How KEYNOTE-522 Permanently Changed Triple-Negative Breast Cancer Care

Nearly 8 years out, KEYNOTE-522's final data confirm that adding pembrolizumab to neoadjuvant chemotherapy doesn't just boost short-term response in high-risk early triple-negative breast cancer (TNBC) but also durably cuts the risk of recurrence and death, cementing a new curative-intent standard of care.

The landmark phase 3 KEYNOTE-522 trial (NCT03036488) established a new standard of care for patients with previously untreated, high-risk, early-stage TNBC.¹ The study evaluated a perioperative approach by combining pembrolizumab (Keytruda; an anti–PD-1 therapy) with neoadjuvant chemotherapy, followed by adjuvant pembrolizumab monotherapy after definitive surgery.

A total of 1174 patients were randomized in a 2:1 ratio to receive either pembrolizumab (n = 784) or a placebo (n = 390). In the neoadjuvant phase, both arms received a robust backbone consisting of 4 cycles of paclitaxel plus carboplatin, followed by 4 cycles of an anthracycline-based regimen (doxorubicin or epirubicin plus cyclophosphamide). Following surgical resection, patients continued their assigned treatment (adjuvant pembrolizumab or placebo) for an additional 9 cycles to complete approximately 1 year of immunotherapy. The trial's dual primary end points were pathologic complete response (pCR)—defined as the absence of invasive cancer in the breast and lymph nodes (ypT0/Tis ypN0)—and event-free survival (EFS). Overall survival (OS) served as the crucial secondary end point.

Final Long-Term Analysis Results (7.8-Year Follow-up)

At the definitive data cutoff on October 14, 2025, the trial reached a median follow-up of 93.8 months (approximately 7.8 years), providing highly mature survival metrics.² The final analysis confirmed that the clinical benefits of adding pembrolizumab are both durable and statistically profound:

• Event-Free Survival: The 7-year EFS rate reached 78.3% in the pembrolizumab cohort compared to 69.8% in the placebo group. This represents a 32% reduction in the risk of disease progression, local/distant recurrence, second primary cancer, or death (HR, 0.68; 95% CI, 0.54-0.86).
• Overall Survival: The 7-year OS rate was 85.1% for the pembrolizumab group vs 77.2% for the placebo group. This demonstrates a clear, long-term survival advantage, reducing the risk of death by 36% (HR, 0.64; 95% CI, 0.49-0.85).

The survival advantages spanned across most pre-specified subgroups, demonstrating that therapeutic efficacy is independent of nodal involvement, clinical disease stage, or PD-L1 expression levels.

Regarding safety, the addition of pembrolizumab slightly increased severe toxicity, with grade ≥3 treatment-related adverse events (AEs) occurring in 77.1% of the pembrolizumab group compared WITH 73.3% of the placebo group. Treatment-related fatalities were rare (0.5% vs 0.3%). However, immune-mediated AEs of any grade were substantially higher in the pembrolizumab arm (35.0% vs 13.1%), reflecting the expected safety profile of checkpoint inhibitors.

Historical Context and Evolution of KEYNOTE-522 Data

The final 7.8-year results are the culmination of a series of interim analyses that progressively shaped international clinical guidelines. Tracking the historical data underscores how early surrogate endpoints translated into long-term survival benefits.

1. Initial Pathologic Complete Response (pCR) Discoveries

In the trial's first formal publication, the addition of pembrolizumab demonstrated a statistically significant and unprecedented surge in pCR. The early evaluation revealed a pCR rate of 64.8% in the chemoimmunotherapy arm compared with 51.2% in the chemotherapy-alone control arm—a definitive absolute improvement of 13.6 percentage points.³ Later evaluations settled the final trial-wide pCR comparison at 63.0% vs 55.6%.

2. The 3-Year (Fourth Interim) Analysis

At a median follow-up of 39.1 months, the trial proved that improvements in pCR directly translated to a reduction in clinical recurrences. The 3-year EFS rate was 84.5% for the pembrolizumab cohort compared with 76.8% for the placebo cohort (HR, 0.63; 95% CI, 0.48–0.82), which triggered full FDA approval in July 2021.

Crucially, exploratory analyses highlighted the prognostic importance of residual disease via the residual cancer burden (RCB) index:

• With pCR (RCB-0): Patients achieving a pCR had excellent 3-year outcomes across both arms (94.4% with pembrolizumab vs. 92.5% with placebo).
• Without pCR (Residual Disease): For patients who failed to achieve a pCR, pembrolizumab acted as a vital buffer. The 3-year EFS for non-responders was 67.4% in the pembrolizumab arm versus a dismal 56.8% in the placebo arm, proving that perioperative immunotherapy remains highly beneficial even when residual tumor cells are found at surgery.

3. The 5-Year (75-Month) Analysis

Presented at the 2024 ESMO Congress, the 5-year data offered the first definitive look at overall survival. At a median follow-up of 75.1 months, the 5-year EFS rate maintained a stark gap at 81.2% vs 72.2% (HR, 0.65). More importantly, it established a statistically significant OS advantage, showing a 5-year survival rate of 86.6% for pembrolizumab vs 81.7% for the control group (HR, 0.66; P =.00150).

This trajectory confirms that the absolute benefit of pembrolizumab remains stable over nearly 8 years, cementing the perioperative regimen's curative potential in high-risk early TNBC.

REFERENCES

1. Hu H, Kaklamani V. Updates on the preoperative immunotherapy for triple-negative breast cancer. Transl Breast Cancer Res. 2023 Apr 30;4:17. doi: 10.21037/tbcr-23-16. PMID: 38751476; PMCID: PMC11092992.

2. Schmid P, Cortes J, Dent R, et al. Neoadjuvant pembrolizumab or placebo plus chemotherapy followed by adjuvant pembrolizumab or placebo for high-risk early-stage TNBC: final analysis results from the phase 3 KEYNOTE-522 study. J Clin Oncol. 2026;44(suppl 16):507. doi:10.1200/JCO.2026.44.16_suppl.507
3. Schmid P, Cortes J, Pusztai L, et al. Pembrolizumab for Early Triple-Negative Breast Cancer. N Engl J Med. 2020 Feb 27;382(9):810-821. doi: 10.1056/NEJMoa1910549. PMID: 32101663.