Tumor immune dysfunction and exclusion subtypes in bladder cancer and pan-cancer: a novel molecular subtyping strategy and immunotherapeutic prediction model

Molecular subtyping is expected to enable bladder cancer (BC) precise treatment. However, reliable subtyping strategies for clinical application remains defective and controversial. Given the significance of tumor immune dysfunction and exclusion (TIDE) in tumor immune escape and immunotherapy, we aimed to develop a novel TIDE-based subtyping method to facilitate personalized management. Transcriptome data of BC was used to evaluate the heterogeneity and the status of TIDE patterns. We identified 69 TIDE biomarker genes and classified BC samples into three subtypes using consensus clustering. Subtype I showed the lowest TIDE status and malignancy with the best prognosis and highest sensitivity to immune checkpoint blockade (ICB) treatment, which was enriched of metabolic related signaling pathways. Subtype III represented the highest TIDE status and malignancy with the poorest prognosis and resistance to ICB treatment, resulting from its inhibitory immune microenvironment and T cell terminal exhaustion. Subtype II was in a transitional state with intermediate TIDE level, malignancy, and prognosis. We further confirmed the existence and characteristics of our novel TIDE subtypes using real-world BC samples. This subtyping method was proved to be more efficient than previous known methods in identifying non-responders to immunotherapy. We also propose that combining our TIDE subtypes with known biomarkers can potentially improve the sensitivity and specificity of these biomarkers. Moreover, besides guiding ICB treatment, this classification approach can assist in selecting the frontline or recommended drugs. Finally, we confirmed that the TIDE subtypes are conserved across the pan-tumors. In conclusion, our novel TIDE-based subtyping method can serve as a powerful clinical tool for BC and pan-cancer patients, and potentially guiding personalized therapy decisions for selecting potential beneficiaries and excluding resistant patients of ICB therapy. After obtaining patient consent and approval from the institutional research ethics committee, we collected 31 surgical resection samples of bladder tumors from 20 patients who hospitalized in the Department of Urology, Shanghai Sixth People's Hospital. For all resected tumor samples, we invited pathology experts to review the samples for accurate pathological diagnosis of the patients. All tumor samples were collected within 5 minutes after excision and immediately placed in liquid nitrogen for preservation. Subsequently, they were operated for bulk RNA sequencing (LY dataset) to validate the TIDE subtypes. RNA quality was measured with Bioanalyzer 2100 system and RNA Nano 6000 Assay Kit. mRNA was purified from total RNA with magnetic beads with poly-T oligos. mRNA was fragmented with divalent cations and high temperature. First and second strand cDNA were made with random primers, M-MuLV Reverse Transcriptase (RNase H-), DNA Polymerase I and RNase H. cDNA ends were polished and adenylated. Adaptor with hairpin loop structure was ligated to cDNA. cDNA fragments of about 370~420 bp were selected with AMPure XP system. PCR was done with Phusion High-Fidelity DNA polymerase, Universal PCR primers and Index (X) Primer. PCR products were cleaned with AMPure XP system and library quality was checked with Agilent Bioanalyzer 2100 system. Index-coded samples were clustered with cBot Cluster Generation System and TruSeq PE Cluster Kit v3-cBot-HS (Illumia). Libraries were sequenced on Illumina Novaseq platform and 150 bp paired-end reads were obtained. Raw data of fastq format were processed with fastp software. Reads with adapter, ploy-N and low quality were removed. Clean data were obtained and Q20, Q30 and GC content were calculated. Reference genome and gene model annotation files were downloaded from genome website. Reference genome index was built with Hisat2 v2.0.5. Clean reads were aligned to reference genome with Hisat2 v2.0.5. Gene model annotation file was used by Hisat2 to make splice junctions database and get better mapping result. Mapped reads were assembled in reference-based way with StringTie (v1.3.3b). StringTie used new network flow algorithm and optional de novo assembly step to make and quantify full-length transcripts for each gene locus. Reads numbers for each gene were counted with featureCounts v1.5.0-p3. FPKM of each gene was calculated based on gene length and reads count.