The study, led by Dean G. Tang and colleagues across multiple institutions, integrated 87,339 transcriptomic profiles drawn from 33 preclinical and clinical datasets spanning the prostate cancer continuum: normal prostate tissue, treatment-naïve primary tumors, tumors treated with neoadjuvant androgen deprivation therapy, and metastatic castration-resistant prostate cancer. On top of this bulk-tumor resource, the team layered single-cell RNA sequencing data covering 115,197 individual cells, allowing them to locate stemness programs within specific cell populations rather than averaging them across whole tumors.

To quantify stemness, the researchers used a transcriptome-derived mRNA-based Stemness Index, abbreviated mRNAsi, which scores tumors on how closely their gene-expression patterns resemble those of stem cells. They also built a new, prostate-specific measure: a 12-gene signature called PCa-Stem, designed to capture stemness programs relevant to this particular cancer rather than borrowing from stem-cell signatures defined in other tissues.

One of the study’s central findings concerns the relationship between stemness and androgen receptor signaling, the pathway on which most prostate cancer therapies depend. During early tumor formation, stemness and canonical androgen receptor activity rose together. But as tumors progressed to higher Gleason grades, the two trajectories diverged: androgen receptor activity declined while stemness continued to climb. Metastatic castration-resistant prostate cancer showed the highest stemness scores and the lowest canonical androgen receptor activity of any stage examined—a pattern the team also reproduced in mouse models lacking the tumor suppressors Pten, Rb1, and Trp53.

This divergence is clinically meaningful. Androgen receptor signaling is the target of androgen deprivation therapy and androgen receptor pathway inhibitors, the backbone of advanced prostate cancer treatment. Tumors that escape these therapies often do so by shifting into states that no longer depend on androgen receptor signaling, a phenomenon known as lineage plasticity, in which cancer cells change identity and adopt features of other cell lineages. The new data place rising stemness at the center of that transition, showing that globally increasing stemness across the disease continuum is enriched in aggressive molecular subtypes—specifically the PAM50-Luminal B and PCS1 subtypes—and is associated with both proliferative programs and lineage plasticity programs.

Network analyses added a mechanistic layer to the picture. The team identified a coordinated mitotic regulatory program linking four molecular features: elevated MYC activity, loss of the RB1 tumor suppressor, a castration-reprogrammed form of androgen receptor activity, and the PCa-Stem signature itself. In other words, the high-stemness state of advanced prostate cancer appears to be reinforced by the cooperation of MYC activation, RB1 loss, and altered androgen receptor signaling—changes that are themselves hallmarks of therapy-resistant disease.

Spatial and single-cell transcriptomic analyses sharpened this conclusion by showing where the stemness program resides. The PCa-Stem program was localized to lineage plasticity-related epithelial cells, and the fraction of PCa-Stem-positive epithelial cells expanded progressively as disease advanced. This cellular expansion provides a direct, single-cell-level explanation for why bulk tumor stemness scores rise with progression: the stem-like compartment is literally growing within the tumor.

The findings were not purely correlational. In functional experiments, the researchers depleted representative genes from the PCa-Stem signature and inhibited MYC, both genetically and pharmacologically. In diverse prostate cancer models, these perturbations consistently suppressed the phenotypes associated with high stemness. While such experiments were performed in preclinical models and do not by themselves establish clinical benefit, they support the signature’s role as an active driver rather than a passive marker of aggressive disease.

The clinical implications are twofold. First, the stemness score and the 12-gene PCa-Stem signature both predicted poor patient survival in independent patient cohorts, suggesting they could serve as prognostic markers to identify patients at risk of aggressive progression. Second, because the high-stemness state is sustained by identifiable molecular cooperation—MYC activity, RB1 loss, and castration-reprogrammed androgen receptor signaling—the work points toward therapeutic strategies aimed at dismantling that state, including MYC inhibition, which the study shows can suppress stemness-associated phenotypes in experimental models.

The authors conclude that cancer stemness functions as a quantifiable molecular and clinical determinant of prostate cancer aggressiveness, lineage plasticity, disease progression, therapy resistance, and patient survival. By unifying stemness, androgen receptor biology, and resistance mechanisms within a single trajectory-integrated transcriptomic framework, the study offers a common quantitative language for describing how prostate tumors escape treatment—a framework that could guide both risk stratification and the development of therapies targeting the stem-like state itself. As with any transcriptomic analysis, the findings rest on associations and preclinical validation, and prospective clinical testing of the signature’s predictive value would be the necessary next step.

Subject of Research: Quantitative cancer stemness and its role in prostate cancer progression, plasticity, and therapy resistance

Article Title: Increasing cancer stemness drives prostate cancer progression, plasticity, therapy resistance and poor patient survival

Article References: Liu, X., Jamroze, A., Cortes, E., Ji, Y., Zhao, K., Zhao, F., Ho, J., Liu, Y., Davicioni, E., Wu, S., Li, W. J., Feng, F. Y., Alumkal, J. J., Spratt, D. E., Sweeney, C. J., Yu, H., Hu, Q., Zou, C., Zhang, D., … Tang, D. G. (2026). Increasing cancer stemness drives prostate cancer progression, plasticity, therapy resistance and poor patient survival. Molecular Cancer. https://doi.org/10.1186/s12943-026-02795-3

Image Credits: AI Generated

DOI: 10.1186/s12943-026-02795-3

Keywords: prostate cancer, cancer stemness, lineage plasticity, androgen receptor, MYC, RB1 loss, castration resistance, single-cell RNA-seq, PCa-Stem signature, therapy resistance, prognosis, Molecular Cancer