Disitamab vedotin (RC48-ADC), an emerging targeted drug, shows promising efficacy and manageable safety in patients with muscle-invasive bladder cancer (MIBC). Antibody drug conjugates (ADCs) can activate the immune microenvironment for immune checkpoint inhibitors (ICIs) combination. However, the lack of efficacy-predicting biomarkers restricts MIBC precision treatment.

We collected 11 bladder tumor specimens from 10 patients before and after neoadjuvant RC48-ADC plus ICI treatment (6 specimens before treatment and 5 specimens after treatment), and performed single-cell RNA sequencing on them. After quality control, we aligned the FASTQ files to the GRCh38 human reference genome using Cell Ranger software. 45603 cells remained for further analysis.

21206 bladder cancer (BLCA) epithelial cells were categorized into 4 clusters: C1, C2, C3, and C4 (Figure 1). HSPA1A was significantly overexpressed in all subclusters after treatment. As the target of RC48-ADC, HER2 was specifically highly expressed in C3 before treatment. In the samples that achieved complete response (CR), HER2 was highly expressed in C3 while HSPA1A was not expressed. Notably, HER2 high-expression was unique to C3, and HSPA1A high-expression always included C3. In C3, patients with positive HSPA1A (regardless of the HER2 status) could not achieve CR (Figure 2). Pseudo-time analysis showed an evolutionary path: C3/C4→C1→C2. The expression characteristics of HER2 and HSPA1A originated from the C3 and remained stable during the evolutionary process. Therefore, C3 may determine the sensitivity and resistance of BLCA to RC48-ADC treatment.

Detecting the expression of HER2 and HSPA1A in C3 subcluster based on single-cell sequencing is expected to serve as a biomarker to predict the efficacy of neoadjuvant RC48-ADC combined with ICIs in the treatment of MIBC.