase trials in first-line use, patients and clinicians alike have been awaiting the results of phase 3 studies of these agents used in first-line. At the 2016 congress of the European Society for Medical Oncology (ESMO), such results were presented for the first time, with efficacy and safety data made available for pembrolizumab (KEYNOTE-02413,15) and nivolumab (CHECKMATE-02614). These trial results also shed further light on the utility of PD-L1 testing, which has been the subject of intense scrutiny during the development and evaluation of these agents.16

The primary goal of this report is to provide a summary of the clinical trial data presented at ESMO 2016, and an interpretation of how these may impact the care of our patients with NSCLC.

Additionally, the evidence base for PD-1 blockade for first-line treatment of NSCLC was further strengthened at ESMO 2016 by the presentation of phase 2 data with pembrolizumab used in combination with chemotherapy (KEYNOTE-021,17-18). Important data in previously treated patients also were presented for the PD-L1 inhibitor, atezolizumab (efficacy and safety results of the phase 3 OAK study19) and for pembrolizumab (updated survival results and quality-of-life data from the phase 2/3 KEYNOTE-010 study20,21). These data also are summarized briefly below.

PD-1 Inhibitors for First-line Treatment of NSCLC: KEYNOTE-024 and CHECKMATE-026

KEYNOTE-024 was an open-label, phase 3 trial involving 305 patients with previously untreated advanced NSCLC and PD-L1 expression on ≥ 50% of tumor cells.13-15 Subjects were randomized to receive either pembrolizumab 200 mg every three weeks or the investigator’s choice of platinum-based chemotherapy. Individuals in the chemotherapy group were allowed to cross over to the pembrolizumab group in the event of disease progression. The primary endpoint was PFS; key secondary endpoints included OS, objective response rate (ORR), and safety.

As shown in Figure 1, the median PFS was 10.3 months for pembrolizumab (95% CI 6.7 to [not reached]) and 6.0 months (95% CI 4.2 to 6.2) for chemotherapy. The relative risk reduction was 50% (HR 0.50; 95% CI 0.37 to 0.68; p < 0.001).Figure01

Despite crossover, OS also was significantly greater with pembrolizumab vs. chemotherapy (Figure 1). While median survival was not reached in either group, the six-month OS rate was 80.2% vs. 72.4% in favor of pembrolizumab (HR for death 0.60; 95% CI 0.41 to 0.89; p = 0.005). The ORR also was significantly higher with pembrolizumab (44.8%; 95% CI 36.8 to 53.0) vs. chemotherapy (27.8%; 95% CI 20.8 to 35.7). The median duration of response was not reached in the pembrolizumab group and was 6.3 months in the chemotherapy group.

There were no new safety signals detected in this study. Overall treatment-related adverse events (AEs) and treatment-related serious AEs were less frequent in the pembrolizumab group than in the chemotherapy group. Immune-related events were as expected, and for the most part easily managed.

KEYNOTE-024 represents a paradigm shift in the management of advanced or recurrent NSCLC. The win-win combination of superior efficacy and reduced toxicity make it clear that, in patients whose tumors demonstrate high expression of PD-L1, first-line pembrolizumab should certainly be considered as an option. Indeed, on October 24, 2016, the U.S. FDA approved pembrolizumab for treatment of patients with metastatic NSCLC whose tumors have high PD-L1 expression (≥ 50%) with no EGFR or ALK genomic tumor aberrations and no prior systemic chemotherapy treatment for metastatic NSCLC. It should be noted, though, that this high-expression subgroup represents a minority of patients with NSCLC. Among almost 2,000 patients screened, only 300 of 1,653 samples adequate for testing were positive (30%). A 50% cut-point for expression was used in this trial, but different cut-points to define positivity, different scoring systems and different antibodies for IHC testing have been used for each drug under development. Is 45% expression really different from 50% expression? Exactly what level of expression will be associated with statistically significant, clinically meaningful, and cost-effective survival benefits remains to be defined. Sadly, not all studies have been powered adequately to examine cut-point intervals by 10% increments or even quartiles. Also, because of different scoring techniques, different antibody avidity and proprietary data, pooled individual-patient-data analyses may not even be possible in the future.

CHECKMATE-026 was an open-label, phase 3 study of 541 patient with advanced NSCLC with no prior systemic therapy for advanced disease.14 The inclusion criteria with respect to PD-L1 expression was ≥ 1%. Subjects were randomized to nivolumab 3 mg/kg IV every two weeks or investigator's choice of platinum-based doublet chemotherapy. Crossover was allowed for patients in the chemotherapy group who experienced disease progression.

The primary endpoint of the study was PFS among patients with ≥ 5% PD-L1 expression (n = 423). Secondary endpoints included OS and ORR in this subset, as well as PFS in the entire study population of patients with ≥ 1% PD-L1 positivity. Subgroup analyses also included PFS and OS among patients with ≥ 50% PD-L1 positivity (n = 214).

For the primary endpoint, there was no significant difference between nivolumab (n = 211) and chemotherapy (n = 212) in the ≥ 5% PD-L1 expression cohort (median PFS 4.2 months for nivolumab [95% CI 3.0 to 5.6] and 5.9 months for chemotherapy [95% CI 5.4 to 6.9]; HR 1.15 [95% CI 0.91 to 1.45; p = 0.2511]) (Figure 2). One-year PFS rates were 23.6% for nivolumab and 23.2% for chemotherapy. A similar trend was seen in the entire trial cohort (PD-L1 expression ≥ 1%): HR for PFS 1.17 (95% CI 0.95 to 1.43). Among the subgroup with ≥ 50% PD-L1 expression (n = 88 for nivolumab and n = 126 for chemotherapy), the unstratified HR for PFS was 1.07 (p = ns).Figure02

The median OS was 14.4 months for nivolumab (95% CI 11.7 to 17.4) and 13.2 months for chemotherapy (95% CI 10.7 to 17.1); HR 1.02 (95% CI 0.80 to 1.30; Figure 2). In the subgroup with ≥ 50% PD-L1 expression, the unstratified HR for OS was 0.90 (p = ns).

The ORR was 26.1% for nivolumab (95% CI 20.3 to 32.5) and 33.5% for chemotherapy (95% CI 27.2 to 40.3).

The AE and safety profile of nivolumab in this study was consistent with previous reports, with no new safety signals and with fewer treatment-related AEs and serious AEs in the nivolumab group compared with the chemotherapy group.

How shocking and totally unexpected these results were! Perhaps the greatest surprise was the lack of benefit from nivolumab, even in the high PD-L1 expression subgroup, and only modest benefit in squamous cancers. Although the CHECKMATE-017 randomized trial comparing nivolumab to docetaxel demonstrated no effect of PD-L1 expression in patients with squamous cell lung cancer, the CHECKMATE-057 trial in non-squamous cancers showed definite trends for increasing benefit with increasing PD-L1 expression, similar to the trials of pembrolizumab and atezolizumab. Unfortunately, this trial was not powered adequately to delve any deeper into expression subgroups and adjust for other potentially important prognostic factors.

Other PD-1/PD-L1 Studies of Interest at ESMO 2016

KEYNOTE-021 was a phase 2 study designed to determine whether adding pembrolizumab to platinum-doublet chemotherapy improved efficacy among 123 patients with chemotherapy-naïve advanced non-squamous NSCLC.17,18 Subjects were randomized to receive carboplatin/pemetrexed chemotherapy plus pembrolizumab 200 mg every three weeks, or the same chemotherapy alone. Baseline PD-L1 expression was recorded as < 1%, 1-49%, or ≥ 50%. The primary endpoint was the proportion of patients achieving an objective response.

As shown in Figure 3, the proportion of patients with an objective response was 55% for the pembrolizumab-chemotherapy combination and 29% for chemotherapy alone. There were no complete responses in either group. Response rates were similar across the PD-L1 expression groups in the chemotherapy-alone arm. Objective response for the combination also was similar when comparing the no PD-L1 expression to any expression subgroups: 57% among patients with < 1% expression and 54% among those with ≥ 1% expression. However, among those with ≥ 50% PD-L1 expression, ORR was 80% (16/20 patients) for the combination and 35% (6/17 patients) for chemotherapy alone.Figure03

Overall, the six-month PFS was 77% for the combination and 63% for chemotherapy alone (HR 0.53; 95% CI 0.31 to 0.9; p = 0.0102). Median PFS was 13.0 months for the combination and 8.9 months for chemotherapy alone. There were no differences detected in terms of OS (HR 0.90; 95% CI 0.42 to 1.91).

The incidence of serious AEs was higher in the combination arm compared to chemotherapy alone. In the combination group, the rates of grade 3, 4 and 5 treatment-related AEs were 31%, 7% and 2%, respectively. For chemotherapy alone, grade 3, 4 and 5 treatment-related AEs were observed in 19%, 3% and 3%, respectively. The most common AEs (experienced by ≥ 15% of patients in either group) are shown in Figure 4.Figure04

The efficacy and safety of pembrolizumab-chemotherapy combination in first-line use is being explored further in the phase 3 KEYNOTE-189 study.22

This small study lends further support to the concept of patient selection for PD-1 monoclonal antibody therapy, and suggests patients with high PD-L1 expression have the greatest potential to derive benefit. Response rates in patients with 1%-49% expression were not superior to chemotherapy alone; in contrast, the ORR of 80% in the high-expression subgroup is virtually unprecedented in advanced NSCLC! However, only 20 patients were in this subgroup. Therefore, this small phase 2 trial must not be considered definitive, and in view of the increased toxicity, and increased cost, the results of large phase 3 trials must be awaited before this combination can become a new standard of care.

KEYNOTE-010. The primary efficacy and safety data for KEYNOTE-010 were published in The Lancet in early 2016.23 At ESMO 2016, updated survival data were reported for a median follow-up of 19.2 months.20 Median overall survival for previously treated patients with ≥ 1% PD-L1 expression was 8.6 months for docetaxel and 10.5 months for pembrolizumab 2 mg/kg (HR 0.72; 95% CI 0.60 to 0.87). Among those with ≥ 50% PD-L1 expression, median overall survival was 8.2 months and 15.8 months for docetaxel and pembrolizumab 2 mg/kg, respectively (HR 0.54; 95% CI 0.39 to 0.73). Eighteen-month overall survival rates were 37% with pembrolizumab 2 mg/kg and 24% with docetaxel in the ≥ 1% PD-L1 expressing analysis and 46% vs. 24%, respectively, among those with ≥ 50% expression.

The KEYNOTE-010 investigators also presented data on quality-of-life parameters from baseline to week 12 in this trial.21 They reported that there was either numeric or significant improvement in the EORTC QLQ-C30 global health status/QoL score for pembrolizumab compared with docetaxel; however there was prolongation in the time to deterioration in the EORTC QLQ-LC13 symptom endpoints of cough, dyspnea, and chest pain.

PD-L1 monoclonal antibody inhibitors now have a clearly defined role in previously treated patients with NSCLC. These updated results suggest that, in terms of efficacy outcomes, symptom control and QoL they are superior to docetaxel, and should become the new standard of care after platinum-based chemotherapy. However, this population of patients may become considerably smaller once approval for pembrolizumab in the first-line is approved globally.

The OAK study enrolled 1,225 patients with advanced NSCLC with any PD-L1 status who had received one or two prior lines of chemotherapy.19 Subjects were randomized to receive atezolizumab 1,200 mg IV or docetaxel every three weeks. The primary endpoints were OS in the overall population and OS among patients with PD-L1 expression in ≥ 1% of tumor cells or tumor-infiltrating immune cells. Preliminary results were presented at ESMO for the first 850 patients enrolled in the study.

Overall, atezolizumab was associated with a significantly better OS compared to docetaxel (Figure 5): median 13.8 months (95% CI 11.8 to 15.7) vs. 9.6 months (95% CI 8.6 to 11.2); HR 0.73 (95% CI 0.62 to 0.87; p = 0.0003). Twelve- and 18-month OS were 55% and 40% for atezolizumab and 41% and 27% for docetaxel, respectively. Among subjects with ≥ 1% PD-L1 expression on tumor or immune cells, median OS was 15.7 months for atezolizumab (95% CI 12.6 to 18.0) and 10.3 months for docetaxel (95% CI 8.8 to 12.0); HR 0.74 (95% CI 0.58 to 0.93; p = 0.0102). Overall objective response rates were 18% for atezolizumab and 16% for placebo in this subgroup. However, in the high-expression subgroup (137 patients, 16%) with ≥ 50% tumor cell expression or ≥ 10% immune cell expression, the OS benefit was considerably greater (Figure 5): median OS was 20.5 months for atezolizumab (95% CI 17.5 to not reached) and 8.9 months for decetaxel (95% CI 5.6 to 11.6); HR 0.41 (95% CI 0.27 to 0.64; p < 0.0001). The ORR was 31% for atezolizumab compared to 11% for docetaxel in this subgroup. There was no significant difference between the groups in terms of overall median PFS, although increasing benefit with atezolizumab was seen with increasing PDL-1 expression. No new safety signals were identified with atezolizumab in this study.Figure05

Atezolizumab, a PD-L1 inhibitor, is the third agent to demonstrate the superiority of immunotherapy over single-agent docetaxel in previously treated patients with NSCLC. As seen in KEYNOTE-010, the greatest survival benefit was seen in the subgroup of patients with high tumor or immune cell expression of PD-L1 (16% of subjects in OAK). In this subgroup, the HR of 0.41 (95% CI 0.27 to 0.64; p < 0.0001) is impressive indeed. Cross-comparisons of different trials are inappropriate, and so it remains unclear how closely the > 50% subgroup in the KEYNOTE studies compares to the TC3 or IC3 subgroup in the OAK and other pembrolizumab trials. Whereas approximately one third of patients fell into the high-expression group of KEYNOTE-024, the high-expression group defined in the OAK study was only half that. Does this really represent a different and extremely sensitive subgroup of patients, or do these apparent differences arise more from the properties of the antibodies used to detect PD-L1 expression? Comparative studies of the various antibodies in use and being developed as companion diagnostics for each agent are ongoing.

Discussion and Conclusions

The findings of KEYNOTE-024 show that PD-1 blockade with pembrolizumab represents a significant improvement in PFS and OS compared to platinum-doublet for first-line treatment of patients with NSCLC and a PD-L1 expression in ≥ 50% of tumor cells. These data establish the therapeutic value of pembrolizumab in first-line NSCLC patients with strong PD-L1 expression and further suggest that testing first-line NSCLC patients for PD-L1 expression would allow for the identification of those who could benefit the most from this treatment. Ongoing phase 3 research (e.g., KEYNOTE -04224) will determine the efficacy among patients with weaker PD-L1 expression (1-49% of tumor cells).

The updated and post-hoc findings with pembrolizumab in the phase 2/3 KEYNOTE-010 study also are encouraging, showing that this agent not only prolongs survival, but also is associated with better quality of life relative to chemotherapy.

Contrary to KEYNOTE-024, the results of CHECKMATE-026 were negative. Given the benefit seen in previous trials with nivolumab, these findings are somewhat surprising. However, there were substantial baseline differences between the nivolumab and chemotherapy groups that may have skewed the results in favor of chemotherapy (lower proportion of females, higher proportion of liver metastases, and lower number of ≥ 50% PD-L1 expressers in the nivolumab arm). Additionally, this trial was not powered to detect a between-group difference among patients with ≥ 50% PD-L1 expression (i.e., the population in KEYNOTE-024). It remains possible that nivolumab would demonstrate benefit in this group of patients. However, at the present time, evidence-based medicine dictates that this agent should not be used in first-line treatment. The MYSTIC trial of first-line treatment with durvalumab (a PD-L1 inhibitor) compared to platinum-based chemotherapy may shed further light on this topic. PD-L1 was not an entry criterion for this study, but tumor samples were required for all patients, and expression analyses are planned. This trial has completed accrual and results are expected in mid 2017.

Whether or not chemotherapy should be administered together with PD-1 blockade for first-line NSCLC treatment is a question that remains unanswered. While the findings of KEYNOTE-021 were positive for pembrolizumab + chemotherapy in this setting, we await the results of a phase 3 study to provide further information.

The results of the OAK study with atezolizumab in second-line treatment indicate that PD-L1 blockade also is an efficacious and safe approach in previously treated patients with high PD-L1 expression. Whether or not this will also be demonstrated in first-line use remains to be seen.


1. Califano R, Kerr K, Morgan RD, et al. Immune checkpoint blockade: a new era for non-small cell lung cancer. Curr Oncol Rep 2016; 18(9):59.

2. Lynch TJ, Bondarenko I, Luft A, et al. Ipilimumab in combination with paclitaxel and carboplatin as first-line treatment in stage IIIB/IV non-small-cell lung cancer: results from a randomized, double-blind, multicenter phase II study. J Clin Oncol 2012; 30(17):2046-54.

3. Rizvi NA, Mazieres J, Planchard D, et al. Activity and safety of nivolumab, an anti-PD-1 immune checkpoint inhibitor, for patients with advanced, refractory squamous non-small-cell lung cancer (CheckMate 063): a phase 2, single-arm trial. Lancet Oncol 2015; 16(3):257-65.

4. Gettinger SN, Horn L, Gandhi L, et al. Overall survival and long-term safety of nivolumab (Anti-Programmed Death 1 Antibody, BMS-936558, ONO-4538) in patients with previously treated advanced non-small-cell lung cancer. J Clin Oncol 2015; 33(18):2004-12.

5. Brahmer JR, Tykodi SS, Chow LQ, et al. Safety and activity of anti-PD-L1 antibody in patients with advanced cancer. N Engl J Med 2012; 366(26):2455-65.

6. Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N Engl J Med 2015; 373(17):1627-39.

7. Brahmer J, Reckamp KL, Baas P, et al. Nivolumab versus docetaxel in advanced squamous-cell non-small-cell lung cancer. N Engl J Med 2015; 373(2):123-35.

8. Gettinger S, Rizvi NA, Chow LQ, et al. Nivolumab monotherapy for first-line treatment of advanced non-small-cell lung cancer. J Clin Oncol 2016; 34(25):2980-7.

9. Garon EB, Rizvi NA, Hui R, et al. Pembrolizumab for the treatment of non-small-cell lung cancer. N Engl J Med 2015; 372(21):2018-28.

10. Herbst RS, Baas P, Kim DW, et al. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet 2016; 387(10027):1540-50.

11. Fehrenbacher L, Spira A, Ballinger M, et al. Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial. Lancet 2016; 387:1837-46.

12. Besse B, Johnson, M, Jänne, PA, et al. Phase II, single-arm trial (BIRCH) of atezolizumab as first-line or subsequent therapy for locally advanced or metastatic PD-L1-selected non-small cell lung cancer (NSCLC). Presented at the European Cancer Congress 2015; abstract 16LBA.

13. Reck M, Rodríguez-Abreu D, Robinson A, et al. KEYNOTE-024: Pembrolizumab (pembro) vs platinum-based chemotherapy (chemo) as first-line therapy for advanced NSCLC with a PD-L1 tumour proportion score (TPS) >50%. Presented at ESMO 2016; presentation LBA8-PR.

14. Socinski MA, Creelan B, Horn L, et al. CheckMate 026: A phase 3 trial of nivolumab vs investigator's choice (IC) of platinum-based doublet chemotherapy (PT-DC) as first-line therapy for stage IV/recurrent programmed death ligand 1 (PD-L1)-positive NSCLC. Presented at ESMO 2016; presentation LBA7-PR.

15. Reck M, Rodríguez-Abreu D, Robinson A, et al. Pembrolizumab or chemotherapy in PD-L1-positive non-small-cell lung cancer. N Engl J Med 2016; [epub ahead of print].

16. Villaruz LC, Socinski MA. The clinical utility of PD-L1 testing in selecting non-small cell lung cancer patients for PD1/PD-L1-directed therapy. Clin Pharmacol Ther 2016; 100(3):212-4.

17. Langer CJ, Gadgeel SM, Borghaei H, et al. Randomized phase 2 study of carboplatin and pemetrexed ± pembrolizumab as first-line therapy for advanced NSCLC: KEYNOTE-021 cohort G. Presented at ESMO 2016; presentation LBA46-PR.

18. Langer CJ, Gadgeel SM, Borghaei H, et al. Carboplatin and pemetrexed with or without pembrolizumab for advanced, non-squamous non-small-cell lung cancer: a randomised, phase 2 cohort of the open-label KEYNOTE-021 study. Lancet Oncol 2016; [epub ahead of print].

19. Barlesi F, Park K, Ciardiello F, et al. Primary analysis from OAK, a randomized phase III study comparing atezolizumab with docetaxel in advanced NSCLC. Presented at ESMO 2016; Presentation #LBA44_PR.

20. Herbst RS, Baas P, Kim D, et al. Pembrolizumab (pembro) vs docetaxel (doce) for previously treated, PD-L1–expressing NSCLC: Updated outcomes of KEYNOTE-010. Presented at ESMO 2016; poster LBA48.

21. Barlesi F, Garon EB, Kim D-W, et al. Assessment of health-related quality of life in KEYNOTE-010: a phase 2/3 study of pembrolizumab versus docetaxel in patients with previously treated advanced NSCLC. Presented at ESMO 2016; poster 1219P.

22. Hall RD, Gadgeel SM, Garon EB, et al. Phase 3 study of platinum-based chemotherapy with or without pembrolizumab for first-line metastatic, nonsquamous non-small cell lung carcinoma (NSCLC): KEYNOTE-189. Presented at ASCO 2016; poster TPS9104.

23. Herbst RS, Baas P, Kim DW, et al. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial. Lancet 2016; 387(10027):1540-50.

24. Mok T, Wu Y-L, Watson PA, et al. Phase 3 KEYNOTE-042 trial of pembrolizumab (MK-3475) versus platinum doublet chemotherapy in treatment-naive patients (pts) with PD-L1–positive advanced non-small cell lung cancer (NSCLC). Presented at ASCO 2015; poster TPS8105.

Copyright 2016 STA HealthCare Communications Inc. All rights reserved. This report was made possible through the support of Merck Canada Inc. The opinions and information contained herein are those of the author and do not necessarily reflect the views or opinions of Merck Canada Inc. or STA HealthCare Communications Inc. Any products mentioned herein should be used in accordance with the prescribing information contained in their respective product monograph.