Combined ctDNA and PSA Testing Predicts Survival in Metastatic Prostate Cancer
A prospective cohort study shows combining ctDNA with PSA levels dramatically improves survival prediction in high-volume metastatic prostate cancer patients.
Summary
A UK national cohort study of 114 men with high-volume metastatic prostate cancer found that measuring circulating tumor DNA (ctDNA) alongside PSA during first-line treatment powerfully predicts survival. Before treatment, 70% of patients tested positive for ctDNA; after 6–12 weeks of therapy, only 29% remained positive. Patients who still had detectable ctDNA at that point had a 12-month survival rate of 73%, compared to 99% for those who cleared it. When ctDNA was combined with PSA levels in statistical models, the highest-risk group had a hazard ratio for death of over 20. Importantly, ctDNA alone predicted outcomes even before combination therapy began. These findings suggest that routine liquid biopsy monitoring could guide treatment decisions and identify patients who may benefit from treatment escalation or modification.
Detailed Summary
Metastatic prostate cancer is a leading cause of cancer death in men, but outcomes vary enormously between patients. Identifying who is responding to treatment — and who is not — early in the treatment course could allow clinicians to intervene before disease progresses. This study addresses that critical clinical gap using blood-based biomarkers.
The PARADIGM study prospectively enrolled 114 men with high-volume metastatic prostate cancer who were beginning standard first-line treatment: androgen deprivation therapy (ADT) combined with either docetaxel chemotherapy or an androgen receptor pathway inhibitor. Blood was collected at the start of each of the first six treatment cycles, allowing dynamic monitoring of two key biomarkers: circulating tumor DNA (ctDNA) and serum PSA.
At baseline, before any treatment, 70% of patients had detectable ctDNA. After 6–12 weeks of combination therapy, that figure dropped to 29% — but patients who remained ctDNA-positive faced dramatically worse outcomes. Their 12-month overall survival was 73% versus 99% for ctDNA-negative patients; at 24 months, survival was 50% versus 85%. When ctDNA was combined with PSA in multivariable models, both emerged as independent risk factors, with the poorest-prognosis group carrying a hazard ratio for death exceeding 20.
A particularly notable finding was that ctDNA alone — but not PSA — predicted shorter survival during the ADT lead-in phase, before combination therapy was added. This suggests ctDNA captures biological information that PSA cannot, especially early in treatment.
The authors propose that integrating ctDNA with PSA and clinical variables into a composite monitoring strategy could enable ctDNA-informed treatment modifications, such as escalating therapy for patients who remain ctDNA-positive. These results support prospective evaluation of ctDNA-guided treatment adaptation in future clinical trials.
Key Findings
- ctDNA positivity after 6–12 weeks of treatment was associated with 73% vs. 99% 12-month survival.
- Combined ctDNA and PSA yielded a hazard ratio for death exceeding 20 in the highest-risk group.
- ctDNA predicted survival during ADT alone, before combination therapy began; PSA did not.
- ctDNA prevalence dropped from 70% pre-treatment to 29% after early combination therapy cycles.
- Combining ctDNA, PSA, and clinical features improves survival prediction accuracy over either alone.
Methodology
PARADIGM was a prospective national cohort study enrolling 114 biologically male patients with high-volume metastatic prostate cancer across multiple UK NHS centers. Serial blood samples were collected at the start of each of the first six treatment cycles to enable dynamic biomarker monitoring. Multivariable models were used to assess the independent prognostic value of ctDNA and PSA.
Study Limitations
This summary is based on the abstract only, as the full text is not openly accessible, limiting assessment of methodology, subgroup analyses, and raw data. The cohort was limited to 114 patients, which constrains statistical power, particularly for subgroup analyses. As a prospective cohort study rather than a randomized trial, ctDNA-guided treatment modification has not yet been tested for efficacy.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.