225P - A novel multianalyte signature for stratifying Indian non-muscle invasive bladder cancer: A single center observational study

Background

Common evolution of non-muscle invasive bladder cancer (NMIBC) is local recurrence or progression to muscle invasive bladder cancer (MIBC). Some patients (pts) with NMIBC might develop metastases without progression to MIBC. Clinical evolution of these atypical presentations is unknown. Bladder cancer over the last 5 years has surged by over 13% in India. 50-70% of NMIBC cases recur, and around 10-20% progress to MIBC. Challenge in treatment is identifying the patients with mutations or gene expression patterns for targeted therapy, which is not feasible for all owing to high cost. While omics categorized Western patients, molecular landscape of Indian NMIBC remains incomplete. Herein we report a potential predictive signature, derived by combining transcriptome and whole exome sequencing data of a cohort of NMIBC patients treated in our centre.

Methods

From 2018 to 2021, 25 pts treated at our centre were included for the study, which has been approved by Institutional Ethics committee. Median age at diagnosis was 64Y. 88% received intravesical BCG. 60% were T1 stage and recurrence rate 48%. RNA sequencing and whole exome sequencing were performed on 25 primary NMBIC tumors. Machine learning model was used to predict recurrence status. Mutational status between recurrent and non-recurrent patients was also obtained. The gene expression signature was validated in external datasets including GSE13507 and GSE154261.

Results

435 genes were differentially expressed between recurred and non-recurred patients (p<0.05). Machine learning model predicted recurrence with 90% accuracy. Top pathways upregulated include signalling by nuclear receptors, Shigellosis etc. Downregulated include cell adhesion, PIK-AKT, and ECM reorganization. Validation in external dataset correlated positively with p<0.05. In addition to most previously reported signatures and APOBEC, we also found mutations in FGFR3, ABC transporters, and DAMP, differentially regulated between recurred vs non recurred patients.

Conclusions

We identified novel differentially regulated genes and mutations in genes that predicts recurrence with 90% accuracy. Further analysis in a larger set is underway to complete the validation of the predictive model.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Sri Shankara Cancer Hospital and Research Center.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.