This study built a longitudinal single-cell RNA sequencing atlas from 441 samples across 241 patients with ten cancer types to map how the tumor immune microenvironment (TIME) changes during immune checkpoint inhibitor (ICI) therapy, and whether those changes predict response and survival.
Four conserved TIME subtypes were identified; ~40% of tumors shifted between subtypes during therapy. Shifts toward inflamed or B cell-enriched subtypes linked to better response and survival, while persistence in or shifts toward myeloid-dominant subtypes indicated resistance. A derived 59-gene signature predicted response and survival across 1,383 baseline tumors in 19 independent cohorts.
The dataset spans ten cancer types, which may obscure tumor-type-specific biology; the longitudinal atlas is observational, so causal directionality of TIME transitions cannot be confirmed; file access is restricted, limiting independent replication from the shared dataset.
Tracking shifts in the tumor immune microenvironment during ICI therapy — not just the baseline state — may be more informative for predicting outcomes. The 59-gene signature offers a potential tool for baseline response prediction across multiple cancer types, pending prospective validation.
Explore related topics
