Medical Reading. Your HealthGuard.

Determinants Of PSA Change Over Time And Its Link To Recurrence After External Beam Radiation Therapy For Prostate Cancer

October 30, 2017

UroToday - Following external-beam radiation therapy, many criteria based on prostate-specific antigen (PSA) measurements are known to be associated with the risk of clinical disease progression when examined in large cohorts. When faced with a desire to implement salvage therapy before the development of symptomatic disease in individuals with rising PSA levels, the clinician may not be clear as to which features should be used when deciding when to implement salvage therapy. To address this, we used sophisticated statistical techniques to assimilate the combination of prognostic factors, treatment details and post-radiation PSA kinetics into a more accurate understanding of clinical recurrence. This then would enable all the available prognostic information to be used in the individual patient at any time point to derive an optimal profile of clinical failure risk from that time point onwards.

We used multiple large datasets from five cohorts to analyze the risk-model, with each having known values of pretreatment PSA, Gleason Score, T stage, age and radiation dose. A total of 40,324 post-treatment PSA measurements from 4,247 patients treated with external-beam radiation therapy (without androgen deprivation therapy) were included. A linear mixed model was used to model PSA levels over time following radiotherapy, and a Cox model simultaneously determined the association between prognostic factors, the longitudinal post-treatment PSA profile and the time to recurrence.

Of the total patients included, clinical recurrences were observed in 12.2% of patients; the median time to first recurrence was 4.1 years. The initial decline in PSA levels after radiation was significantly associated with pretreatment PSA and T-stage. A subsequent rising PSA profile, indicating treatment failure, was associated with higher values of pretreatment PSA, T-stage and Gleason score and lower radiation doses. When the PSA profile was included in the Cox model as PSA slope (analogous to the doubling time) and the absolute PSA level at any given time, the pretreatment PSA, Gleason score and stage remained significant predictors of clinical failure, as did the PSA profile indices, while radiation dose did not. Interestingly, patient age at treatment had no identifiable impact on the PSA profile, yet young age did significantly increased the risk of clinical recurrence both with and without PSA profile adjustment in the model.

The joint model of the PSA profile and the simultaneous actuarial clinical failure risk is aimed to be a clinically relevant tool. After incorporating the known prognostic information to date, both pre and post radiation therapy, an estimate of the expected risk of clinical failure can be found. To simplify usage, a web-based calculator based on the model has been developed ( This enables a clinician to input all the relevant details along with the PSA history to produce a pair of estimates: a range of curves depicting the estimated future PSA readings, and a survival curve (with uncertainty intervals) showing the estimated risk of clinical failure from that time onwards. Using this, one can counsel individual patients regarding their risk of failure in the subsequent years, and make a more informed decision regarding therapy.

S. G. Williams, J. M. G. Taylor and C. Proust-Lima as part of Beyond the Abstract on UroToday

UroToday - the only urology website with original content global urology key opinion leaders actively engaged in clinical practice.

To access the latest urology news releases from UroToday, go to: www.urotoday

Copyright © 2008 - UroToday