Inverse relationship between neoantigen clonality and T-cell activity reveals distinct immune phenotypes in HNSCC.

backgroundNeoantigens derived from tumour-specific mutations are critical targets for T-cell recognition, yet the relationship between neoantigen clonality and immune response in head and neck squamous cell carcinoma (HNSCC) remains poorly understood. We hypothesised that clonal neoantigens, being present in all tumour cells, would drive T-cell exhaustion through persistent antigen exposure.

methodsWe analysed 527 HNSCC tumours from The Cancer Genome Atlas. Neoantigens were predicted using pVACseq with MHCflurry binding prediction for patient-specific HLA class I alleles. We computed a Clonality Score (binder-weighted variant allele frequency normalised by neoantigen count) to quantify clonal origin independent of total burden. T-cell exhaustion was assessed using a 17-gene expression signature, alongside TIDE dysfunction scores and the Pan-Immune Score. Cox proportional hazards models with interaction terms tested whether prognostic effects of clonality depended on immune context.

resultsContrary to our hypothesis, neoantigen clonality showed strong inverse correlations with T-cell infiltration-associated signatures, including a 17-gene exhaustion/activation panel (ρ = -0.41, P = 9.8 × 10⁻²²), TIDE dysfunction (ρ = -0.53, P = 6.7 × 10⁻³⁸), and Pan-Immune Score (ρ = -0.50, P = 8.5 × 10⁻³³). Tumours stratified into four phenotypes: Hot/Low Clonality (33%) exhibited high immune infiltration and cytolytic activity; Cold/High Clonality (33%) showed minimal immune engagement despite abundant clonal neoantigens. High clonality also associated with reduced antigen presentation machinery expression (ρ = -0.31, P = 1.0 × 10⁻¹²), suggesting impaired antigen visibility as a potential upstream mechanism. Survival analysis revealed a significant clonality × immune status interaction (P = 0.034). In 'hot' tumours, high clonality predicted improved survival (HR = 0.72, P = 0.043), whereas clonality had no prognostic effect in 'cold' tumours (HR = 1.00, P = 0.98).

conclusionsNeoantigen clonality inversely predicts T-cell activity in HNSCC, identifying high-clonality tumours as immunologically 'cold' despite abundant tumour-specific antigens. Crucially, the prognostic benefit of clonality is realised only when T-cell effectors are present, reconciling our findings with the paradigm that clonal neoantigens represent optimal immunological targets. This context-dependent relationship generates the hypothesis that neoantigen clonality, when interpreted alongside immune context, may help identify patients who could benefit from immune-priming strategies prior to checkpoint inhibitor therapy, pending validation in immunotherapy-treated cohorts.