Comment on ‘adjuvant anti‐PD‐1 antibody versus observation for resected nail apparatus melanoma’

We read with great interest the retrospective study by Komori et al.1 investigating the efficacy of adjuvant anti-programmed death-1 antibody (adj-PD-1) therapy versus observation (OBS) in patients with resected Stage IIB–III nail apparatus melanoma (NAM). As a rare and biologically distinct subtype of melanoma, NAM has long lacked targeted therapeutic evidence, and this study fills a critical gap by providing the first direct comparative data on adj-PD-1 use in this population. While the study design incorporates propensity score matching (PSM) to mitigate selection bias, several methodological considerations merit discussion to enhance result validity. A key limitation is sample generalizability. The study exclusively included Japanese patients, raising questions about applicability to ethnically diverse populations. NAM, which constitutes 0.7%–3.5% of all cutaneous melanomas, shows marked ethnic variation in its prevalence, accounting for 0.18%–2.8% of cases in Europeans, 10%–23% in Asians and 25% in African Americans.2 NAM harbours unique genetic features (e.g. copy-number alterations, low tumour mutational burden)3, 4 distinct from acral melanoma (AM) and non-acral cutaneous melanoma, and previous research on AM has highlighted racial variations in PD-1 responsiveness.5, 6 This suggests ethnic differences in immune microenvironment or genomics may further modify adj-PD-1 efficacy, limiting extrapolation to global clinical practice. The retrospective design introduces inherent biases. While PSM adjusted for age, sex and stage, critical confounders (comorbidities, surgical margin status, lymph node dissection extent) and mutational profiles (BRAF, NRAS) were incomplete. These factors influence melanoma recurrence/survival and could confound efficacy comparisons. Integrating mutational status into subgroup analyses might identify genomic subsets benefiting from adj-PD-1, even without overall survival advantage.3, 4 Survival outcome interpretation is constrained by short follow-up (median 2.2 years) and a small matched cohort (69 pairs). NAM's heterogeneous clinical course requires longer follow-up to capture recurrence patterns, and the small sample limits power to detect modest but meaningful differences (e.g. distant metastasis-free survival). Unreported subgroup analyses (e.g. Stage IIB vs. IIIC–IIID) leave unanswered whether adj-PD-1 benefits high-risk subgroups like Stage IIID NAM (hazard ratio 5.69 for recurrence-free survival vs. IIB1). Lack of immune-related adverse event (irAE) and treatment adherence data is another concern. IrAEs impact treatment continuation and quality of life, and early discontinuation due to irAEs could underestimate adj-PD-1 potential—critical for risk–benefit assessment. Despite limitations, Komori et al.'s study contributes valuable evidence that adj-PD-1 confers no survival benefit in resected NAM, aligning with recent evidence showing no benefit of immune checkpoint inhibitors in acral melanoma.6-8 Future prospective, multicentre studies with ethnically diverse cohorts should incorporate genomic profiling, detailed surgical/comorbidity data and irAE collection. Exploring combination therapies (e.g. anti-PD-1 + anti-CTLA-4) is warranted given NAM's unique biology. In conclusion, these preliminary findings require cautious clinical interpretation. Adj-PD-1 use in NAM should remain reserved for selected patients pending more robust evidence. No funding was received for this study. The authors declare no conflict of interest. Ethical approval is not applicable to this comment, as it is a scholarly discussion of previously published research and does not involve new human or animal subjects or new experimental work. This comment complies with the ethical standards of the Declaration of Helsinki and relevant institutional guidelines. No new research involving human participants or animals was conducted for this work. Data sharing is not applicable to this comment as no new data were created or analysed.