Genomic classifier–driven NCCN risk reclassification to track distinct transcriptomic signatures in early prostate cancer.

D Daniel Keizman (Tel-Aviv Sourasky Medical Center, Tel-Aviv, Israel) T Tamar Zahavi A Avivit Peer (Fishman Oncology Center at Rambam Health Care Campus, Haifa, Israel) D David Leonid Sarid (Ichilov Hospital, Sourasky Medical Center, Tel Aviv-Yafo, Israel) A Avishay Sella (Sourasky Medical Center, Tel Aviv, Israel) K Keren Rouvinov (Soroka Medical Center, Beer Sheva, Israel) N Noa Shani Shrem (Soroka University Medical Center, Be'er Sheva, Israel) H Hadas Dresler (Shaare Zedek, Jerusalem, NA, Israel) Y Yoav Manaster (Progenetics IL, Azur, Israel)

Abstract

5013 Background: NCCN clinicopathologic risk stratification in early prostate cancer (EPC) may under- or overestimate biologic risk and contribute to over- or undertreatment. The RNA-based Decipher genomic classifier (DGC), endorsed by NCCN, frequently drives NCCN risk reclassification, yet the transcriptomic biology that tracks with reclassification remains incompletely defined. We investigated whole-transcriptome signatures underlying DGC-based risk stratification and NCCN risk reclassification in a real-world EPC cohort. Methods: We performed a retrospective, real-world analysis, of prostate biopsy Decipher testing of 1,272 men with EPC, using the Decipher genomic classifier clinical database (Veracyte, San Diego, CA) and the GRID RWD registry (NCT02609269). Associations between NCCN risk reclassification and clinicopathologic variables (PSA, clinical T stage, and Gleason Grade Group) were evaluated, along with genomic risk scores, transcriptomic signatures, and a panel of gene expression signatures associated with adverse molecular features, including androgen receptor (AR) signaling, PAM50 subtype (Luminal B), inferred TP53/PTEN activity, and pathways related to proliferation, DNA repair, and immune response. Results: By DGC, 21% of tumors were classified as low risk, 34% as intermediate risk, and 45% as high risk. Decipher testing resulted in NCCN risk up-classification in 38% and down-classification in 19%, while 43% remained in the same NCCN risk category. Whole-transcriptome analysis revealed distinct molecular patterns across reclassification groups. Tumors that were up-classified exhibited a shift toward a more proliferative, less androgen-dependent phenotype, with significant increases (p < 0.05) in proliferation, DNA-repair, and immune-related pathways, increased Luminal B prevalence, reduced androgen receptor (AR) signaling, and decreased TP53 and PTEN activity. In contrast, down-classified tumors demonstrated significant downregulation of proliferation, DNA-repair, and immune-related pathways, with preservation of AR signaling and PTEN activity (p < 0.05). Conclusions: Genomic classifier–driven NCCN risk reclassification tracks coherent transcriptomic patterns that extend beyond clinicopathologic stratification. Higher genomic risk and up-classification align with Luminal B enrichment, reduced AR signaling, and loss of TP53/PTEN activity, supporting the clinical and biological validity of genomic risk refinement beyond conventional NCCN classification.

Article Details

Volume / Issue Vol. 44, Issue 16_suppl
Published June 01, 2026
Pages 5013-5013
ISSN 0732-183X
Publisher Lippincott Williams & Wilkins

Journal Info

Journal of Clinical Oncology

Lippincott Williams & Wilkins

ISSN: 0732-183X Health Sciences

Authors (9)

D

Daniel Keizman

Tel-Aviv Sourasky Medical Center, Tel-Aviv, Israel

T

Tamar Zahavi

A

Avivit Peer

Fishman Oncology Center at Rambam Health Care Campus, Haifa, Israel

D

David Leonid Sarid

Ichilov Hospital, Sourasky Medical Center, Tel Aviv-Yafo, Israel

A

Avishay Sella

Sourasky Medical Center, Tel Aviv, Israel

K

Keren Rouvinov

Soroka Medical Center, Beer Sheva, Israel

N

Noa Shani Shrem

Soroka University Medical Center, Be'er Sheva, Israel

H

Hadas Dresler

Shaare Zedek, Jerusalem, NA, Israel

Y

Yoav Manaster

Progenetics IL, Azur, Israel