Adjuvant treatment for acral melanoma

We read with great interest the article by Sanogo et al., reporting a propensity score–matched analysis evaluating adjuvant immune checkpoint inhibitors (ICI) in resected stage IIB–IV acral melanoma.1 After matching, 64 patients receiving adjuvant ICI were compared with 64 patients without adjuvant ICI. No significant improvement in relapse-free or distant metastasis-free survival was observed, while Grade 3–4 immune-related adverse events occurred in approximately 19% of treated patients. These findings add to inconsistent observational evidence. Koizumi et al. reported no survival benefit of adjuvant anti–programmed death-1 (PD-1) therapy in sole melanoma,2 whereas Jacques et al. described improved outcomes compared with historical observation controls.3 A recent multicentre study focusing on nail apparatus melanoma similarly did not demonstrate a clear survival benefit of adjuvant PD-1 therapy compared with observation.4 Differences among studies likely reflect variation in study design, control definitions and handling of temporal confounding. Heterogeneity also arises from the composition of the adjuvant ICI cohort. In the study by Sanogo et al., the adjuvant group included anti-PD-1 monotherapy but also ipilimumab monotherapy and combined ipilimumab plus nivolumab. In clinical practice, adjuvant therapy is predominantly anti-PD-1 monotherapy, whereas ipilimumab-containing regimens are rarely used. Pooling biologically distinct immunotherapy regimens within a single exposure category may therefore complicate interpretation of treatment effects. Although the authors attempted to address temporal bias by incorporating treatment timing into the propensity score model, structural concerns remain. Adjuvant PD-1 therapy became available in Europe only in 2018, whereas the registry includes patients from 2012 onward. Consequently, the comparator group inevitably includes more patients treated in earlier therapeutic eras. Matching cannot fully eliminate differences in treatment availability, staging practice and follow-up intensity across eras. Post-matching balance diagnostics are incompletely reported. Standardized mean differences, essential for assessing covariate balance after propensity score matching, are not presented. Moreover, Table 1 raises concerns regarding internal consistency: Stage IIIA appears absent in the adjuvant ICI cohort before matching but present afterward. Under conventional matching procedures in which the treated cohort is fixed, such a change would be unexpected and may represent a tabulation or reporting error requiring clarification. Eligibility criteria also warrant clarification. The Methods section specifies inclusion of Stage IIB–IV melanoma, yet Table 1 lists Stage IA, IB and IIA cases. If early-stage patients were included, their lower recurrence risk could dilute treatment effects. The strength of this study is the incorporation of common melanoma driver mutations within the propensity score framework. Alterations such as BRAF, NRAS and KIT are key biological features of acral melanoma and may influence prognosis and treatment decisions. This genomic information is not routinely available in many registries and, in some healthcare settings including Japan, testing beyond BRAF is not commonly performed in routine practice. Interpretation of survival analyses would benefit from greater transparency regarding follow-up. Kaplan–Meier curves are presented without numbers at risk, limiting assessment of effective sample size over time in this small cohort. The specification of the propensity score model also merits clarification. The model includes AJCC stage together with pathological components such as Breslow thickness, ulceration and mitotic index. Because these variables are structurally related, simultaneous inclusion may introduce redundancy and potentially destabilize estimation. Finally, statistical power is limited. With only 64 patients per group and relatively few events, the study may be underpowered to detect modest treatment differences. The study provides real-world insight into acral melanoma, but temporal heterogeneity, incomplete balance reporting and limited statistical precision complicate causal interpretation. Further studies specifically designed for this biologically distinct melanoma subtype are needed to clarify the role of adjuvant PD-1 therapy. None. Dr. Yasuhiro Nakamura has received research, speaking and/or consulting support from Alexion Pharma, BMS, Dai-ichi Sankyo, HUYA Bioscience International, Kyowa Kirin, LEO Pharma, Maruho, MSD, Novartis, Ono Pharmaceutical, Pierre Fabre, Sanofi, Sun Pharma and Tanabe-Mitsubishi Pharma. Not applicable. Not applicable. Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.