← Back to Home

Should Treatment Be Intensified for High-Risk Prostate Cancer? Genomic Scores Offer Clues for Stratification

An analysis combining large-scale real-world data with 401 randomized trial participants suggests that Decipher scores may help identify which patients have a greater need for additional hormone therapy. However, independent validation is needed before the scores can be used to reduce medication use in patients with low scores.

By SURL BioNews

For patients with high-risk prostate cancer that has not yet metastasized to distant sites, additional treatment may provide longer disease control, but it may also cause side effects such as fatigue. The challenge is that tumors may follow different courses even when patients share the same high-risk classification. A new analysis by a University College London (UCL) team offers a clue: genomic scores from tumor tissue may help align treatment intensity more closely with individual patients’ needs.

UCL announced the study on October 23 and said it would be presented on October 24 at the European Society for Medical Oncology (ESMO) annual meeting in Madrid. The team analyzed the Decipher tissue-based molecular test developed by Veracyte, using two data sources: real-world data covering 27,607 men, of whom 5,956 were included in the clinical outcomes analysis; and 401 randomized prostate cancer patients in the STAMPEDE trial who had not yet developed distant metastases.

Both datasets showed a consistent pattern: higher Decipher scores were associated with worse subsequent outcomes. UCL and *Clinical Lab Products* reported that, for each 0.1 increase in score, the adjusted hazard ratio for distant metastasis was 1.49, and the hazard ratio for prostate cancer death was 1.38. These figures describe relative hazards during follow-up and cannot be directly converted into an individual’s probability of developing metastases or dying.

A finding more directly relevant to treatment decisions emerged among patients whose cancer had not spread to the lymph nodes. The research team said Decipher scores helped identify patients who might benefit from adding abiraterone to standard treatment to reduce the risk of subsequent distant metastases. *Healthcare Management* also reported this subgroup finding, noting its potential use in helping weigh the benefits and burdens of additional treatment.

This question builds on existing treatment evidence from STAMPEDE: adding abiraterone can improve disease control and survival for some patients with high-risk prostate cancer that has not yet metastasized to distant sites. The current research asks whether those most in need of intensified treatment can be identified more precisely. Sarah Howlett, a researcher at the UCL Cancer Institute who led the study, said patients with low scores had lower rates of cancer spread and might have an opportunity to receive less additional treatment in the future, while patients with high scores might warrant greater priority for intensified treatment.

However, predicting a worse disease course and demonstrating that a particular group benefits especially from a specific drug are two different questions. The currently available news reports do not fully detail the score thresholds, the absolute benefits in each group, or the results of statistical tests examining whether treatment effects actually varied by score. The available information is therefore insufficient to interpret a low score as indicating that abiraterone can safely be omitted, or to establish stratification rules that can be directly applied.

Howlett emphasized that seeing similar trends in large-scale real-world data and a randomized trial adds credibility to the research, but validation in other independent datasets is still needed before the scores can routinely inform these treatment decisions. The study offers a direction for tailoring treatment more precisely. The next step is to establish whether making treatment choices based on the scores can maintain cancer control while meaningfully reducing the treatment burden on patients.

References

  1. University College London
  2. Clinical Lab Products
  3. Healthcare Management