​Impact of Molecular Subtyping in Muscle Invasive Bladder Cancer by mRNA expression clustering on Predicting Survival and Response of Treatment

Abstract

Recently discovered molecular classifications for urothelial bladder cancer appeared to be promising prognostic and predictive biomarkers. It is a major challenge for clinical work to study the molecular subtypes of BC. Outcome of bladder cancer (BC) treatment still need establish and explored the molecular subtypes of bladder cancer and potential clusters. The present study was conducted to evaluate the prognostic impact of molecular subtypes assessed by mRNA expression in a consecutively collected, mono-institutional muscle-invasive bladder cancer (MIBC) cohort, performed by unsupervised clustering and validate subtypes of our institutional cohort with data from The Cancer Genome Atlas (TCGA) and possible to correlate the mRNA expression with tumor molecular subtype membership. Our overall goal was to determine whether mRNA expression have shown significant difference in specific molecular subtypes and correlation with clinical outcomes. Molecular subtyping of muscle-invasive bladder cancer (MIBC) predicts disease progression and treatment response. However, present subtyping techniques are based primarily on transcriptomic analysis, which is relatively expensive. Subtype classification of protein levels by immunohistochemistry (IHC) are more affordable and feasible to perform in a general pathology laboratory. Recent data demonstrated that GATA3, CK20, CK5/6, and CK14 protein levels were correlated with MIBC molecular subtypes. We aimed to evaluate the correlation of those IHC markers with survival outcomes after radical cystectomy in Thai patients. Moreover, we aim to evaluate molecular subtypings by mRNA expression analysis. vii Method 30 MIBC were pathologically re-evaluated and molecular subtypes were assessed on mRNA. Fresh-frozen primary tumor samples from a single cohort in Songklanagarind hospital who underwent radical cystectomy between 2015 and 2020. First, we screened the expression profiles of differentially genes expression and of BC by comparing DEG and principle component analysis with K-mean clustering. Moreover, external validation set from the Cancer Genome Atlas (TCGA) database was done by using significant gene expression. We used the complete TCGA dataset with our subtype gene expression and assign TCGA’s bladder cancers to molecular subtypes. Taken together, we explored the molecular subtypes and their outcome treatment of BC. Institutional cohort (n= 30 MIBC) and The Cancer Genome Atlas (TCGA)- dataset (n=231 MIBC) were subtyped using unsupervised genes and analyzed for predicting of survival, cancer-specific survival (CSS), overall-survival (OS), and recurrence–free survival (RFS). Moreover, we evaluated the IHC-based subtypes in MIBC, as classified by GATA3, CK20, CK5/6, and CK14 expression in 132 MIBC patients who underwent radical cystectomy followed by adjuvant chemotherapy (2008–2016). All individual markers and clinicopathological parameters were analyzed against treatment outcomes after radical cystectomy and some selected tissues were sent for whole transcriptome sequencing and clustering from mRNA expression. Result Unsupervised consensus hierarchical clustering applied to gene expression data and identified 3 molecular subtypes. These subtypes were associated with distinct clinicopathological characteristics and molecular expression. The clustering was validated in the TCGA dataset. We identified different clinical characteristics and identified 3 molecular subtypes MIBC specimens from cohort dataset successfully. In multivariable analyses, N- stage, T-stage, M-stage and/or age predicted CSS/OS and/or cisplatin- based adjuvant- chemotherapy response. In the TCGA-dataset, publications report that subtypes risk-stratify patients for OS. For IHC study section, the result showed that the mean patient age was 65.6 years, and the male to female ratio was 6.8:1. Positive IHC expression rates of GATA3, CK20, CK5/6, and CK14 were 80.3%, 50.8%, 42.4%, and 28.0%, respectively. The 5-year overall survival (OS) was 27.0% (95% confidence interval (CI) 19.6%–35.0%). Only GATA3 and CK5/6 were significantly associated with survival outcome (log-rank p-values = 0.004 and 0.02). GATA3 and CK5/6 were then used to establish subtypes, which were luminal (GATA+ viii

and CK5/6−, 38.6%), basal (GATA− and CK5/6+, 12.9%), mixed (GATA+ and CK5/6+, 37.9%), and double-negative (GATA− and CK5/6−, 10.6%). Patients with the mixed subtype had a significantly better 5-year OS at 42.8%, whereas patients with the double-negative subtype had the worst prognosis among the four groups (5-year OS 7.14%). In the multivariable analysis, lymph node status and subtype independently predicted survival probability. The double-negative subtype had a hazard ratio of 3.29 (95% CI 1.71–6.32). Conclusion The results further reinforce the conclusion that the molecular subtypes of bladder cancer are distinct disease entities with specific molecular subtype. In our cohorts/subtyping- classifications, clinical and novel molecular subtypes for predicting outcome of treatment. For immunohistochemistry subtyping using GATA3 and CK5/6 was applicable in MIBCs, and patients with the double-negative subtype were at the highest risk and may require more intensive therapy and mRNA subtyping by mRNA expression must showed the significant relationship with survival rate.