The New England Journal of Medicine February 27, 2020

Pembrolizumab for Early Triple-Negative Breast Cancer

Peter Schmid, Javier Cortes, Lajos Pusztai, Heather McArthur, Sherko Kummel, Jonas Bergh, Carsten Denkert et al.

Bottom Line

In patients with early-stage triple-negative breast cancer, the addition of neoadjuvant and adjuvant pembrolizumab to standard chemotherapy significantly increased the pathological complete response rate and showed early improvements in event-free survival compared to chemotherapy alone.

Key Findings

1. Among the first 602 randomized patients, pathological complete response (pCR) was significantly higher in the pembrolizumab group compared to the placebo group (64.8% vs. 51.2%; estimated absolute difference of 13.6 percentage points, P<0.001) [3.2.2].
2. At a median follow-up of 15.5 months across all 1,174 patients, disease progression, recurrence, or death occurred in 7.4% of the pembrolizumab group versus 11.8% of the placebo group (HR 0.63; 95% CI, 0.43-0.93).
3. Grade 3 or higher treatment-related adverse events occurred in 78.0% of the pembrolizumab arm and 73.0% of the placebo arm.
4. Immune-mediated adverse events of any grade (e.g., hypothyroidism, severe skin reactions, and adrenal insufficiency) occurred more frequently with pembrolizumab, emphasizing the need for rigorous monitoring in patients treated with curative intent.

Study Design

Design
Phase 3 RCT
Double-Blind
Sample
1,174
Patients
Duration
15.5 mo
Median
Setting
Multicenter, 21 countries
Population Patients with previously untreated, high-risk stage II or III triple-negative breast cancer (TNBC)
Intervention Neoadjuvant pembrolizumab (200 mg q3w) plus paclitaxel and carboplatin for 4 cycles, followed by pembrolizumab plus doxorubicin/epirubicin and cyclophosphamide for 4 cycles; followed by adjuvant pembrolizumab for up to 9 cycles
Comparator Neoadjuvant placebo plus identical chemotherapy regimen; followed by adjuvant placebo
Outcome Pathological complete response (pCR, defined as ypT0/Tis ypN0) and Event-free survival (EFS)

Study Limitations

The primary 2020 report had a short median follow-up of only 15.5 months, limiting mature conclusions on long-term event-free survival and overall survival [3.2.1].
The trial design did not allow for determining the individual relative efficacy contributions of the neoadjuvant versus the adjuvant pembrolizumab phases.
Adjuvant capecitabine, which became standard of care based on the CREATE-X trial for patients with residual disease, was not permitted in the study protocol.
The addition of pembrolizumab substantially increased the risk of irreversible immune-mediated endocrinopathies, adding lifelong morbidity to a potentially curable population.

Clinical Significance

KEYNOTE-522 established a new global standard of care for patients with high-risk, early-stage triple-negative breast cancer (TNBC). The clinically meaningful 13.6% absolute increase in pCR translated into longer-term event-free survival and overall survival benefits (proven in 2022 and 2024 trial updates), justifying the upfront integration of perioperative pembrolizumab independent of PD-L1 status. This regimen fundamentally shifted curative-intent TNBC management, obligating oncologists to balance improved survival against the management of immune-related adverse events.

Historical Context

Historically, early-stage TNBC was managed primarily with anthracycline- and taxane-based neoadjuvant chemotherapy, yet recurrence rates remained distressingly high and overall survival was lower than other breast cancer subtypes. Prior to KEYNOTE-522, immune checkpoint inhibitors had only shown efficacy in the metastatic, PD-L1-positive TNBC setting (e.g., IMpassion130). KEYNOTE-522 was a landmark trial because it successfully moved immunotherapy into the early-stage, potentially curable setting for TNBC, demonstrating that the addition of PD-1 blockade to aggressive platinum-containing neoadjuvant chemotherapy could overcome the aggressive biology of the disease.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

Pembrolizumab targets the PD-1 receptor. Why is the PD-1/PD-L1 pathway particularly relevant in triple-negative breast cancer (TNBC) compared to hormone receptor-positive breast cancers, and how does concurrent neoadjuvant chemotherapy enhance the effect of this immunotherapy?

Key Response

This tests foundational immunology and pathophysiology. TNBC typically exhibits a higher tumor mutational burden, more tumor-infiltrating lymphocytes (TILs), and higher PD-L1 expression than other breast cancer subtypes, making it highly immunogenic. Chemotherapy induces immunogenic cell death, releasing tumor neoantigens and priming the immune system, which creates a synergistic effect when combined with a checkpoint inhibitor like pembrolizumab.

Resident
Resident

In the KEYNOTE-522 trial, adding pembrolizumab increased the rate of immune-related adverse events (irAEs). If a patient on this regimen presents with new-onset profound fatigue, mild hypotension, and vague abdominal pain, what is your differential diagnosis for irAEs, and what specific urgent laboratory workup is indicated?

Key Response

This addresses vital clinical management for residents. The presentation is highly suspicious for immune-mediated endocrinopathies, specifically primary adrenal insufficiency (hypophysitis or adrenalitis), though immune hepatitis and colitis must also be considered. Urgent workup requires a CMP, TSH, free T4, 8 AM cortisol, ACTH levels, and potentially an ACTH stimulation test, as adrenal crisis is a life-threatening oncologic emergency.

Fellow
Fellow

The KEYNOTE-522 regimen mandates adjuvant pembrolizumab for up to 9 cycles post-surgery, regardless of the pathologic complete response (pCR) achieved. Based on the trial's subgroup analyses of event-free survival (EFS) stratified by pCR status, is the adjuvant phase of pembrolizumab truly necessary for those who achieve pCR, and how does this complicate de-escalation trial designs?

Key Response

Explores a major controversy in breast oncology. Patients achieving pCR generally have excellent outcomes regardless of adjuvant therapy. The trial was not powered to isolate the benefit of the adjuvant pembrolizumab phase from the neoadjuvant phase. Fellows must balance the toxicity of continuing immunotherapy against the lack of definitive data, highlighting the rationale for ongoing de-escalation trials (like OptimICE-PCR) which aim to omit adjuvant pembrolizumab in pCR patients.

Attending
Attending

Given that KEYNOTE-522 showed a significant EFS benefit but also introduced life-altering, permanent toxicities (e.g., adrenal insufficiency, type 1 diabetes), how do you frame the risk-benefit shared decision-making discussion for a marginal-risk patient (e.g., T1c N0 TNBC) who might have had an excellent prognosis with standard chemotherapy alone?

Key Response

Focuses on clinical wisdom and practice-changing implications. While the regimen is the standard of care for stage II/III TNBC, attendings must weigh the absolute EFS benefit (which varies by baseline risk) against a ~1-2% risk of permanent, lifelong endocrine morbidity. Navigating this requires nuanced communication with young women about long-term survivorship versus maximal recurrence risk reduction.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

KEYNOTE-522 utilized a dual primary endpoint design (pCR and EFS) with complex alpha-spending boundaries. From a biostatistical perspective, what are the methodological challenges of using pCR as an early trial-level surrogate for EFS in early-stage TNBC, and did the magnitude of delta-pCR in this trial mathematically predict the magnitude of the EFS benefit based on historical meta-analyses?

Key Response

Critiques trial methodology and statistical approaches. The FDA frequently grants accelerated approval based on pCR, but trial-level surrogacy of pCR for EFS/OS is heavily debated (e.g., the Cortazar meta-analysis). Evaluating whether the robust 13.6% pCR increase cleanly translated to the observed EFS hazard ratio of 0.63 requires deep understanding of event-driven trial designs, survival kinetics, and the potential non-proportionality of hazards in immunotherapy trials.

Journal Editor
Journal Editor

During the long enrollment and follow-up period of KEYNOTE-522, the standard of care for early-stage TNBC patients with residual disease post-neoadjuvant chemotherapy evolved to include adjuvant capecitabine based on the CREATE-X trial. As an editor, how do you evaluate the validity of the trial's EFS endpoint given that patients in the placebo arm with residual disease did not uniformly receive capecitabine?

Key Response

Addresses a classic critical appraisal issue: evolving control arms. Because post-neoadjuvant capecitabine became a standard for non-pCR patients during the study, the lack of its uniform use in the control arm could theoretically exaggerate the EFS benefit seen in the pembrolizumab arm. A tough reviewer would scrutinize subsequent therapy tables and demand subgroup analyses adjusting for adjuvant capecitabine use to ensure the experimental arm's superiority is genuine.

Guideline Committee
Guideline Committee

Following KEYNOTE-522, NCCN and ASCO guidelines adopted neoadjuvant pembrolizumab plus chemotherapy as standard for high-risk early-stage TNBC. Based on the trial's subgroup data, should guidelines mandate PD-L1 testing (e.g., CPS score) as a prerequisite for this early-stage regimen, as is strictly required by guidelines for the metastatic TNBC setting (e.g., KEYNOTE-355)?

Key Response

Evaluates how evidence directly informs specific guideline parameters. In metastatic TNBC (KEYNOTE-355), pembrolizumab is only recommended for PD-L1 CPS >= 10. However, in KEYNOTE-522, the addition of pembrolizumab improved pCR and EFS independent of PD-L1 expression. Therefore, guidelines currently recommend against requiring PD-L1 testing to initiate the KEYNOTE-522 regimen, demonstrating how disease stage and primary vs. metastatic tumor biology drastically alter biomarker guideline recommendations.

Clinical Landscape

Noteworthy Related Trials

2019

GeparNuevo

n = 174 · Ann Oncol

Tested

Durvalumab plus neoadjuvant chemotherapy

Population

Patients with early-stage TNBC

Comparator

Placebo plus neoadjuvant chemotherapy

Endpoint

Pathological complete response (pCR)

Key result: The addition of durvalumab to neoadjuvant chemotherapy increased pCR rates, particularly in patients who received a window-phase of durvalumab alone before chemotherapy.
2020

IMpassion031

n = 333 · Lancet

Tested

Atezolizumab plus neoadjuvant chemotherapy

Population

Patients with early-stage TNBC

Comparator

Placebo plus neoadjuvant chemotherapy

Endpoint

Pathological complete response (pCR)

Key result: Atezolizumab combined with neoadjuvant chemotherapy significantly increased pCR rates compared to placebo.
2020

KEYNOTE-355

n = 847 · Lancet

Tested

Pembrolizumab plus chemotherapy

Population

Patients with previously untreated locally recurrent inoperable or metastatic TNBC

Comparator

Placebo plus chemotherapy

Endpoint

Progression-free survival and overall survival

Key result: Pembrolizumab plus chemotherapy significantly improved progression-free survival in patients with metastatic TNBC and a combined positive score of 10 or more.

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