Validating the hybrid epithelial/mesenchymal state as a biomarker of treatment response and outcomes in head and neck cancer

PROJECT SUMMARY Head and neck squamous cell carcinoma (HNSCC), including oral cavity squamous cell carcinoma (OCSCC) is the sixth leading cause of cancer-related mortality, with the majority of deaths attributable to tumor metastasis and failures in treatment. Because most cases of OCSCC result from tobacco and alcohol exposure, these tumors are highly heterogeneous, greatly complicating diagnosis, treatment, and investigations into the biology of this disease. We recently performed single cell RNA-sequencing (scRNA-seq) in OCSCC and identified a hybrid epithelial/mesenchymal state (HEM) with some features of classical EMT, yet persistent expression of epithelial markers (Puram et al., Cell). Further investigation into HEM demonstrated its localization at the leading edge of tumors where it appears to drive invasion and metastasis. Accordingly, HEM is associated with adverse pathology and poor prognosis, with a stronger effect on survival than even smoking. Strikingly, HEM appears to be a strong predictor of outcomes in both Black and White Americans (Mazul et al., Med), suggesting it may be useful in diverse sociodemographic groups. Based on these findings, we hypothesize that the HEM state is regulated by the complex interplay between malignant, immune, and stromal cells, and will reliably predict unfavorable outcomes in diverse OCSCC patients in treatment-naïve and recurrent/metastatic settings. First, we will validate a RNA-seq based HEM heterogeneity score to predict outcomes in a diverse cohort of 400 treatment naïve (including 150 Black American) OCSCC patients (Aim 1). However, we are also curious if HEM may predict outcomes in the recurrent/metastatic setting (Aim 2). Because these patients typically have limited tissue sampling due to tumor accessibility, bulk RNA-seq methods on fresh frozen tissue may be challenging. Thus, we will first define the effect of HEM on immune infiltration using multiplexed, multispectral imaging (MSI) approaches, followed by determining if MSI metrics of HEM predict outcomes in recurrent/metastatic patients. In addition, preliminary MSI analyses reveal that HEM leads to exclusion of tumor infiltrating lymphocytes including T-cells. Thus, we will also test whether HEM may predict immunotherapy response, which is front line therapy in recurrent/metastatic OCSCC patients: We will utilize MSI to analyze patients tumors treated with immune checkpoint inhibitors (ICI) to establish imaging-based metrics that predict response. If successful, our studies will authenticate HEM as a critical biomarker in OCSCC to risk stratify biologically favorable and unfavorable disease, improving patient expectations regarding disease outcomes, better informing patient-physician interactions regarding treatment decisions, and enabling future clinical trials for high risk OCSCC tumors using targeted therapy or potentially intensifying adjuvant therapy beyond standard-of-care treatment. Because no molecular markers of OCSCC are routinely used in clinical practice, these efforts represent an urgent and unmet need, which if addressed could significantly impact the management of these challenging patients and create opportunities for more personalized treatment through future studies.