Single-cell analysis of matched FFPE and frozen tissue samples reveals comparable resolution of intratumoural heterogeneity

Introduction Intra-tumoural heterogeneity contributes to treatment resistance and disease progression in cancer. Single-cell RNA sequencing (scRNA-seq) enables profiling of cellular diversity in the tumour microenvironment. However, current protocols for generating single-cell 3′gene expression data require intact RNA, which excludes archival formalin-fixed paraffin-embedded (FFPE) tissues. A commercial method now enables scRNA-seq on FFPE samples using a probe-based approach. While methods for spatial and single-cell profiling on FFPE have been previously assessed, none have evaluated whether intra-tumoural heterogeneity was comparable between scRNA-seq data from matched FFPE and fresh-frozen cancer samples across a range of cancers. Methods scRNA-seq was performed on 12 pairs of matched FFPE and frozen tumour tissue and results were compared across key stages of a standardized analysis pipeline. This included cell type annotation, identification of malignant cells, optimization of batch correction methods, and characterization of immune cell subtypes. Results ScRNA-seq from fresh frozen material yielded higher median numbers of unique molecular identifiers (UMIs) and genes per cell than data from FFPE tissues. However, across both FFPE and frozen samples, the same cell types were consistently identified, and candidate malignant cells could be detected by inferring copy number alterations (CNAs). FFPE-derived data were able to resolve sub-clonal CNAs and characterize T cell subpopulations. Discussion These findings demonstrate that intra-tumoural heterogeneity can be inferred from FFPE-based scRNA-seq using tissue samples typically prepared in hospital pathology laboratories. By enabling analysis of widely available FFPE specimens, this approach opens new avenues for population-scale single-cell profiling in cancer research.