By: Nick Turner, PhD, Senior DirectorLung cancer is the leading cause of cancer mortality worldwide, accounting for more than 1.6 million deaths annually. Non-small cell lung cancer (NSCLC) accounts for 85% of all new cases. Current first-line treatment decisions for metastatic NSCLC mandate analysis of the genetic status of the cancer to determine the presence of driver mutations, such as activating mutations in the epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK). In mutation-positive patients, first-line treatment comprises small molecule inhibitors of EGFR-tyrosine kinase, e.g., gefitinib, afatinib or osimertinib, or ALK inhibitors, e.g., alectinib or crizotinib. Roughly 85% of patients with NSCLC, however, do not harbor these oncogenic drivers; in these patients, treatment options have been limited to platinum-based chemotherapy. The emergence of immune checkpoint inhibitors, which attenuate immunosuppression within the tumor microenvironment and enable the immune system to attack tumor cells, opens the door to more effective therapy with durable responses and manageable toxicity.

Pembrolizumab and the Changing Treatment Paradigm

KEYNOTE-024 showed that median progression-free survival (mPFS) in treatment-naive NSCLC was significantly improved in patients who received pembrolizumab relative to the standard of care platinum-doublet chemotherapy (10.3 months vs. 6.0 months), with a hazard ratio (HR) for disease progression or death of 0.51. More strikingly, the risk of death was 37% lower in the pembrolizumab arm of the study, with median overall survival (mOS) of 30 months compared to 14.2 months on chemotherapy2.

KEYNOTE-024, however, is limited in that the trial was restricted to patients with high expression of the biomarker PD-L1 (i.e., those patients with a tumor proportion score (TPS) ≥50%), which represent only around one-third of the patient pool.

Moreover, the water was muddied by the failure of BMS’ nivolumab in CheckMate-026 to show a survival benefit in patients with a broader range of PD-L1 expression (TPS ≥5%)3.

The question of what to do in patients with low PD-L1 expression (TPS <50%) was answered by KEYNOTE-1894,5 This study looked at potential synergies between pembrolizumab and four cycles of platinum-pemetrexed chemotherapy and showed that the risk of death was reduced by 51% in patients treated with the combination relative to chemotherapy (HR 0.49, mOS >21 months). Crucially, this survival benefit was seen regardless of PD-L1 expression, with pronounced risk reduction seen in PD-L1-high (TPS ≥50%; HR 0.42, 12-month OS rate 73% vs. 48%), PD-L1-low (TPS 1-49%; HR 0.55, 12-month OS rate 71.5% vs. 51%) and PD-L1-negative (TPS <1%; HR 0.59, 12-month OS rate 61.7% vs. 52%) patients5.

Based on these outcomes, draft guidelines6 propose single-agent pembrolizumab as the standard of care for NSCLC in patients in whom PD-L1 is expressed in ≥50% of tumor cells, while patients whose tumor PD-L1 expression (TPS) is between 1% and 50% should be considered for triplet therapy with pembrolizumab plus carboplatin-pemetrexed. At the moment, pembrolizumab is the only checkpoint inhibitor approved for the first-line treatment of NSCLC.

PD-L1 Inhibitor Uncertainties

Merck has an application undergoing priority review that, if approved, would expand the patient population eligible for pembrolizumab monotherapy to those with PD-L1 expression TPS ≥1%, based on KEYNOTE-0427. Although the trial reported superiority of pembrolizumab over chemotherapy, it is questionable whether expansion of the drug’s label is supported by the data. Thus, whereas the efficacy of pembrolizumab monotherapy in the PD-L1 TPS ≥50% population is unequivocal (HR 0.69, mOS of 20 months vs. 12.2 months), there is little if any impact on OS (HR 0.92, mOS of 13.4 months vs. 12.1 months) in patients with PD-L1 TPS 1-49% or PFS in those with PD-L1 TPS ≥1% (HR 1.07) or ≥20% (HR 0.94)7.

Moreover, KEYNOTE-042 reveals that more than half of subjects with PD-L1 TPS ≥1% have accelerated disease progression on pembrolizumab monotherapy relative to those treated with chemotherapy (mPFS of 5.4 months vs. 6.5 months). This result suggests that pembrolizumab  monotherapy carries a survival liability early in treatment relative to chemotherapy. Furthermore, early disease progression is seen across all strata of PD-L1 expression (≥50%, ≥20%, ≥1%). Whether this is perceived as clinically meaningful is a matter of debate but may be clarified by the FDA’s review (which is anticipated on January 11, 2019).

In the context of the debate over whether pembrolizumab plus chemotherapy is preferable to pembrolizumab monotherapy in patients with PD-L1 TPS ≥50%, not only does KEYNOTE-189 raise the efficacy bar, it also establishes that pembrolizumab combined with chemotherapy mitigates any potential early survival liability.

Superiority of the PD-L1 inhibitor atezolizumab plus carboplatin-pemetrexed over chemotherapy alone is also established in Roche’s IMpower132 study8, which showed improvements in tumor response and PFS comparable to those seen in KEYNOTE-1895. The absence of a significant impact on OS (HR 0.81 at interim analysis), however, should mean that pembrolizumab’s position as a backbone of therapy for metastatic NSCLC is likely to remain unchallenged for the foreseeable future. Pembrolizumab plus chemotherapy should therefore be standard of care in all patients with NSCLC except those with low performance status or those who otherwise cannot tolerate chemotherapy.

The Combination of PD1/PD-L1 inhibitors with Other Checkpoint Inhibitors

Approaches combining PD1/PD-L1 inhibitors with other checkpoint inhibitors have proven less successful, with both nivolumab (CheckMate-227) and durvalumab (MYSTIC) in combination with CTLA-4 inhibitors disappointing on PFS and OS9,10. Although patients with high tumor mutational burden (TMB) derive a survival benefit from dual checkpoint blockade with nivolumab plus ipilimumab (mPFS of 7.2 months vs. 5.5 months; mOS of 23 months vs. 16.7 months, HR 0.77)9,11, efficacy appears inferior to that achieved with pembrolizumab and chemotherapy (mPFS of 8.8 months vs. 4.9 months; mOS of NR vs. 11.3 months, HR 0.49)4,5 and, paradoxically, is independent of PD-L1 expression level9. An exception may exist in patients with high-TMB and PD-L1 <1%12, where the HR on PFS favors nivolumab plus ipilimumab (HR, 0.48)12 over pembrolizumab plus chemotherapy (HR 0.75)5.

What this means is that, if nivolumab plus ipilimumab has a role in NSCLC, it is likely to be limited to PD-L1-negative (TPS <1%), TMB-high disease where cross-trial comparison is suggestive of a greater PFS benefit compared to pembrolizumab plus chemotherapy. BMS, however, will have to convince the clinical community of the utility of TMB as a predictive biomarker. TMB’s value was called into doubt recently by an exploratory analysis that showed that the HR for OS in patients with TMB >10 mut/Mb was comparable to that in patients with TMB <10 mut/Mb11. One explanation for the apparent dichotomy is that TMB may select for PFS but not OS, which could leave TMB and nivolumab out in the cold.

If, indeed, pembrolizumab in combination with chemotherapy is established as standard of care in metastatic NSCLC, there remain several open questions:

  • Can we predict the responder population? So far, PD-L1 status has the greatest traction. CheckMate-0263 and -2279 suggest that PD-L1-low but TMB-high patients may derive benefit from nivolumab, although this has been called into doubt by a recent exploratory analysis (see above). Other markers such as tumor gene expression, tumor-infiltrating lymphocytes and circulating tumor DNA are under investigation.
  • Do we need to predict tumor response at all, if it has been shown to be a poor indicator of OS?
  • What therapeutic options are open to those patients with disease progression on first-line PD1/PD-L1 inhibitors? The durable OS responses to pembrolizumab plus chemotherapy in KEYNOTE-189 suggest that pembrolizumab treatment should continue beyond progression.
  • Can patients with progression on first-line PD1/PD-L1 inhibitors mount an anti-tumor immune response on second-line agents?
  • Are there therapeutic combinations that can lift the tail of the survival curve?
  • What options are there for fast progressors?

KEYNOTE-189 provides a tantalizing glimpse into long-term survival. Extended follow-up of patients in this trial and other chemotherapy combination trials should be encouraged.


  1. Reck, M. “Pembrolizumab versus Chemotherapy for PD-L1–Positive Non–Small-Cell Lung Cancer.” New England Journal of Medicine, Vol. 375 (2016): 1823-1833.
  2. Gottfried, M. “Updated Analysis of KEYNOTE-024: Pembrolizumab vs Platinum-Based Chemotherapy for Advanced NSCLC With PD-L1 TPS ≥50%.” WCLC 2017, Abstract OA 17.06.
  3. Carbone, D.P. “First-Line Nivolumab in Stage IV or Recurrent Non–Small-Cell Lung Cancer.” New England Journal of Medicine, Vol. 376 (2017): 2415-2426.
  4. Gandhi, L. “Pembrolizumab plus Chemotherapy in Metastatic Non–Small-Cell Lung Cancer.” New England Journal of Medicine, Vol. 378 (2018): 2078-2092.
  5. Gandhi, L., et al. “KEYNOTE-189: Randomized, double-blind, phase 3 study of pembrolizumab (pembro) or placebo plus pemetrexed (pem) and platinum as first-line therapy for metastatic NSCLC.” AACR 2018, Abstract CT075.
  6. Brahmer, J.R., et al. “The Society for Immunotherapy of Cancer consensus statement on immunotherapy for the treatment of non-small cell lung cancer (NSCLC).” Journal for Immunotherapy of Cancer, Vol. 6 (2018): 75-90.
  7. Lopes, G., et al. “Pembrolizumab (pembro) versus platinum-based chemotherapy (chemo) as first-line therapy for advanced/metastatic NSCLC with a PD-L1 tumor proportion score (TPS) ≥ 1%: Open-label, phase 3 KEYNOTE-042 study.” ASCO 2018, Abstract LBA4.
  8. Barlesi, F. “IMpower132: efficacy of atezolizumab (atezo) + carboplatin (carbo)/cisplatin (cis) + pemetrexed (pem) as 1L treatment in key subgroups with stage IV non-squamous non-small cell lung cancer (NSCLC).” ESMO 2018, Abstract LBA54.
  9. Hellmann, M., et al. “Nivolumab plus Ipilimumab in Lung Cancer with a High Tumor Mutational Burden.” New England Journal of Medicine, Vol. 378 (2018): 2093-2104.
  10. AstraZeneca press release.
  11. BMS press release.
  12. Borghaei, H., et al. “Nivolumab + ipilimumab, nivolumab + chemotherapy, and chemotherapy in chemo-naïve patients with advanced non-small cell lung cancer and <1% tumor PD-L1 expression: Results from CheckMate 227.” ASCO 2018, Abstract 9001.