Investigating circulating immunosuppression as novel biomarker and potential therapeutic target to overcome immunotherapy resistance in (R/M) HNSCC patients

Introduction: Recurrent/metastatic head and neck squamous cell carcinoma ((R/M) HNSCC) is one of the most aggressive and immunosuppressive cancers. Despite immunotherapy advances, the initial response rate remains low highlighting the urgent need to further investigate the immune evasion mechanisms driving therapeutic resistance and immune-related biomarkers leading to tumor progression. In HNSCC, resistance to immunotherapy is driven by a strong immunosuppressive tumor microenvironment (TME) characterized by the accumulation of myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs) which suppress antitumor immune responses and promote tumor growth. In addition to the TME, soluble immune checkpoints (sICs) have also emerged as modulators of the immune response to tumor cells, influencing cancer development, prognosis and treatment. Despite the significant efforts to characterize the immunosuppressive landscape within the TME, the role of circulating immunosuppressive cells and sICs in influencing immunotherapy efficacy in (R/M) HNSCC remains poorly defined. Aims: The aim is to investigate the potential role of circulating immunosuppressive cells (Tregs and MDSCs) and sICs as possible predictive and prognostic biomarkers in the blood of (R/M) HNSCC patients undergoing pembrolizumab therapy. Results: Eighty (R/M) HNSCC patients receiving anti-PD-1 therapy were enrolled during the study and divided into a discovery cohort (n=40), and a validation cohort (n=40). Peripheral blood samples were collected at baseline (T0) and after one cycle of anti-PD-1 treatment (T1). In the discovery cohort, higher baseline levels of CD137⁺Tregs (>0.081%) as well as LOX-1⁺PMN-MDSCs (>0.26%), identified patients with significantly worse survival. CD137⁺Tregs also positively correlated with performance status (PS), while high levels of LOX-1⁺PMN-MDSCs negatively affected response to pembrolizumab (p=0.021), showing a significant increase in non-responsive patients at T1. Moreover, the functional characterization of CD137⁺Tregs and LOX-1⁺PMN-MDSCs isolated from (R/M) HNSCC patients revealed a significantly stronger immunosuppressive capacity compared to CD137-Tregs and LOX-1⁺PMN-MDSCs, respectively. Multivariate analysis (MVA) identified LOX- 2 1⁺PMN-MDSCs (>0.26%) as independent prognostic factors correlated with PFS (p=0.009) and OS (p=0.018) and further confirmed in a prospective validation cohort (PFS: p=0.007; OS: p=0.001). Concurrently, serum levels of IDO molecule and 13 soluble immune checkpoints (sICs) (sCTLA-4, sCD137, sPD-L1, sPD-L2, sPD-1, sLAG-3, sBTLA, sTIM-3, sCD80, sGITR, sHVEM, sCD27, sCD28) were evaluated, at T0 and T1, by Luminex assay. Among them, sLAG-3 (>152.8 pg/mL) emerged as the only independent prognostic factor for PFS (p=0.02) and OS (p=0.04) in MVA, highlighting its crucial role in the clinical outcome of (R/M) HNSCC patients. These data were further confirmed in the validation cohort. sLAG-3 levels significantly decrease at T1 only in responsive patients (p=0.009), while no significant change was found in non-responsive. Moreover, patients with higher levels of sLAG-3 (>168.8 pg/mL) at T1 exhibited worse overall survival. A combined analysis revealed that patients with LOX-1⁺PMN-MDSCshigh+sLAG-3high, at baseline, had significantly worse survival (PFS: p=0.007; OS: p=0.013) and did not benefit from treatment compared to LOX-1⁺PMN-MDSCslow+sLAG-3low patients. To further elucidate the circulating immune landscape associated with treatment response, a network analysis was performed to characterize the immunological profiles of responsive and non-responsive patients. This integrative approach revealed that responsive and non-responsive patients exhibited different immune network correlation, suggesting that a more structured, organized and coordinated network is associated with a more favourable clinical response. Conclusions: Our study provides novel insights into the immunosuppressive mechanisms that may underlie resistance to anti-PD-1 immunotherapy in (R/M) HNSCC patients. Overall, these findings suggest that circulating immunosuppressive cells and sICs could be used as non-invasive biomarkers to predict resistance to immunotherapy treatment and monitor disease progression in (R/M) HNSCC patients.