Can you summarize the phase III KEYNOTE‑689 trial results of peri‑operative pembrolizumab (programmed death‑1 inhibitor) in patients with resectable locally advanced head‑and‑neck squamous cell carcinoma?

KEYNOTE-689 Trial Summary: Perioperative Pembrolizumab in Resectable Locally Advanced Head and Neck Cancer

The KEYNOTE-689 phase III trial demonstrated that adding perioperative pembrolizumab (2 cycles neoadjuvant + 15 cycles adjuvant) to standard surgery and adjuvant therapy significantly improved event-free survival in patients with resectable locally advanced head and neck squamous cell carcinoma, with a 27% reduction in risk of progression, recurrence, or death in the overall population. 1

Trial Design and Population

This open-label, randomized phase III trial enrolled 714 patients with locally advanced HNSCC:

• Pembrolizumab arm (n=363): 2 cycles neoadjuvant pembrolizumab 200 mg every 3 weeks → surgery → adjuvant radiotherapy ± cisplatin → 15 cycles adjuvant pembrolizumab
• Control arm (n=351): Surgery → adjuvant radiotherapy ± cisplatin alone

The trial stratified patients by PD-L1 expression using Combined Positive Score (CPS), with sequential analysis planned for CPS ≥10, CPS ≥1, and all patients. 1

Key Efficacy Results (First Interim Analysis at 38.3 Months Median Follow-Up)

Event-Free Survival at 36 Months:

CPS-10 Population (n=465):

• Pembrolizumab: 59.8%
• Control: 45.9%
• Hazard ratio: 0.66 (95% CI: 0.49-0.88; P=0.004)

CPS-1 Population (n=682):

• Pembrolizumab: 58.2%
• Control: 44.9%
• Hazard ratio: 0.70 (95% CI: 0.55-0.89; P=0.003)

Total Population (n=714):

• Pembrolizumab: 57.6%
• Control: 46.4%
• Hazard ratio: 0.73 (95% CI: 0.58-0.92; P=0.008) 1

The benefit was consistent across all PD-L1 expression levels, though the magnitude of benefit appeared numerically greater in higher PD-L1 expressors.

Surgical Feasibility

Critically, neoadjuvant pembrolizumab did not compromise surgical outcomes:

• Surgery completion rate: ~88% in both arms
• This addresses a key concern about neoadjuvant immunotherapy potentially delaying or preventing definitive surgery 1

Safety Profile

Grade ≥3 Treatment-Related Adverse Events:

• Pembrolizumab arm: 44.6%
• Control arm: 42.9%
• Treatment-related deaths: 1.1% vs 0.3%

Immune-Mediated Adverse Events (Grade ≥3): 10.0% in pembrolizumab arm

The safety profile was manageable and consistent with known pembrolizumab toxicity, with no unexpected signals. 1

Clinical Context and Implications

Comparison with Current Guidelines:

This trial represents a paradigm shift. Current guidelines 2 recommend immunotherapy only for recurrent/metastatic HNSCC with PD-L1 expression (CPS ≥1), not for locally advanced resectable disease. The 2023 ASCO guideline 3 explicitly states that "the role of ICIs in locoregionally advanced disease is not yet established" and notes that multiple trials found no benefit when adding ICIs to definitive chemoradiotherapy.

Why KEYNOTE-689 Succeeded Where Others Failed:

1. Perioperative timing: Combined neoadjuvant + adjuvant approach, rather than concurrent with chemoradiation
2. Surgical population: Resectable disease treated with surgery as primary modality
3. Early immune priming: Neoadjuvant dosing may activate immune response before tumor removal
4. Extended adjuvant therapy: 15 cycles of maintenance immunotherapy

Pattern of Benefit:

The trial showed reduction in distant recurrence 4, suggesting systemic immune activation that prevents metastatic spread—a critical outcome for long-term survival.

Comparison with NIVOPOSTOP Trial

A parallel trial (NIVOPOSTOP) used adjuvant-only nivolumab with postoperative chemoradiotherapy in high-risk patients, also showing benefit (HR 0.76; 95% CI: 0.60-0.98). 4 Key differences:

• KEYNOTE-689: Broader perioperative strategy, lower treatment-related serious adverse events
• NIVOPOSTOP: Focused adjuvant approach in high-risk only, higher compliance

Both trials validate perioperative immunotherapy but with different strategic approaches. 4

Practical Considerations

Patient Selection:

• Resectable locally advanced HNSCC (stage III-IVA typically)
• Benefit seen across all PD-L1 levels, though numerically greater in CPS ≥10
• No autoimmune disease or immunosuppression requirements

Treatment Algorithm:

1. Neoadjuvant phase: Pembrolizumab 200 mg IV every 3 weeks × 2 cycles
2. Surgery: Proceed at ~6 weeks after last neoadjuvant dose
3. Adjuvant therapy: Standard risk-adapted (chemo)radiotherapy
4. Adjuvant immunotherapy: Pembrolizumab 200 mg IV every 3 weeks × 15 cycles

Caveats:

• Overall survival data remain immature—event-free survival is a validated surrogate but not definitive
• Cost-effectiveness analysis pending
• Optimal patient selection criteria need refinement
• Response-adapted de-escalation strategies not yet defined 4

Future Directions

The field is moving toward:

• Refining biomarker-driven patient selection beyond PD-L1
• Optimizing neoadjuvant regimen intensity and duration
• Implementing response-adapted de-escalation to reduce treatment burden
• Defining role in HPV-positive vs HPV-negative disease separately 4

This represents the first major advance in locally advanced HNSCC treatment in two decades, establishing perioperative immunotherapy as a new standard of care for resectable disease. 4, 5