493 Distinct correlative signatures of CD4+ helper and regulatory T cells associate with response to different neoadjuvant immune checkpoint inhibitor combinations in HNSCC

Background Clinical advantages observed with immune checkpoint inhibitors (ICI) in the recurrent/metastatic (R/M) head and neck squamous cell carcinoma (HNSCC) setting have led to increased interest in investigating immunotherapy at an earlier stage in the HNSCC treatment regimen. Meanwhile, the response rate of HNSCC patients is limited, and mechanisms of immune evasion remain poorly understood. Here, we investigate T cell transcriptional and spatial dynamics of HNSCC patients who received neoadjuvant nivolumab (anti-PD-1, Nivo) ± ipilimumab (anti-CTLA-4, Ipi) or relatlimab (anti-LAG-3, Rela) to further elucidate underlying mechanisms of ICI therapy on T cells associated with favorable clinical response. Methods Pre- and post-treatment tumor samples from HNSCC patients (n=42) were processed for single-cell RNA-sequencing, TCR sequencing and cellular indexing of transcriptomes and epitopes (CITE)-seq. Thirty-nine patients completed sequencing and were analyzed using Seurat (v.5.0.1). Specific T cell marker genes and signatures characterize cellular subpopulations with specific transcriptional and clonal dynamics. FFPE tissue sections were used to perform multispectral imaging to evaluate changes in T cell composition within the tumor. Pathologic response was scored using standard criteria.1 Results We observed 4 pathologic major responders (pMR) in each combination arm (Nivo+Ipi: 4/12 (30%), Nivo+Rela: 4/15 (27%)). Lower CTLA4 gene expression at baseline by both CD4 helper (Th) and regulatory (Treg) cells positively correlate with a favorable pathological response to Nivo+Ipi therapy (Th: R = -0.71, p = 0.047, Treg: R= -0.73, p=0.039). In contrast, increased expression of interferon response and T cell exhaustion-associated genes in Th and Treg cells are more abundant in pMR at baseline in the Nivo+Rela arm. Multispectral imaging data indicate an increased frequency of Th cells post treatment in the stroma and tumor, which positively correlate with a better pathological response only to Nivo+Ipi therapy (stroma: R=0.96, p=0.0021, tumor: R=0.76, p=0.046). The overall frequency of activated, TNF-positive Treg subpopulations is increased only after Nivo+Ipi treatment (p<0.001) in all responding patients. Additionally, TCR clonotype analyses reveal clonal sharing between Th and Treg cells suggesting a possible reprogramming of CD4+ T cells populations during therapy. Conclusions Immune checkpoint receptor expression, TCR clonal dynamics and the frequency of Th and Treg cells in the tumor directly correlate with the clinical outcome of neoadjuvant ICI therapy combinations. These findings suggest a possible biologic basis for precision ICI treatment selection. Moreover, this novel neoadjuvant study highlights the possibility to overcome the unsatisfactory response rates to single immunotherapy approaches by implementing a double ICI strategy. Acknowledgements This research was funded by BMS CA223049 (RLF), P50 CA097190 (RLF), R01 CA206517 (RLF), R01 DE031947 (RLF, LV), and DFG research fellowship #522169450 (MB). This research was supported in part by the University of Pittsburgh Center for Research Computing through the resources provided by using the HTC cluster, supported by NIH award number S10OD028483. This research utilized the UPMC Hillman Cancer Center Flow Cytometry Core Facility, supported in part by award P30 CA047904 (RLF). Trial Registration NCT04080804. Reference Cottrell TR, et al. Pathologic features of response to neoadjuvant anti-PD-1 in resected non-small-cell lung carcinoma: a proposal for quantitative immune-related pathologic response criteria (irPRC). Ann Oncol 2018;29(8):1853–1860. Ethics Approval This study was approved by University of Pittsburgh’s Institutional Review Board; approval number HCC 18-139/CA224-056.