A phase II open-label, single-arm study to evaluate the efficacy of pembrolizumab for leukoplakia.

10544 Background: Oral leukoplakia and erythroleukoplakia represent high-risk premalignant lesions of the oral cavity, with malignant transformation to head and neck squamous cell carcinoma reported in up to 36% of cases. Therapies are limited, with surgical resection remaining the standard of care despite high recurrence rates. Immune checkpoint inhibitors (ICIs) have emerged as a potential preventive strategy. We evaluated whether the efficacy of the PD-1 inhibitor pembrolizumab may serve as a preventive strategy to reduce progression to invasive carcinoma in patients (pts) with these premalignant lesions. Methods: This nonrandomized, open-label, multi-center phase II clinical trial enrolled pts between June 2019 and December 2025. Eligible pts had oral leukoplakia, erythroleukoplakia, or proliferative verrucous leukoplakia with moderate dysplasia, severe dysplasia, or carcinoma in situ. Patients with previously treated squamous cell carcinoma (SCC) were included. Pts received pembrolizumab 200 mg intravenously every 3 weeks for 6 months. The primary endpoint was clinical response at 6 months, defined as the proportion of patients achieving a complete response (CR) or partial response (PR). CR was defined as complete lesion resolution on visual inspection sustained for ≥4 weeks. PR was defined as a ≥50% reduction in the product of the two largest perpendicular dimensions of a single lesion or the sum of all measurable lesions. The secondary objectives included clinical response rate at 9 and 12 months, and toxicity. Results: Sixteen pts received at least one dose of pembrolizumab. The median age was 63.5 years (range, 33–82), and 44% were female. Three patients had previously treated SCC. Baseline pathology demonstrated moderate dysplasia in 31% and severe or high-grade dysplasia in 69% of patients. At 6 months, the clinical response rate was 44% (7/16; 95% CI, 23–67%), including CR in 19% (3/16) and PR in 25% (4/16) of pts. Progressive disease occurred in 38% (6/16) of patients. Clinical response rates were 31% at 9 months (5/16) and 19% at 12 months (3/16). Median follow-up was 9 months. Treatment-related adverse events (TRAE) occurred in 69% of pts, most commonly fatigue (31%), mucositis/oral discomfort (31%), hypothyroidism (25%), and gastrointestinal immune-related events, including diarrhea or colitis (25%). Grade ≥3 treatment-related adverse events occurred in 19%, with no grade 4–5 events or treatment-related deaths. One pt discontinued due to TRAE. Conclusions: These findings suggest that PD-1 blockade may represent a feasible preventive approach for high-risk oral premalignant lesions and warrants further investigation as a preventive option. Biomarker evaluation of pre-treatment and on-treatment biopsies are ongoing to inform predictors of response. Clinical trial information: NCT03603223 .