New England Journal of Medicine May 31, 2018

Pembrolizumab plus Chemotherapy in Metastatic Non-Small-Cell Lung Cancer

Leena Gandhi, Delvys Rodríguez-Abreu, Shirish Gadgeel et al.

Bottom Line

In patients with previously untreated metastatic nonsquamous non-small-cell lung cancer without targetable mutations, the addition of pembrolizumab to platinum-pemetrexed chemotherapy significantly improved overall survival and progression-free survival compared to chemotherapy alone.

Key Findings

1. After a median follow-up of 10.5 months, the estimated 12-month overall survival (OS) rate was 69.2% in the pembrolizumab-combination group versus 49.4% in the placebo-combination group (HR for death, 0.49; 95% CI, 0.38 to 0.64; P<0.001).

2. Median progression-free survival (PFS) was significantly prolonged in the pembrolizumab arm (8.8 months vs. 4.9 months; HR for disease progression or death, 0.52; 95% CI, 0.43 to 0.64; P<0.001).

3. The overall survival benefit of adding pembrolizumab was consistent across all evaluated PD-L1 expression categories, including tumors with less than 1% PD-L1 expression.

4. The objective response rate (ORR) was substantially higher in the pembrolizumab-combination group compared to the placebo-combination group (47.6% vs. 18.9%; P<0.001).

5. Rates of grade 3 or higher adverse events were similar between the two groups (67.2% with pembrolizumab vs. 65.8% with placebo), demonstrating that the addition of immunotherapy did not drastically alter the toxicity profile of the chemotherapy backbone, though acute kidney injury was slightly more common in the pembrolizumab cohort.

Study Design

Design

Phase 3 RCT

Double-Blind

Sample

616

Patients

Duration

10.5 mo

Median

Setting

Multicenter, international

Population Patients with previously untreated metastatic nonsquamous NSCLC without sensitizing EGFR or ALK mutations.

Intervention Pembrolizumab (200 mg q3w) + pemetrexed + carboplatin/cisplatin for 4 cycles, followed by maintenance pembrolizumab + pemetrexed.

Comparator Placebo + pemetrexed + carboplatin/cisplatin for 4 cycles, followed by maintenance placebo + pemetrexed.

Outcome Overall survival (OS) and progression-free survival (PFS)

Study Limitations

• The initial reporting had a relatively short median follow-up of 10.5 months, which limited the early assessment of late toxicities and long-term survival plateaus (though later updates have confirmed durability).

• There was a high effective crossover rate (approximately 50% of patients in the control arm subsequently received anti-PD-1/PD-L1 therapy), which biases overall survival analyses toward the null, though pembrolizumab still showed a massive OS benefit.

• The trial did not directly compare chemoimmunotherapy to pembrolizumab monotherapy in the subgroup of patients with PD-L1 expression ≥50%, leaving the optimal regimen choice for this high-expresser population open to clinical debate.

• Patients with untreated brain metastases, active autoimmune disease, or actionable EGFR/ALK genomic aberrations were excluded, limiting the direct generalizability of the findings to these specific populations.

Clinical Significance

KEYNOTE-189 was a landmark paradigm-shifting trial that established pembrolizumab in combination with platinum-based chemotherapy and pemetrexed as the undisputed first-line standard of care for patients with metastatic nonsquamous NSCLC lacking targetable drivers. By demonstrating a halving of the risk of death compared to chemotherapy alone—regardless of PD-L1 expression—it effectively rendered standalone cytotoxic chemotherapy obsolete in this clinical setting.

Historical Context

Prior to KEYNOTE-189, pembrolizumab monotherapy had recently revolutionized the treatment of advanced NSCLC, but its use was restricted as a first-line option only to patients whose tumors had high PD-L1 expression (TPS ≥50%) based on the KEYNOTE-024 trial. For the vast majority of patients with PD-L1 <50%, platinum-doublet chemotherapy remained the standard. While the phase 2 KEYNOTE-021G trial provided an early signal that adding pembrolizumab to chemotherapy could improve outcomes, it was the definitive phase 3 KEYNOTE-189 trial that cemented chemoimmunotherapy as the universal backbone for metastatic nonsquamous NSCLC wild-type for EGFR/ALK.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

Pembrolizumab targets the PD-1 receptor. How does blocking the PD-1/PD-L1 pathway synergize mechanistically with the cytotoxic effects of platinum-pemetrexed chemotherapy in non-small-cell lung cancer?

Key Response

Chemotherapy induces immunogenic cell death, leading to the release of tumor neoantigens and potentially upregulating PD-L1 expression on the tumor and within the microenvironment. Concurrently, blocking PD-1 prevents exhaustion of the newly primed cytotoxic T-cells, allowing the immune system to mount a robust and sustained response against the exposed tumor antigens.

Resident

Before initiating the KEYNOTE-189 regimen (pembrolizumab plus platinum-pemetrexed), what specific molecular testing must be resulted, and why might starting immunotherapy prematurely be harmful if certain actionable mutations are present?

Key Response

EGFR and ALK testing (among others like ROS1) must be resulted as negative, because this chemoimmunotherapy regimen is less effective in oncogene-driven NSCLC. Furthermore, starting pembrolizumab and later switching to targeted therapies like an EGFR TKI (e.g., osimertinib) upon late resulting of biomarkers significantly increases the risk of severe immune-mediated pneumonitis and hepatotoxicity.

Fellow

KEYNOTE-189 demonstrated an overall survival benefit across all PD-L1 subgroups, including those with a Tumor Proportion Score (TPS) <1%. When counseling a fit patient with a TPS of 90%, how do you weigh the KEYNOTE-189 chemoimmunotherapy data against the KEYNOTE-024 single-agent pembrolizumab data?

Key Response

While chemoimmunotherapy benefits all comers and offers a faster time to response (critical for highly symptomatic patients or those with high disease burden), patients with TPS >50% also have the option of single-agent pembrolizumab. Fellows must balance the potentially higher objective response rate of the combination against the increased and cumulative toxicity of chemotherapy, relying on shared decision-making since no head-to-head trial has definitively proven combination superiority over monotherapy in the >50% subgroup.

Attending

Given the robust survival curves and durable responses seen in KEYNOTE-189, how do you manage the decision to discontinue pembrolizumab at the 2-year mark (35 cycles) as done in the trial, considering the lack of definitive prospective data guiding cessation?

Key Response

Attendings must navigate the clinical uncertainty when a patient reaches the 2-year mark with a complete or deep partial response. The trial capped therapy at 35 cycles, suggesting durable benefit even after cessation. The decision involves balancing the risk of late-onset, potentially irreversible immune-related adverse events and financial toxicity against the patient's psychological fear of disease progression, requiring highly nuanced shared decision-making.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The KEYNOTE-189 trial utilized a crossover design allowing patients in the placebo-chemotherapy arm to receive pembrolizumab upon progression. How does this crossover affect the interpretation of the Overall Survival (OS) endpoint, and what advanced statistical methods are required to estimate the true OS benefit?

Key Response

Crossover naturally dilutes the apparent OS benefit of the experimental arm, making the intention-to-treat Hazard Ratio an underestimate of the true efficacy of early versus late pembrolizumab. Advanced statistical methods, such as the Rank Preserving Structural Failure Time (RPSFT) model or inverse probability of censoring weighting (IPCW), are necessary to adjust for this confounding crossover effect and estimate the true magnitude of the survival advantage.

Journal Editor

In evaluating the methodology and trial design of KEYNOTE-189, does the lack of a pembrolizumab-monotherapy arm limit the trial's ability to definitively answer the optimal first-line strategy, and how should this influence editorial acceptance of broad claims regarding the 'universal' superiority of chemoimmunotherapy?

Key Response

A critical peer reviewer would note that without a pembrolizumab-only arm, KEYNOTE-189 cannot answer whether the addition of chemotherapy is truly necessary for patients with high PD-L1 expression (>50%). Editors should flag conclusions that overstate the necessity of the combination for all patients, ensuring the authors acknowledge this structural limitation and the reliance on cross-trial comparisons to address this specific high-expressor population.

Guideline Committee

Based on the KEYNOTE-189 efficacy data, how should clinical practice guidelines (e.g., NCCN, ASCO) classify the strength of recommendation for this regimen in metastatic nonsquamous NSCLC without actionable mutations, and how does this fundamentally alter the algorithm for the PD-L1 <1% population?

Key Response

Following KEYNOTE-189, guidelines updated this regimen to a Category 1, preferred recommendation regardless of PD-L1 expression level. This fundamentally shifted the treatment paradigm for the PD-L1 <1% subgroup, retiring cytotoxic chemotherapy alone as the standard of care and mandating the integration of upfront immunotherapy for all wild-type nonsquamous NSCLC patients who do not have contraindications to immune checkpoint inhibitors.

Clinical Landscape

Noteworthy Related Trials

2016

KEYNOTE-024

n = 305 · NEJM

Tested

Pembrolizumab 200 mg every 3 weeks

Population

Previously untreated advanced NSCLC with PD-L1 expression >= 50% without EGFR/ALK mutations

Comparator

Platinum-based chemotherapy

Endpoint

Progression-free survival (PFS)

Key result: Pembrolizumab significantly prolonged progression-free survival and overall survival compared to chemotherapy and was associated with fewer adverse events.

2018

KEYNOTE-407

n = 559 · NEJM

Tested

Pembrolizumab plus carboplatin and paclitaxel/nab-paclitaxel

Population

Previously untreated metastatic squamous NSCLC

Comparator

Placebo plus chemotherapy

Endpoint

Overall survival and progression-free survival

Key result: The addition of pembrolizumab to chemotherapy significantly prolonged overall and progression-free survival in patients with squamous NSCLC, regardless of PD-L1 expression.

2020

CheckMate 9LA

n = 719 · Lancet Oncol

Tested

Nivolumab plus ipilimumab combined with 2 cycles of chemotherapy

Population

Previously untreated advanced NSCLC without EGFR/ALK mutations

Comparator

Chemotherapy alone

Endpoint

Overall survival (OS)

Key result: Nivolumab plus ipilimumab with limited chemotherapy provided a significant improvement in overall survival versus chemotherapy alone.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis