Purpose

Integrating ultrasensitive prostate specific antigen with surveillance in patients at high risk after radical prostatectomy potentially optimizes treatment by correctly identifying recurrence, promoting an early salvage strategy and minimizing overtreatment. We tested the power of postoperative ultrasensitive prostate specific antigen to identify eventual biochemical failure.

Materials and methods

We identified 247 patients at high risk with a median followup of 44 months who underwent radical prostatectomy from 1991 to 2013. Each patient had extraprostatic extension and/or a positive margin. Surgical technique, initial prostate specific antigen, pathology findings and postoperative prostate specific antigen were analyzed. The ultrasensitive prostate specific antigen assay threshold was 0.01 ng/ml. Conventional biochemical relapse was defined as prostate specific antigen 0.2 ng/ml or greater. Kaplan-Meier and Cox multivariate analyses were done to compare the rates of ultrasensitive prostate specific antigen recurrence and conventional biochemical relapse.

Results

Sensitivity analysis revealed that ultrasensitive prostate specific antigen 0.03 ng/ml or greater was the optimal threshold to identify recurrence. A first postoperative ultrasensitive value of 0.03 ng/ml or greater, Gleason grade, T stage, initial prostate specific antigen and margin status predicted conventional biochemical relapse. On multivariate analysis only a first postoperative ultrasensitive value of 0.03 ng/ml or greater, Gleason grade and T stage independently predicted conventional biochemical relapse. First postoperative ultrasensitive prostate specific antigen 0.03 ng/ml or greater conferred the highest risk (HR 8.5, p < 0.0001) and identified conventional biochemical relapse with greater sensitivity than undetectable first conventional prostate specific antigen (70% vs 46%). Any postoperative prostate specific antigen 0.03 ng/ml or greater captured all failures missed by the first postoperative value (100% sensitivity) with accuracy (96% specificity). Defining failure at an ultrasensitive value of 0.03 ng/ml or greater yielded a median lead time advantage of 18 months (mean 24) over the conventional definition of prostate specific antigen 0.2 ng/ml or greater.

Conclusions

Ultrasensitive prostate specific antigen 0.03 ng/ml or greater is an independent factor that identifies biochemical relapse more accurately than any traditional risk factors and confers a significant lead time advantage. This factor enables critical decisions on the timing of and indication for postoperative radiotherapy in patients at high risk after radical prostatectomy.