Pulse.

a daily field guide to health research that matters

◆ Console

‹ Wed · 1 Jul 2026
Promising but preliminary

Machine learning algorithms develop a tumor-educated platelets-related gene signature to predict colorectal cancer prognosis and therapy response.

A blood test using immune cells predicts colorectal cancer outcomes better than standard staging, potentially guiding treatment decisions.

Using tumor-educated platelets (TEPs) as a liquid biopsy substrate, this study applied multiple ML algorithms to derive a robust gene expression signature for colorectal cancer prognosis and treatment response prediction. The TEP-based signature demonstrated superior performance to standard staging across multiple validation cohorts, offering a minimally invasive prognostic tool that could guide adjuvant treatment decisions in stage II-III CRC patients.

What the study was

Study design
Machine learning prognostic study
Category
Diagnostics
Maturity
Validated
Journal
iScience

Why it surfaced

Tumor-educated platelets as liquid biopsy is an emerging field; ML consensus approach adds methodological rigor; iScience (Cell Press) is a solid venue; CRC is high-incidence with major unmet need for treatment response prediction.

A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.