Related Experiment Video
Updated: Sep 3, 2026

Retzius-Sparing Robot-Assisted Radical Prostatectomy
Published on: May 19, 2022
De-Escalation Therapy in Metastatic Prostate Cancer: A Retrospective Study on Intermittent ADT and ARPI Combination
Barak Talmor1, Jalal Baranseh1, Tomer Charas1
1Department of Oncology, Rambam Medical Centre, Haifa, Israel.
Introduction:
Androgen deprivation therapy (ADT) combined with an androgen receptor pathway inhibitor (ARPI) is standard care for metastatic prostate cancer (mPCa), but prolonged treatment causes clinical and financial toxicity. The efficacy of intermittent treatment in the ARPI era remains unclear. We evaluated outcomes after planned treatment cessation following a deep response to ADT plus ARPI.
Patients:
We included 69 men with metastatic hormone-sensitive (mHSPC) or castrationresistant prostate cancer (mCRPC) treated between January 2014 and May 2025. Eligible patients received ADT plus ARPI for ≥ 12 months without progression and achieved deep biochemical and radiological responses before treatment cessation.
Methods:
This retrospective cohort study evaluated treatment-free survival (TFS), progression-free survival (PFS), and duration of response upon rechallenge.
Results:
Median induction therapy duration was 26 months, and median TFS was 35 months. Two-and five-year TFS rates were 56.2% and 33.8%, respectively. On multivariable analysis, initial hormone sensitivity (mHSPC versus mCRPC) was the sole independent predictor of a durable treatment break (HR 2.27; P = .029). In the mHSPC cohort, median PFS from induction initiation was 65.5 months, while median TFS was not reached. Of 33 patients who progressed, 27 were rechallenged with the same regimen, achieving a 96.3% response rate.
Conclusion:
A planned treatment holiday after a deep response to modern ADT plus ARPI is feasible and durable, particularly in mHSPC. These deep responders may remain off therapy longer than their initial induction period without evident loss of treatment sensitivity upon rechallenge.
