Entinostat Benefit in HR+ Breast Cancer Is a Luminal B-Specific Effect: The DNMT3A/HDAC2 Chromatin Lock on ER Output as the Mechanistic Basis for Subtype-Enriched Patient Selection, with Implications for NCT07235618

Entinostat is now approved in China (NMPA, April 2024) for HR+/HER2- advanced breast cancer, and a May 2025 meta-analysis of four RCTs (n=1,371) confirmed PFS benefit (HR=0.80, p=0.01) with no OS benefit. No published trial has stratified entinostat benefit by Luminal A (LumA) versus Luminal B (LumB) PAM50 subtype. This document reports the OrganismCore attractor geometry finding that entinostat benefit is a LumB-specific phenomenon, not a class effect across all HR+ disease. The mechanistic basis, derived from Waddington landscape analysis of 19,542 single cancer cells (GSE176078) and replicated in METABRIC (n=1,980), is a LumB-specific co-repressor circuit: DNMT3A and HDAC2 are co-expressed in LumB at r=+0.267 compared to r=+0.071 in LumA (p=5.68×10^-56 for the difference). This circuit acts as a chromatin lock on ER output, suppressing downstream oestrogen-responsive gene transcription (TFF1) despite elevated ESR1 transcript — the TFF1/ESR1 decoupling replicated in METABRIC (p=0.0019). Entinostat (class I HDAC inhibitor) inhibits HDAC2 directly, dissolving the co-repressor complex specifically in LumB. In LumA, the circuit is absent; entinostat has no HDAC2-dependent lock to dissolve and no mechanism for improving ET response. The modest class-level HR=0.80 from the published meta-analysis is consistent with a strong LumB signal diluted by non-responding LumA patients (approximately 50%/50% mixture in unselected HR+ trials). The prediction is directly testable in NCT07235618 (entinostat + fulvestrant, post-CDK4/6i failure, Sun Yat-sen University Cancer Center, Guangzhou; PI: Dr. Shusen Wang; N=50; start January 2026). NCT07235618's post-CDK4/6i eligibility criterion selects a de facto LumB-enriched population (LumA patients with intact CDKN1A response are more likely to benefit from CDK4/6i and less likely to progress). PAM50 subtype stratification or FOXA1/EZH2 IHC ratio (CS-LIT-22, DOI: 10.5281/zenodo.18892788) could be added as a correlative endpoint on baseline tissue without protocol amendment, change to primary endpoint, or additional patient burden. Verdict: CONVERGENT-NOVEL. Entinostat clinical activity in HR+ breast cancer is confirmed. The LumB-specific mechanism (DNMT3A/HDAC2 coupling r=+0.267, TFF1/ESR1 decoupling p=0.0019), the subtype-enrichment prediction, and the NCT07235618 correlative proposal have no direct published equivalent as of 2026-03-06.