B7-H3: A Continuum-Wide, AR-Repressed Anchor for Combinatorial Prostate Cancer Therapy

A landmark 2026 study in Clinical Cancer Research reframes surface-target therapy in advanced prostate cancer by placing B7-H3 (CD276) at the center of a continuum-wide strategy. Integrating the JHU-PANORAMA single-cell atlas (~1M cells, 213 patients) with patient-derived xenografts, metastatic rapid-autopsy tissues, and mechanistic cell-line work, the authors show that B7-H3 is the most uniformly expressed and least heterogeneous cell-surface target across hormone-sensitive, castration-resistant, neuroendocrine, and “double-negative” prostate cancer. Crucially, B7-H3 is negatively regulated by androgen receptor (AR) signaling, creating a strong rationale for combining B7-H3–directed therapeutics with standard androgen deprivation to overcome resistance.

This addresses a persistent bottleneck: many promising targets (PSMA, STEAP-1/2, KLK2, PSCA) are AR-dependent and drop in AR-negative or neuroendocrine states, precisely where resistance is entrenched. In contrast, B7-H3 remains high and stable across AR-positive adenocarcinomas, AR-negative adenocarcinomas, and neuroendocrine models, with minimal interpatient and intratumoral variability. Functional assays reinforce the therapeutic logic: combining an investigational B7-H3 ADC (MGC018) with enzalutamide produced synergistic tumor growth inhibition, supporting B7-H3–centric regimens layered onto androgen deprivation therapy.

The study also introduces a systematic framework for designing bispecific and combinatorial strategies using “AND” and “OR” gating logic to score target pairs for specificity and phenotypic coverage. B7-H3 achieved high combinatorial scores with TROP-2, NECTIN1, KLK2, and NECTIN4 (targets already in clinical development) suggesting bispecific antibodies or dual-ADC strategies anchored on B7-H3 could maintain coverage across disease states while limiting escape. With B7-H3 agents advancing across modalities (ADCs such as DS-7300a, Fc-engineered antibodies, CAR-T), these data support trials that combine B7-H3 therapeutics with standard ADT from earlier lines and include patients with neuroendocrine or “double-negative” phenotypes who are typically excluded from PSMA-targeted approaches. The authors have released JHU-PANORAMA as an interactive R Shiny app to enable direct exploration and trial refinement.

Source.