Early prostate specific antigen (PSA) response predicts PSA nadir in patients with prostate cancer treated with external beam radiation

Abstract

AbstractSerum prostate specific antigen (PSA) values decrease after external beam radiation (EBRT) for clinically localized prostate cancer, reaching their nadir 6–12 months after treatment. The nadir PSA value is predictive of disease‐free survival. This prospective study was initiated to evaluate whether the early response of PSA as measured by 1) the rate of PSA decrease during treatment (PSA slope) and 2) the immediate (within 1 month) post‐treatment PSA would predict the PSA nadir.Thirty patients treated with EBRT for clinical stages A2‐C2 (T1c‐T3c) adenocarcinoma of the prostate were enrolled in this prospective study. Nine patients were subsequently excluded because they were lost to follow‐up after less than 1 year (n = 6) or their pretreatment PSA was measured by a method other than Hybritech (n = 3). Serum PSA measurements were obtained at weeks 2, 4, and 6 of treatment (n = 21) as well as immediate post‐treatment PSA at the first follow‐up 2–4 weeks after treatment (n = 17). The end‐point analyzed was whether or not the PSA nadir was ≦ 1.5 ng/ml. A linear regression model of log(PSA) vs. time (treatment week) was fit for each patient's data. The rate of decrease of PSA (PSA slope) was estimated from this regression. The PSA slope, pretreatment PSA, immediate post‐treatment PSA, stage, and grade were studied by Cox life table regression analysis to determine predictors of PSA nadir. In the Cox model, time was measured from the start of radiotherapy until PSA ≦ 1.5 ng/ml or until the last measurement for those who did not reach this level. Models combining the significant individual factors were then constructed. The minimum follow‐up is 52 weeks and the median is 66 weeks.The PSA slope (P = 0.05) and the immediate post‐treatment PSA (P = 0.02) predicted a PSA nadir ≦ 1.5 ng/ml. Pretreatment PSA approached significance (P = 0.09). A model which combined pretreatment PSA and PSA slope predicted PSA nadir ≦ 1.5 ng/ml (P = 0.05) as did PSA slope combined with immediate post‐treatment PSA (P = 0.01). Within each model, PSA slope was the stronger predictor.In conclusion, the early response of serum PSA as measured by the PSA slope and the immediate post‐treatment PSA appears to be predictive of PSA nadir in patients treated with EBRT. More patients and longer follow‐up are needed to confirm this finding and determine whether the early response predicts disease‐free survival. © 1995 Wiley‐Liss, Inc.