Evaluation of Prostate Specific Antigen Acceleration for Prostate Cancer Diagnosis

The Calculus of Serum PSA

OBJECTIVES

To review the mathematics of kinetic changes in serum prostate-specific antigen (PSA) and to use a compartmental model to derive a new kinetic measure, alpha.

METHODS

The calculus of kinetic measures of PSA changes with time is presented, and a compartmental model is then used to derive alpha of serum PSA. Alpha is then tested for prognostic importance in 119 men who underwent prostatectomy.

RESULTS

The percentage of tumor in the prostate is closely related to alpha and to tumor length in diagnostic needle biopsies, but not to tumor grade. The presence of adverse pathology in the prostatectomy specimens (positive margins or T3 stage) is significantly associated with alpha, but not to tumor length or grade.

CONCLUSIONS

The derived kinetic parameter, alpha, shows promise as a preoperative prognostic parameter, and may help sort patients into those with low vs high probability for adverse pathology features in the prostatectomy specimens.

A comparison of early prostate-specific antigen decline between prostate brachytherapy and different fractionation of external beam radiation-Impact on biochemical failure

PURPOSE

The aim of this study was to compare early prostate-specific antigen (PSA) decline patterns and PSA nadirs between low-dose-rate seed prostate brachytherapy (LDR-PB) and different fractionations of external beam radiotherapy (EBRT) and their predictive importance for biochemical failure (bF).

METHODS AND MATERIALS

Patients with D'Amico low- or intermediate-risk prostate cancer who underwent a single-modality treatment without androgen deprivation were included in this study. Three different treatment groups were compared: (1) normofractionation EBRT up to 70.2-79.2 Gy/1.8-2.0 Gy, (2) LDR-PB, and (3) EBRT with hypofractionation 60 Gy/3 Gy daily or 5-7.25 Gy once a week over 9-5 weeks, to a total dose of 45-36.25 Gy, respectively. The log-rank test, Cox regression analysis, and nonparametric tests were used.

RESULTS

We analyzed 892 patients: the median followup for patients without bF was 84 months (interquartile range 60-102 months), with 12% of patients experiencing bF. The PSA decline within the first 15 months was generally exponential. LDR-PB showed a faster early exponential decline compared with EBRT treatments, but whether decline was fast or slow had no influence on recurrence. The only factors that were positive predictive factors in univariate and multivariate analyses were the time to nadir >48 months (median), PSA nadir <0.5 ng/mL, and <0.2 ng/mL (all p < 0.001).

CONCLUSIONS

Although there are significant differences in early exponential PSA decline between different treatments, only the PSA nadir and longer time to nadir were predictive factors for bF.