Updated outcomes from STAMP: Surgically treated adjuvant Merkel cell carcinoma with pembrolizumab, a phase III trial—ECOG-ACRIN EA6174.

Abstract

LBA9505 Background: EA6174 (STAMP) is the first randomized phase III trial of adjuvant pembrolizumab (pembro) in resected Merkel cell carcinoma (MCC), a rare, aggressive skin cancer with high relapse risk after surgery. This abstract presents new analyses of MCC-specific outcomes and radiation therapy (RT) effects. Methods: Patients (pts) with resected stage I without a sentinel lymph node biopsy to IIIB MCC were randomized 1:1 to pembro 200 mg IV q3 weeks ×17 doses (n = 147) or standard of care (SOC) (n = 146). Randomization was stratified by stage (I/II vs III) and intended RT. Co-primary endpoints were relapse-free survival (RFS) and overall survival (OS). MCC-specific outcomes were captured as relapse due to recurrence or death due to any cause (RFS_a), and MCC-specific recurrence and death (RFS_b). Distant metastasis-free survival (DMFS) was defined as time to distant recurrence or death from MCC. 280 pts provided 80% power with one-sided type 1 error rate of 0.05 to evaluate the hazard ratio (HR) of 0.56 hierarchically (RFS before OS). One-sided p-values from the log-rank test are reported in the analysis (data as of 2/19/26, median follow-up 47 months). Results: 293 pts enrolled (67.9% male; 84.3% stage III, median age 69, > 50% accrued during the COVID-19 pandemic). Overall RFS (RFS_a) was numerically improved with pembro but not statistically significant between arms (p = 0.10). DMFS was improved in pts treated with pembro (p = .05). Significant improvement was observed in MCC-specific outcomes (RFS_b) in pts treated with pembro, with HR = 0.66 [90%CI (0.45,0.96)], resulting in 1-year and 2-year RFS_b rates of 84% [90%CI(78,88)] and 77% [90%CI(70,82)] in the pembro arm, compared with 73% [90%CI( 67,79)] and 67% [90%CI (60,74)], in the SOC arm (p = 0.032). The impact of pembro on RFS_b was maintained (p = 0.039) among the subset that received any RT (n = 131 pembro, n = 123 SOC). For pts treated concurrently with RT, pembro did not affect outcomes; however, for pts treated sequentially with RT prior to randomization (n = 68 pembro and n = 65 SOC), pembro improved RFS_a (P = 0.014), RFS_b (p = 0.009), and DMFS (p = 0.004). Conclusions: EA6174 demonstrates that pts treated with pembro experienced significantly fewer MCC-related events than pts who received SOC and that adjuvant pembro improves DMFS. RFSa may have been influenced by the average age of enrolled pts and by disruptions and morbidity associated with patient enrollment during the COVID-19 pandemic; as such RFSb may be a more accurate indicator of pembro efficacy. These findings support a meaningful impact of pembro on MCC-specific outcomes, particularly for distant relapse. Pembro may be of particular benefit when RT is delivered sequentially rather than concurrently with pembro, similar to other studies of RT and immunotherapy in solid tumors. Mature OS and MCC-specific survival data will further guide use of adjuvant immunotherapy for MCC. Clinical trial information: NCT03712605 .