This is the authors' abstract. We don't add key points for this paper.
Locally advanced rectal cancer (LARC) is treated with neoadjuvant chemoradiotherapy (nCRT), but only a minority of patients achieve a pathological complete response (pCR). Predictive biomarkers of response could help guide treatment decisions, yet none have reached clinical practice. In this exploratory study, we integrated six publicly available transcriptomic datasets and applied machine learning to derive a 186-gene signature predictive of nCRT response. The signature showed good performance in cross-validation (AUC 0.80) and was associated with consensus molecular (CMS4) and immune (iCMS3) subtypes enriched in responders. Gene set enrichment analyses highlighted pathways involved in tumor growth, immune regulation, and resistance. Spatial transcriptomic profiling of pre-treatment biopsies further identified compartment-specific markers, with tumor-associated genes showing greater predictive value. These results provide biological insights into response mechanisms and generate hypotheses for future validation. Larger prospective studies are required to assess the clinical utility of this approach.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Corrò et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: