Related Experiment Video
Updated: Aug 17, 2026
A New Technique for Treating Low-risk Prostate Cancer—Super Active Surveillance
Published on: November 7, 2025
PSA bounces after neoadjuvant androgen deprivation and external beam radiation: impact on definitions of failure
Anthony L Zietman1, John P Christodouleas, William U Shipley
1Department of Radiation Oncology, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA. azietman@partners.org
Purpose:
To determine the characteristics of prostate specific antigen (PSA) bounces after external beam radiation therapy (EBRT) with neoadjuvant androgen deprivation and their impact on definitions of biochemical failure.
Methods And Materials:
Characteristics of bounce were calculated for all patients treated by EBRT with neoadjuvant androgen deprivation at our institution between 1992 and 1998 (preexclusion analysis). Calculations were repeated for the subgroup that satisfied additional inclusion/exclusion criteria (postexclusion analysis). The percentage of bounces scoring as false positives according to the ASTRO definition of biochemical failure was compared with those for three alternative definitions (Vancouver, Nadir-plus-two, and Nadir-plus-three) using McNemar's tests.
Results:
Thirty-nine percent (preexclusion cohort) and 56% (postexclusion cohort) of patients demonstrated a PSA bounce. Twenty percent (preexclusion analysis) and 25% (postexclusion analysis) of these bounces scored as biochemical failure according to the ASTRO definition. The Nadir-plus-three definition scored the smallest percentage of bounces as failure, but the difference between this definition and the ASTRO definition reached statistical significance in neither preexclusion nor postexclusion analyses (p >/= 0.070).
Conclusions:
A substantial proportion of patients treated by EBRT with neoadjuvant deprivation experienced a PSA bounce. A large percentage of these bounces scored as biochemical failure according to the ASTRO definition. The Nadir-plus-three definition is less vulnerable to this bias.
