109P A novel, comprehensive glimpse at NY-ESO-1 expression, mRNA to protein translation, & potential impact on clinical studies

Letetresgene autoleucel (lete-cel; GSK3377794) are autologous T cells engineered to express a high affinity TCR capable of recognizing the HLA-A*02:01, *02:05, *02:06 and SLLMWITQC antigen complex. In this multi-modality study we compared expression of NY-ESO-1 protein (IHC) to mRNA (NGS), evaluated correlation between HLA-A and NY-ESO-1 mRNA expression, and determined LAGE-1a mRNA expression in indications that have not been well defined for these biomarkers. Correlative analysis was performed on HLA typing between Sanger sequencing and NGS. NY-ESO-1 IHC was performed on procured gastric, bladder, HNSCC, and ovarian tumors (cut-off: >10% proportion of cells at >1+ staining intensity). A subset of samples had comparative IHC and NGS. An optimal cutoff for NY-ESO-1 mRNA expression to give the best concordance to the IHC results was determined by ROC curve analysis. LAGE-1a mRNA expression was determined by WTS (mRNA CTAG2 > 0). HLA typing and expression levels were determined by WES and WTS. For tumors with matched blood, concordance of HLA typing between Sanger sequencing and NGS was determined. The human samples were sourced ethically and research use was in accord with terms of the ICF. The prevalence of NY-ESO-1 expression (IHC) was 17% in gastric, 7% in ovarian, 20% in HNSCC and 21% in bladder. Comparison of the NY-ESO-1 mRNA and protein data yielded a positive percent agreement (PPA; sensitivity) = 1 and a negative percent agreement (NPA; specificity) = 0.97. Within subset, NY-ESO-1 mRNA expression was observed in 30% gastric, 12% ovarian, 30% HNSCC, and 36% bladder; LAGE-1a expression was found in 50% gastric, 50% ovarian, 30% HNSCC, and 64% bladder. Sanger sequencing and NGS showed a 97.6% concordance for HLA. NY-ESO-1 & LAGE-1a were expressed in all indications investigated. NGS was highly correlated to IHC and Sanger sequencing and may be an alternative method for identifying a higher number of tumors that express biomarkers of interest. Our data show a unique platform for comprehensive assessment of disease which may improve patient stratification and management strategies. These findings may reduce the need for multiple assays and increase the opportunity to expand into new tumor types.