Verification of molecular subtyping of bladder cancer in the GUSTO clinical trial

Abstract

The GUSTO clinical trial (Gene expression subtypes of Urothelial carcinoma: Stratified Treatment and Oncological outcomes) uses molecular subtypes to guide neoadjuvant therapies in participants with muscle‐invasive bladder cancer (MIBC). Before commencing the GUSTO trial, we needed to determine the reliability of a commercial subtyping platform (Decipher Bladder; Veracyte) when performed in an external trial laboratory as this has not been done previously. Here, we report our pre‐trial verification of the TCGA molecular subtyping model using gene expression profiling. Formalin‐fixed paraffin‐embedded tissue blocks of MIBC were used for gene expression subtyping by gene expression microarrays. Intra‐ and inter‐laboratory technical reproducibilities, together with quality control of laboratory and bioinformatics processes, were assessed. Eighteen samples underwent analysis. RNA of sufficient quality and quantity was successfully extracted from all samples. All subtypes were represented in the cohort. Each sample was subtyped twice in our laboratory and once in a separate reference laboratory. No clinically significant discordance in subtype occurred between intra‐ or inter‐laboratory replicates. Examination of sample histopathology showed variability of morphological appearances within and between subtypes. Overall, these results show that molecular subtyping by gene expression profiling is reproducible, robust and suitable for use in the GUSTO clinical trial.

Predicting response to cisplatin-based neoadjuvant chemotherapy for muscle-invasive bladder cancer: transcriptomic features outrank genomic biomarkers

“…While some studies have shown that molecular subtypes could be potentially used to personalize BC treatment 17,21 , we achieved poor AUCs for RF models using the LundTax and TCGAclas subtypes (0.41 and 0.51, respectively; Supplementary Fig. 5D, Supplementary Table 10).…”

“…Some studies suggest that MIBC molecular subtypes predict NAC response, and can be useful for treatment selection and improving outcomes 17 . However, these findings have been demonstrated to be inconsistent over time.…”