Spatiotemporal single-cell analysis decodes cellular dynamics underlying different responses to immunotherapy in Colorectal Cancer

Expanding the efficacy of immune checkpoint blockade (ICB) in colorectal cancer (CRC) patients presses for a comprehensive understanding of treatment responsiveness. Here, we analyzed 169 single-cell samples from CRC patients at multiple sequential time points during the course of anti-PD-1 neoadjuvant therapy to map the evolution of local and systemic immunity. In tumors, exhausted T (Tex) cells or tumor-reactive-like CD8 T (Ttr-like) cells were closely related to treatment efficacy, and we observed correlated dynamics between Tex cells and multiple other tumor-enriched cell types following the treatment. Accordingly, several coordinated cellular programs exhibiting distinct response associations were identified. From a systemic perspective, we found divergent replenishment patterns of Ttr-like cells underlying different response statuses and decoded the phenotypic transitions of Ttr-like cells as they infiltrated tissues from the periphery. Finally, a predictive signature was established using circulating CD8 T cells. Our study provides novel insights into the spatiotemporal cellular dynamics following PD-1 blockade in CRC. Leveraging single-cell RNA-seq and TCR-seq technologies, we generated single-cell transcriptome profiles of primary tumor tissues, adjacent normal tissues, and peripheral blood of 22 CRC patients underwent neoadjuvant anti-PD-1 treatment. Raw sequence data were provided at: China Genomic Sequence Archive (GSA), and we will provide the accession number once the paper is accepted. **Raw data will be uploaded to China Genomic Sequence Archive (GSA), according to the Regulations on the Management of Human Genetic Resources in China.**