The aim of this doctoral dissertation was to investigate the association between response to immune check-point blockade (ICB)-based therapy and stress keratin 17 (K17) mediated immune evasion axis in head and neck squamous cell carcinoma (HNSCC). The predictive role of K17 in ICB-treated HNSCC has not been investigated to date.
As part of this work, we set out to investigate the following hypotheses: 1. K17 expression in HNSCC is inversely correlated with the expression of markers of immune activation in HNSCC. 2. Low K17 expression is associated with improved clinical response to ICB-based therapy in HNSCC.
We investigated the differences in immune-related gene expression profile relative to K17 expression by employing targeted spatial transcriptomic profilling using K17 as a morphology marker. To bypass the limitations of spatial transcriptomics technology, we performed highly multiplex immunofluorescence staining. We next performed a clinical biomarker validation study using two independent retrospective cohorts: 1) a discovery cohort consisted of 48 patients with HNSCC that were treated with ICB-based therapy at the University of Wisconsin-Madison, followed by 2) a validation HNSCC cohort treated with ICB-based therapy at Yale University Hospital (n=22). The expression of K17 was determined by immunohistochemistry (IHC) on formalin fixed, parafin embedded (FFPE) samples, and subsequently correlated to clinical outcomes and the expression of other established predictive biomarkers (programmed death ligand 1 (PD-L1), T cell infiltration).
We found K17 low expressing tumors had higher CD4, CD8, B-cell, macrophage and T-reg infiltration. Of the 48 patients in the discovery cohort (60% male, median age 61.5 years), 21 (44%) were K17 high, and 27 (56%) were K17 low. Seventeen patients (35%, 77% K17 low) had disease control, while 31 patients (65%, 45% K17 low) had progressive disease. High K17 expression was associated with lack of disease control (p=0.037), shorter time to treatment failure (p=0.025), progression-free (p = 0.004), but not overall survival (p = 0.06). High K17 expression was associated with lack of disease control in an independent validation cohort (p=0.011). PD-L1 expression or HPV status did not correlate with K17 expression or clinical outcome.
This study represents a comprehensive evaluation of a novel predictive biomarker including a biomarker assay standardization and clinical utility assessment , and an analysis of the correlation between K17 and clinicopathologic characteristics, targetable genomic mutation profile, effect size analysis and a hypothesis-driven spatial transcriptomic experiment, followed by mIF validation. We demonstrate the independent predictive value of K17 in predicting non-response to pembrolizumab-based therapy in HNSCC. We further elucidate the relationship between K17 and the immune microenvironment. Our findings may inform clinicians treating HNSCC patients with ICB how to identify patients that are most likely to respond to this therapy.
|