Panel-derived TMB and immune-related GEP are associated with the response to immune checkpoint inhibitors (ICIs) in urothelial cancers.

Abstract

e17047 Background: Tumor mutational burden (TMB) and gene expression profile (GEP) have emerged as potential biomarkers for the prediction of response to the immune checkpoint inhibitor (ICI) treatment. An interferon-g gene signature was shown recently to predict ICI response in a pan-cancer setting. For TMB, not only did TMB-high patients experience better clinical outcome with ICI, but panel-derived TMB was also shown to be comparable to the gold standard, TMB derived from whole exome sequencing. Here, we evaluate possible correlations between panel-derived TMB, as well as immune-related GEP, with ICI response in urothelial cancers. Methods: FFPE tumor tissues from 30 patients with urothelial cancers who had previously received ICI as monotherapies or combination with chemotherapies at Kaohsiung Chang Gung Memorial Hospital were retrospectively tested for targeted next-generation sequencing (ACTOnco) and quantitative PCR-based GEP (ACTTME), for the identification of nonsynonymous variants across 440 cancer-associated genes and expression levels of > 90 immune-related genes, respectively. TMB was calculated by using the sequenced regions of ACTOnco to estimate the number of somatic nonsynonymous mutations per megabase of all protein-coding genes. For ACTTME, Cq values were normalized to internal control before group analysis. RECIST criteria were used to categorize tumor response to the treatment. Results: Patients were defined as responders (CR/PR) and non-responders (SD/PD) according to the RECIST criteria. In the ICI monotherapy cohort, the responders (n = 8) had significantly higher TMB than non-responders (n = 15) (Median 16.2 muts/Mb vs. 6.5 muts/Mb, p= 0.0035). In contrast, for the cohort receiving ICI combination therapy, several genes implicated in hypoxia (HIF1A), suppressive cell types (Treg & MDSC) and immune checkpoint (PD-L2) were significantly elevated ( p< 0.05) in the non-responder group by 2 to 8 folds. Conclusions: Despite the fact that the cohort size is small, this study showed panel-derived TMB, as well as immune-related GEP, can potentially serve as predictive biomarkers to identify urothelial cancer patients for immune checkpoint inhibitor therapy.