New England Journal of Medicine NOVEMBER 11, 2021

Pembrolizumab for Persistent, Recurrent, or Metastatic Cervical Cancer (KEYNOTE-826)

Nicoletta Colombo, Coraline Dubot, Domenica Lorusso, et al. (KEYNOTE-826 Investigators)

Bottom Line

The addition of pembrolizumab to first-line platinum-based chemotherapy, with or without bevacizumab, significantly improves overall and progression-free survival in patients with persistent, recurrent, or metastatic cervical cancer.

Key Findings

1. In the overall population, pembrolizumab plus chemotherapy (with or without bevacizumab) significantly improved overall survival (median OS 26.4 months vs. 16.8 months, HR 0.63; 95% CI, 0.52–0.77).
2. Progression-free survival was similarly superior in the pembrolizumab arm, with a median of 10.4 months compared to 8.2 months in the placebo arm (HR 0.65; 95% CI, 0.53–0.79).
3. Clinical benefits were robust in the PD-L1 combined positive score (CPS) ≥1 population (median OS 28.6 months vs. 16.5 months, HR 0.60) and were consistent regardless of bevacizumab use.
4. Grade 3 or higher adverse events were slightly more frequent in the pembrolizumab group (82.4%) compared to the placebo group (75.4%), indicating a manageable safety profile consistent with known immunotherapy toxicities.

Study Design

Design
RCT
Double-Blind
Sample
617
Patients
Duration
39.1 mo
Median
Setting
Multicenter, International
Population Patients with persistent, recurrent, or metastatic cervical cancer not previously treated with systemic chemotherapy and not amenable to curative treatment.
Intervention Pembrolizumab (200 mg every 3 weeks) plus platinum-based chemotherapy (paclitaxel with cisplatin or carboplatin), with or without bevacizumab.
Comparator Placebo plus platinum-based chemotherapy (paclitaxel with cisplatin or carboplatin), with or without bevacizumab.
Outcome Overall survival (OS) and progression-free survival (PFS).

Study Limitations

The study was not randomized for the use of bevacizumab, making definitive conclusions on its specific interaction with pembrolizumab challenging, though benefit was observed regardless of its inclusion.
Data regarding the efficacy of pembrolizumab in patients with PD-L1 CPS 0 remain less definitive compared to the PD-L1 positive populations.
The potential for immune-related adverse events requires diligent long-term monitoring in a population often burdened by high symptom scores from advanced disease.

Clinical Significance

This trial established a new standard of care for the first-line treatment of persistent, recurrent, or metastatic cervical cancer, providing a transformative therapeutic option that demonstrates durable survival benefits for patients across varying PD-L1 expression levels.

Historical Context

Prior to KEYNOTE-826, treatment options for advanced cervical cancer were limited primarily to platinum-doublet chemotherapy, often combined with bevacizumab based on GOG 240, with historically poor outcomes. The KEYNOTE-826 trial introduced immune checkpoint inhibition into the first-line setting, drastically changing the therapeutic landscape and offering significantly improved life expectancy.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the mechanism of action of pembrolizumab in cervical cancer, and why is the Combined Positive Score (CPS) used instead of the Tumor Proportion Score (TPS) as a biomarker?

Key Response

Pembrolizumab is a humanized monoclonal antibody that binds to the PD-1 receptor on T-cells, blocking its interaction with PD-L1 and PD-L2 on tumor cells and restoring the anti-tumor immune response. In cervical cancer, the CPS is used because it accounts for PD-L1 expression on both tumor cells and tumor-infiltrating immune cells (macrophages, lymphocytes), which is more representative of the inflamed microenvironment often caused by chronic HPV infection than TPS alone.

Resident
Resident

In a patient with metastatic cervical cancer and a PD-L1 CPS of 5, how should the results of the KEYNOTE-826 trial influence your selection of first-line therapy compared to the previous GOG 240 standard?

Key Response

Before KEYNOTE-826, the standard was a platinum doublet plus bevacizumab (based on GOG 240). KEYNOTE-826 demonstrated that adding pembrolizumab to this backbone significantly improved both progression-free survival (PFS) and overall survival (OS). For a patient with CPS ≥1, the Hazard Ratio for death was 0.64, making the addition of pembrolizumab the new preferred first-line treatment for patients with PD-L1-positive disease.

Fellow
Fellow

Does the survival benefit observed with pembrolizumab in KEYNOTE-826 persist in patients who do not receive concurrent bevacizumab, and how does this impact treatment for those with contraindications to anti-angiogenics?

Key Response

Subgroup analyses in KEYNOTE-826 indicated that the benefit of pembrolizumab was consistent regardless of whether bevacizumab was used. This is clinically significant for patients with contraindications to bevacizumab, such as those with a history of fistulas, uncontrolled hypertension, or major surgery, as they can still achieve superior survival outcomes by adding pembrolizumab to their chemotherapy alone.

Attending
Attending

With pembrolizumab now established in the first-line setting for metastatic cervical cancer, what are the implications for therapeutic sequencing when a patient progresses, particularly regarding the use of tisotumab vedotin or other immunotherapy agents?

Key Response

Moving pembrolizumab to the first line creates a 'sequencing gap' because prior data for second-line immunotherapy (like KEYNOTE-158) were based on IO-naive patients. Upon progression on first-line pembrolizumab-chemo, practitioners must now look toward agents with different mechanisms, such as the antibody-drug conjugate tisotumab vedotin, as there is currently no evidence that continuing PD-1/PD-L1 inhibition beyond progression or switching to another IO agent is effective in this population.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

Evaluate the statistical design of KEYNOTE-826 regarding its dual-primary endpoint strategy and the hierarchical testing across the three PD-L1 CPS populations. How does this design maintain the family-wise error rate?

Key Response

The study utilized a complex graphical method to allocate the type I error (alpha) across the primary endpoints of PFS and OS. Significance was tested sequentially: first in the CPS ≥1 group, then in the CPS ≥10 group, and finally in the all-comers population. This gatekeeping strategy ensures that multiple comparisons across subgroups do not inflate the risk of a false positive finding, ensuring that the results for the all-comer population are statistically valid even though the majority of benefit was driven by PD-L1 positive patients.

Journal Editor
Journal Editor

How do the 'all-comer' results in KEYNOTE-826 potentially mask the efficacy profile in the PD-L1 negative (CPS <1) subgroup, and is there enough evidence to justify regulatory approval in this specific subset?

Key Response

Approximately 88% of the trial population had a CPS ≥1. While the trial was positive for the 'all-comer' population, a critical reviewer would note that the benefit in the small CPS <1 subgroup was not clearly established and likely did not reach statistical significance independently. This raises concerns about whether 'all-comer' labeling is appropriate or if it simply reflects the overwhelming efficacy in the PD-L1 positive majority, a common point of contention in immunotherapy trials.

Guideline Committee
Guideline Committee

Based on KEYNOTE-826, should the recommendation for pembrolizumab + chemotherapy be limited to PD-L1 positive disease, and how does this align with current NCCN and ASCO guidelines?

Key Response

While the FDA-approved label often follows the all-comer data, NCCN and ASCO guidelines specifically emphasize the CPS ≥1 population because that is where the clinical benefit is most robust (Category 1 recommendation). For patients with PD-L1 negative disease (CPS <1), the evidence is less compelling, and guidelines may still suggest the GOG 240 regimen (Chemo + Bevacizumab) alone as an acceptable alternative, highlighting the importance of PD-L1 testing at diagnosis.

Clinical Landscape

Noteworthy Related Trials

2014

GOG 240 Trial

n = 452 · NEJM

Tested

Paclitaxel and cisplatin or topotecan plus bevacizumab

Population

Persistent, recurrent, or metastatic carcinoma of the cervix

Comparator

Paclitaxel and cisplatin or topotecan without bevacizumab

Endpoint

Overall survival

Key result: The addition of bevacizumab to chemotherapy resulted in a 3.7-month improvement in median overall survival compared to chemotherapy alone.
2019

KEYNOTE-158 Trial

n = 98 · JCO

Tested

Pembrolizumab 200 mg every 3 weeks

Population

Previously treated advanced cervical cancer with PD-L1 positive tumors

Comparator

None (single-arm study)

Endpoint

Objective response rate

Key result: Pembrolizumab demonstrated durable antitumor activity with an objective response rate of 14.3% in patients with PD-L1 positive recurrent or metastatic cervical cancer.
2019

CheckMate 358 Trial

n = 19 · Gynecol Oncol

Tested

Nivolumab 240 mg every 2 weeks

Population

Recurrent or metastatic cervical cancer

Comparator

None (single-arm study)

Endpoint

Objective response rate

Key result: Nivolumab showed clinical activity with a manageable safety profile in patients with recurrent or metastatic cervical cancer regardless of PD-L1 expression.

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