Neoadjuvant/Adjuvant Pembrolizumab Yields EFS Benefits Regardless of Surgical Outcomes in HNSCC

The addition of neoadjuvant and adjuvant pembrolizumab (Keytruda) to standard-of-care (SOC) surgery followed by (chemo)radiotherapy improved event-free survival (EFS) irrespective of resection status, extranodal extension status, or surgical margin status vs SOC alone in patients with resectable, locally advanced head and neck squamous cell carcinoma (HNSCC), according to findings from an exploratory analysis of the phase 3 KEYNOTE-689 trial (NCT03765918) presented at the 2026 ASCO Annual Meeting.¹

Among patients who underwent surgery with R0 resection and received pembrolizumab plus SOC (n = 287), the median EFS was not reached (NR; 95% CI, 41.1 months-NR) vs 40.2 months (95% CI, 28.8-NR) for those who received surgery followed by SOC alone (n = 261; HR, 0.67; 95% CI, 0.50-0.88). Moreover, in the population of patients with a PD-L1 combined positive score (CPS) of at least 1 and R0 resection, the median EFS was NR (95% CI, 50.3 months-NR) in the pembrolizumab arm (n = 275) compared with 35.3 months (95% CI, 26.9-NR) in the SOC arm (n = 249; HR, 0.62; 95% CI, 0.47-0.83).

For patients with extranodal extension, the median EFS was 25.2 months (95% CI, 10.9-51.8) in the pembrolizumab arm (n = 82) vs 12.6 months (95% CI, 9.9-22.7) in the SOC arm (n = 112; HR, 0.77; 95% CI, 0.52-1.15). Patients without extranodal extension experienced a median EFS that was NR (95% CI, 50.3 months-NR) with pembrolizumab plus SOC (n = 228) vs 57.0 months (95% CI, 35.8-NR) with SOC alone (n = 188; HR, 0.74; 95% CI, 0.53-1.05).

Among patients with positive surgical margins, the median EFS was 50.3 months (95% CI, 25.5-NR) with pembrolizumab plus SOC (n = 64) compared with 13.8 months (95% CI, 10.1-35.8) with SOC alone (n = 92; HR, 0.58; 95% CI, 0.36-0.93). The median EFS for patients with negative surgical margins were NR (95% CI, 40.5 months-NR) and 51.5 months (95% CI, 31.3-NR) in the pembrolizumab (n = 249) and SOC (n = 209) arms, respectively (HR, 0.78; 95% CI, 0.57-1.07).

“[These] results further support the addition of neoadjuvant and adjuvant pembrolizumab to the SOC for resectable, locally advanced HNSCC,” Douglas R. Adkins, MD, and coauthors wrote in a poster presentation of the data.

Adkins is a professor of medicine in the Division of Oncology at Washington University in St. Louis, Missouri.

How was KEYNOTE-689 designed?

The randomized, open-label study enrolled patients with resectable, locally advanced HNSCC of the larynx, hypopharynx, oral cavity (stage III/IVA), or oropharynx (p16-negative stage III/IVA, or p16-positive stage III T4/N0-2). Patients also needed to have an ECOG performance status of 0 or 1 and evaluable tumor burden based on RECIST 1.1 criteria.^1,2

If patients had cancer located outside the oropharynx, larynx, and hypopharynx or oral cavity, had received prior radiotherapy or systemic anticancer therapy, or had previous allogeneic tissue/solid organ transplant, they were not included in the trial.²

Patients in the neoadjuvant pembrolizumab arm (n = 363) received neoadjuvant pembrolizumab for 2 cycles followed by surgery, then 3 cycles of adjuvant pembrolizumab plus radiotherapy with or without cisplatin, and finally adjuvant pembrolizumab for 12 cycles.¹

Patients in the SOC arm (n = 351) underwent surgery followed by radiotherapy with or without cisplatin.

EFS by R0 resection status, presence or absence of extranodal extension, and positive or negative surgical margins were the end points of the analysis.

The baseline characteristics of all patients who underwent surgery (pembrolizumab arm, n = 322; SOC arm, n = 308) revealed that most patients were male (78.3%; 78.6%), had an ECOG performance status of 0 (55.9%; 60.4%), and were human papillomavirus negative (97.2%; 96.4%). Most patients were also current or former smokers (80.7%; 78.6%) and had engaged in alcohol use (69.6%; 70.8%). Primary tumor sites for patients broke down as located in the oral cavity (59.9%; 61.7%), larynx (22.4%; 21.1%), oropharynx (9.9%; 9.4%), and hypopharynx (7.8%; 7.8%). Most patients had stage IVA cancer at baseline (73.9%; 74.0%) whereas fewer patients had stage III (25.5%; 26.0%) or stage IVB (0.6%; 0%) cancer.

What did surgical and pathological outcomes show in KEYNOTE-689?

A higher proportion of patients in the pembrolizumab arm achieved R0 resection (89.1%) compared with in the SOC arm (84.7%). Patients in the pembrolizumab arm also had lower rates of postsurgical high-risk pathological features compared with the SOC arm, including extranodal extension (25.5%; 36.4%) and positive surgical margins (19.9%; 29.9%), along with a higher rate of overall downstage reassignment (45.7%; 36.4%). A central pathological risk assessment classified 36.6% of patients in the pembrolizumab arm as having high-risk disease vs 50.3% of those in the SOC arm.

What were the additional data for pembrolizumab plus SOC in resectable, locally advanced HSNCC?

Among patients with extranodal extension and a PD-L1 CPS of at least 1, the median EFS was 25.2 months (95% CI, 11.1-NR) in the pembrolizumab arm (n = 76) vs 12.6 months (95% CI, 9.9-22.7) in the SOC arm (n = 109; HR, 0.74; 95% CI, 0.49-1.12). Among patients with extranodal extension and a PD-L1 CPS of at least 1, the median EFS values were NR (95% CI, 44.7 months-NR) and 57.0 months (95% CI, 35.3-NR) in the pembrolizumab (n = 220) and SOC (n = 178) arms, respectively (HR, 0.71; 95% CI, 0.51-1.01).

In the population of patients with a PD-L1 CPS of at least 1 with positive margins, the median EFS was 50.3 months (95% CI, 20.6-NR) with pembrolizumab plus SOC (n = 58) compared with 13.8 months (95% CI, 9.9-26.8) with SOC alone (n = 89; HR, 0.60; 95% CI, 0.37-0.98). Additionally, patients in the same population but with negative margins had median EFS values that were NR (95% CI, 40.8 months-NR) and 50.1 months (95% CI, 30.9-NR) in the pembrolizumab (n = 242) and SOC (n = 198) arms, respectively (HR, 0.72; 95% CI, 0.53-1.00).

References

1. Adkins D, Haddad RI, Tao Y, et al. Neoadjuvant and adjuvant pembrolizumab plus standard of care for resectable locally advanced head and neck squamous cell carcinoma: efficacy by surgical outcomes in the phase 3 KEYNOTE-689 trial. J Clin Oncol. 2026:44(suppl 16);abstr 6057. doi:10.1200/JCO.2026.44.16_suppl.6057
2. Study of pembrolizumab given prior to surgery and in combination with radiotherapy given post-surgery for advanced head and neck squamous cell carcinoma (MK-3475-689). ClinicalTrials.gov. Updated May 12, 2026. Accessed July 7, 2026. https://clinicaltrials.gov/study/NCT03765918