654MO - High dimensional immuno-phenotyping of immunotherapy response in head and neck cancer

Background

Immunotherapies, in particular immune check point inhibitors (ICI), have proven to be game changing treatments of mucosal head and neck squamous cell cancer (HNSCC). Emerging successes with anti-PD-1/PD-L1 ICI therapy have lead to durable responses and prolonged survival in human papillomavirus positive (HPV+) and negative (HPV-) patients, and there is now an urgent need for predictive biomarkers to guide patient selection for highly targeted ICI therapies. There are currently no means by which to determine whether a patient would respond to anti PD-1/PD-L1 immunotherapy. Studies in the tumour microenvironment (TME) have demonstrated that a high degree of T-cell infiltration provides fertile grounds for effective ICI.

Methods

In this project, we used high dimensional spatial biomarker profiling to characterise the TME using targeted (region of interest) and unbiased (whole slide imaging) of metastatic/recurrent HNSCC tumours from a cohort of patients (n=40) treated with Pembrolizumab/Nivolumab. The discovery cohort consisted of patients who had complete/partial/stable/progressive response to ICI therapy. We first analysed tissues using targeted panel, GeoMx digital spatial profiling (Nanostring Technologies) of 80 protein targets simultaneously across immune cell profiles, immune-modulatory targets and IO-drug targets. We then expanded this to ultra-high plex antibody imaging with the Phenocylcer Fusion platform (Akoya Biosciences) of over 50 antibodies to derive single cell phenotypic signatures that could offer cues as to which treatment matches best with certain outcome groups.

Results

Our study identified stromal tissue signatures, specifically, VISTA, CD44, CD127 associated with resistance to immunotherapy. Most notably, high expression of VISTA within the stromal compartment was associated with a worse overall survival (p=0.0062). Using whole slide tissue imaging of single cell proteomic readouts, we were able to co-localise this signature to cell types within the tumour microenvironment.

Conclusions

Our study demonstrates the complementarity of targeted proteomic discovery and whole tissue imaging to identify biomarkers associated with response to ICI therapy in HNSCC.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The University of Queensland.

Funding

The Passe and Williams Foundation.

Disclosure

N. Ma, N. Jhaveri, A. Pratapa: Other, Institutional, Full or part-time Employment: Akoya Biosciences; Financial Interests, Institutional, Stocks/Shares: Akoya Biosciences. B. Ben Cheikh, O. Braubach: Non-Financial Interests, Institutional, Full or part-time Employment: Akoya Biosciences; Financial Interests, Institutional, Stocks/Shares: Akoya Biosciences. All other authors have declared no conflicts of interest.