Spotlight Medical Validates AI Assay for Low-Risk High-Risk Breast Cancer

The assay identifies a biologically low-risk subgroup inside clinically high-risk sufferers, with outcomes just like stage I illness, doubtlessly sparing 30,000 girls every year from pointless remedy intensification.

Highlight Medical as we speak introduced that the Journal of Scientific Oncology has revealed a peer-reviewed research describing a clinicopathologic assay that identifies a sizeable subgroup of sufferers labelled scientific excessive danger (ER-positive, HER2-negative early breast most cancers) who’ve a beneficial long run prognosis after customary chemo-endocrine remedy alone. The research reviews blinded validation on two potential cohorts from the CANTO and UNIRAD trials.

Unmet want. Present tips advocate adjuvant escalation with CDK4/6 inhibitors for clinically high-risk ER+/HER2− illness, but no routinely obtainable software reliably distinguishes sufferers who might do nicely with endocrine remedy plus chemotherapy alone. The newly revealed work addresses this scientific hole by combining routine scientific variables with human-interpretable morphology descriptors derived from a single H&E slide. The mannequin integrates 4 clinicopathologic variables and 10 quantitative digital pathology descriptors.

“Escalating remedy for each clinically outlined high-risk case exposes many ladies to toxicity with out profit. This assay gives actionable perception that lets us goal remedy the place it actually helps, doubtlessly sparing 30,000 girls every year from pointless toxicity,” stated Prof. Fabrice André, MD, PhD, Chief Scientific Officer of Gustave Roussy.

Key outcomes (prespecified, blinded validation). On the mixed CANTO+UNIRAD validation cohorts (n = 633), the assay labeled about one in 5 sufferers as low-risk; 95.4% of those remained freed from distant relapse at 9 years, a 79% relative discount in metastatic occasions in comparison with the remaining sufferers (sHR 0.21; p < 0.001).

Context. The 9-year freedom from distant recurrence on this low-risk subgroup is corresponding to outcomes reported in TAILORx for node-negative, intermediate-risk sufferers handled with endocrine remedy alone (≈94.5%), for whom escalation was not thought of useful.

The way it works. The totally locked mannequin makes use of routinely collected clinicopathologic inputs and quantitative, human-interpretable morphology from a single H&E slide; multivariable analyses within the paper point out the assay gives impartial prognostic info past customary elements. Analytical validation confirmed 96–100% reproducibility throughout tumor heterogeneity, scanners, and lab protocols, enabling strong deployment in multicenter settings.

“Our research exhibits that routine tumor slides comprise new, strong, and clinically significant prognostic alerts that may be extracted with clear quantitative strategies,” stated Marvin Lerousseau, PhD, Chief Scientific Officer at Highlight Medical. “The key power of this work lies in demonstrating a completely locked assay validated below blinded circumstances in two potential trials, with outcomes that have been constant throughout distinct affected person populations.”

Article particulars

Title: “Figuring out sufferers with low relapse price regardless of high-risk ER +/ HER2- early breast most cancers: improvement and validation of a clinicopathological assay”
Journal: Journal of Scientific Oncology (on-line 25 August 2025). 

The put up Spotlight Medical Validates AI Assay for Low-Risk High-Risk Breast Cancer first appeared on AI-Tech Park.