News|Articles|January 5, 2026

Study validates Decipher as a predictor of early biochemical recurrence in prostate cancer

Author(s)Hannah Clarke
Fact checked by: Benjamin P. Saylor
Listen
0:00 / 0:00

Key Takeaways

  • The Decipher Genomic Classifier effectively predicts 2-year BCR in localized prostate cancer, especially in African American men, with an AUC of 0.889, increasing to 0.984 after adjustments.
  • High-risk African American patients showed a 21.56-fold increase in odds of 2-year BCR, highlighting the classifier's strong predictive capability in this group.
SHOW MORE

According to the authors, these findings support the early use of genomic testing to guide treatment decisions.

Data from the prospective VANDAAM study (NCT02723734) suggest that the Decipher Genomic Classifier is a strong predictor of early 2-year biochemical recurrence for patients with localized prostate cancer, particularly among African American men.1

The findings were published in the Journal of the National Comprehensive Cancer Network.

“One of the most important messages from this study is that genomic risk information adds useful detail on top of the tools clinicians already use,” said principal investigator Kosj Yamoah, MD, PhD, chair of the Radiation Oncology Department at Moffitt Cancer Center, in a news release on the findings.2 “For African American men with early prostate cancer, this test helped separate a small group with rapid recurrence from the large group who remained cancer free at 2 years.”​

In total, the study included 226 evaluable men who received treatment for early-stage prostate cancer between 2016 and 2021. Of those, 207 men had completed test results and follow-up (104 African American patients, 103 White patients). Participants were matched by CAPRA score. All patients were treated with either surgery or definitive radiotherapy with or without short-term androgen deprivation therapy. Patients were followed for an average of 24 months after treatment to monitor prostate-specific antigen (PSA) levels.

In the overall cohort, the genomic classifier yielded an area under the curve (AUC) of 0.783. According to the authors, this suggests that the test “fairly discriminates between patients who did and did not experience 2-year BCR.” The AUC increased to 0.814 after adjusting for clinicopathologic variables.

Patients in the high-risk genomic classifier category demonstrated a 6.6-fold increase in the odds of experiencing 2-year BCR compared with those in the low-risk category (OR, 6.6; 95% CI, 1.61 to 26.91; P = .008). The high-risk category continued to demonstrate higher odds of BCR even after adjusting for clinicopathologic variables such as age, PSA levels, Gleason score, tumor stage, and positive biopsy cores (OR, 5.25; 95% CI, 1.27 to 21.66; P = .02).

Among African American patients, the genomic classifier yielded an AUC of 0.889, showing a strong performance in identifying which men experienced 2-year BCR vs those who did not. The AUC increased to 0.984 after adjusting for clinicopathologic variables.

According to the authors, “This substantial improvement underscores the added value of combining genomic and clinical information to enhance predictive accuracy for 2-year BCR outcomes among [African-American] patients.”

Further, African American patients in the high-risk genomic classifier category demonstrated a 21.56-fold increase in the odds of 2-year BCR compared with African American patients in the low-risk category (OR, 21.56; 95% CI, 1.16 to 400.74; P = .04). Patients in the high-risk category continued to demonstrate significantly higher odds of BCR even after adjusting for clinicopathologic variables (OR, 16.90; 95% CI, 1.18 to 240.96; P = .04).

Among White men in the sample, the test demonstrated moderate discriminatory performance in identifying patients who experienced 2-year BCR vs those who did not, with an AUC of 0.675. The AUC increased to 0.731 after adjusting for clinicopathologic variables.

Overall, race and treatment type continued to be significantly associated with an increased odds of BCR even after adjusting for additional variables (OR, 5.46; 95% CI, 2.04 to 8.89; P = .001). However, the authors note that further validation of the test is warranted given the limited number of BCR events in the sample.

“These data support using genomic testing earlier in care to better match African American patients with the treatment intensity and type that fit the biology of their tumor,” Yamoah concluded in the news release.2 “It is one practical step toward narrowing long-standing differences in prostate cancer outcomes.”​

REFERENCES

1. Yamoah K, Trivedi P, Awasthi S, et al. A prospective validation of the Decipher Genomic Classifier in men with early localized prostate cancer: The VANDAAM study. JNCCN. 2025. https://doi.org/10.6004/jnccn.2025.7089

2. Genomic test helps flag early aggressive prostate cancer in African American patients. News release. H. Lee Moffitt Cancer Center & Research Institute. December 19, 2025. Accessed January 5, 2026. https://www.eurekalert.org/news-releases/1110736

Newsletter

Stay current with the latest urology news and practice-changing insights — sign up now for the essential updates every urologist needs.


Latest CME