Document resource
Background Lymphocyte-activation gene 3 (LAG-3) is often coexpressed with PD-L1 in solid tumors, including cSCC. Dual LAG-3 and PD-1 inhibition has demonstrated antitumor activity and manageable safety in resectable and unresectable melanoma. KeyForm-010, a basket study ( NCT06036836), evaluated the favezelimab (anti-LAG-3 antibody) and pembrolizumab (anti-PD-1 antibody) fixed-dose coformulation in solid tumors. We report results for cohort A, a phase 2 study of neoadjuvant favezelimab/pembrolizumab (coformulation) versus pembrolizumab for resectable cSCC.Methods Eligible participants were ≥18 years-of-age with histologically confirmed stage II–IV resectable cSCC (without M1; head/neck tumor staging per AJCC 8th ed; other tumor sites per UICC 8th ed) amenable to curative intent surgery and no prior systemic therapy or radiotherapy to the index lesion. Participants were randomized 1:1 (stratification: non-nodal vs nodal disease and head/neck vs other tumor site) to neoadjuvant favezelimab 800 mg/pembrolizumab 200 mg or pembrolizumab 200 mg Q3W for ≤3 cycles, followed by surgery, then adjuvant favezelimab/pembrolizumab or pembrolizumab Q3W as allocated. Total treatment duration was ≤17 cycles. Primary endpoint was major clinical benefit (composite of major pathologic response [mPR], pathologic complete response [pCR], or clinical complete response [cCR] without surgery). Secondary endpoints were pCR and mPR per local pathology review, ORR per investigator assessment before surgical resection, and safety.Results 83 participants were randomized to receive favezelimab/pembrolizumab (n=41) versus pembrolizumab (n=42); 31/41 (76%) versus 31/42 (74%) had head/neck tumors. Median (range) time from randomization to data cutoff (June 10, 2025) was 10.7 (5.7–19.8) months versus 10.3 (5.9–19.8) months, respectively. Major clinical benefit rate was high in both groups (59% vs 64%); pCR rates were ~40% in each group ( table 1). Among the high proportion of participants who did not go to surgery (~30%, each group), 38% (5/13) versus 50% (7/14) had cCR, 80% (4/5) versus 100% (7/7) of whom had negative biopsies. Treatment-related AEs occurred more frequently with neoadjuvant favezelimab/pembrolizumab versus pembrolizumab (73% vs 48%; grade ≥3, 34% vs 2%; no grade 5 in either group), as did immune-mediated AEs and infusion reactions (46% vs 2%; table 2).Conclusions This is the largest study to date demonstrating high clinical and pathological response with neoadjuvant pembrolizumab (alone or in coformulation) in resectable cSCC. Favezelimab/pembrolizumab had higher AE incidence without additional efficacy over pembrolizumab alone. Safety results were overall consistent with the known safety profile of both drugs. Findings suggest pembrolizumab could introduce a surgery-sparing neoadjuvant treatment for resectable cSCC, potentially salvaging critical head/neck anatomic sites.Trial Registration Clinicaltrials.gov, NCT06036836Ethics Approval The study was conducted in accordance with principles of Good Clinical Practice and was approved by the appropriate institutional review boards and regulatory agencies. All participants provided written informed consent before enrollment. Abstract 1305 Table 1Efficacy results amPR defined as presence of ≤10% viable tumor in surgical resection specimen per central review, therefore mPR rate includes participants with pCR. bPercentages calculated using number of participants in row heading as denominator. cThree participants became inoperable due to disease progression following neoadjuvant therapy: favezelimab/pembrolizumab, n=2; pembrolizumab, n=1. dThere were no recurrences among participants with negative biopsies. eAssessed per RECIST version 1.1 by investigator. fParticipants who had ≥1 postbaseline imaging assessment, none of which were evaluable per RECIST version 1.1 by investigator. gParticipants for whom no postbaseline tumor assessment was performed.Abstract 1305 Table 2Summary of AEs during neoadjuvant phaseaData are n (%). aIncludes AEs that occurred from the first to the last neoadjuvant study treatment, definitive surgery, or radiation therapy before the first adjuvant treatment or, if no adjuvant treatment, up to 30 days after the last neoadjuvant study treatment, definitive surgery, or radiation therapy. bAEs are based on a list of terms specified by the sponsor and considered regardless of attribution by investigators. Related terms are included in the preferred terms listed. cThere were no grade 4 or 5 immune-mediated AEs or infusion reactions.