Mathematical Biomarkers Use Routine Blood Tests to Guide Adaptive Prostate Cancer Therapy

Personalizing systemic therapy for prostate cancer remains challenging because tumors can evolve under treatment pressure, driving resistance and early relapse. Continuous high-dose regimens may initially shrink disease but can also favor resistant tumor clones, limiting long-term control. Clinicians therefore need early, noninvasive indicators to identify treatment strategies most likely to sustain response. Researchers have now developed mathematical biomarkers derived from routine blood tests to help guide adaptive therapy decisions in prostate cancer.

Researchers at the H. Lee Moffitt Cancer Center & Research Institute, working with the Wolfson Center for Mathematical Biology at Oxford University, developed three mathematical biomarkers calculated from prostate-specific antigen (PSA) measurements collected during the first treatment cycle. Adaptive therapy pauses treatment when disease is controlled and resumes it when tumor growth returns, with the goal of limiting the expansion of drug-resistant populations. The new biomarkers are designed to predict treatment response early enough to help determine whether a patient may benefit from an adaptive schedule or should remain on continuous therapy.

Image: Researchers developed three mathematical biomarkers from routine PSA blood tests that may help predict early prostate cancer treatment outcomes and personalize adaptive therapy (Image Credit: Adobe Stock)

The three measures are the Adaptive Therapy (AT) Score, Expected Time to Progression (eTTP), and Expected Mean Daily Dose (eMDD). The AT Score estimates a patient’s potential benefit from adaptive therapy compared with continuous treatment. The eTTP projects how long disease control is likely to persist before regrowth, while the eMDD estimates average drug exposure over time under an adaptive regimen, providing an early indication of treatment burden. The framework relies only on first-cycle PSA data to model tumor behavior over time and can be calculated automatically by decision-support software without requiring additional biopsies.

Investigators evaluated the approach using data from 53 patients enrolled in two independent clinical studies. One cohort included 40 patients whose disease remained highly responsive to hormone therapy, while the second included 13 patients whose cancer had progressed despite effective testosterone suppression. Across both groups, the AT Score was strongly associated with time to progression. In the more advanced cohort, higher AT Scores and longer eTTP values were also associated with longer survival. Conventional PSA-derived measures showed substantially less ability to predict outcomes. The findings were published in JAMA Oncology on August 6, 2026.

The authors note that larger, prospective validations are needed before clinical adoption, but the measures could serve as a decision-support tool to better match patients to adaptive versus continuous strategies. The team is also adapting similar models for cancers with noninvasive tumor markers comparable to PSA, such as CA125 in ovarian cancer. If confirmed, this approach may help clinicians balance disease control with drug exposure while anticipating resistance earlier in the care pathway.

Related Links  
Moffitt Cancer Center