Maladaptive Plateau Detection in Prostate Cancer: Pathological Stabilization Across Cohort Observation-Space and Longitudinal PSA Dynamics

This work extends the maladaptive plateau framework into a real-data prostate-cancer demonstration. Using two open datasets, the study shows that Plateau Dynamics can operate in two complementary modes: cohort observation-space mapping when dense time series are unavailable, and individual longitudinal trajectory analysis when repeated PSA measurements exist. In a 600-patient prostate-cancer cohort, PSA/CRP-associated plateau regions were strongly enriched for biochemical recurrence, with 59.8% BCR inside plateau regions versus 18.3% outside. Within Risk Group 3, plateau-positive patients showed 59.6% BCR compared with 18.2% among plateau-negative patients. In a second dataset of longitudinal PSA after external beam radiotherapy, Plateau Dynamics separated post-treatment PSA stabilization into low remission, mid residual, and high maladaptive plateau regimes. Together, the analyses support the central disease insight: cancer recurrence is not only biomarker elevation, but pathological stabilization — the formation of persistent low-dynamical regimes in which disease-associated markers become stable in the wrong biological direction.