Multi-omics and single-cell analysis link the pan-cancer TOP2A-immune exclusion paradox to a non-cell-autonomous dilution effect

Abstract

TOP2A is universally upregulated in human cancers, yet its negative correlation with immune infiltration in bulk transcriptomes remains mechanistically unresolved at single-cell resolution. We integrated multi-omics data across 34 cancer types with external prognostic validation in 59 independent datasets. Single-cell deconvolution resolved cell-type-specific expression, and confounder-adjusted analyses distinguished proliferation-dependent from TOP2A -specific phenotypes. Pharmacogenomic profiling employed bidirectional Connectivity Map screening with cross-validation across four drug sensitivity databases. TOP2A was broadly upregulated at mRNA and protein levels across cancers, with high expression associated with shorter survival in most malignancies yet a protective effect in THYM and READ. Copy-number amplification, rather than somatic mutation, emerged as the predominant genomic correlate of TOP2A overexpression. TOP2A expression correlated positively with tumor mutation burden, homologous recombination deficiency, aneuploidy, and loss of heterozygosity, indicating widespread genomic instability. Single-cell analysis revealed that TOP2A expression is stringently restricted to malignant epithelial cells and proliferating immune subsets. Tumor purity and proliferation-adjusted analyses demonstrated that the bulk-level immune exclusion signature reflects stoichiometric dilution driven by malignant cell expansion rather than direct immunosuppression, whereas associations with genomic instability were largely proliferation-independent. Pharmacogenomic cross-validation revealed enhanced sensitivity of TOP2A -high tumors to topoisomerase, Aurora kinase, and microtubule inhibitors, but intrinsic resistance to MEK and EGFR inhibitors. CMap screening prioritized the HDAC inhibitor MS-275 as a candidate with pan-cancer reversal potential across 22 cancer types. This study clarifies that TOP2A functions as a pan‑cancer barometer of proliferative burden and genomic instability rather than a direct immune suppressor. Resolving the bulk-level immune paradox as a non-cell-autonomous dilution effect through single-cell deconvolution and confounder-adjusted analyses, and identifying putative therapeutic vulnerabilities, we provide a framework for deploying TOP2A as a prognostic biomarker and a hypothesis-generating therapeutic target.