Introduction: Around 90% of prostate cancers detected using the serum prostate specific antigen (PSA) as a screening test are considered to be localised. However, 20–30% of men treated by radical prostatectomy experience biochemical failure within two years of treatment. The presence of primary circulating prostate cells (CPCs) in the blood of these men implies a dissemination of the tumour and could indicate a greater risk of treatment failure.

Objective: To evaluate the use of the number of primary CPCs detected before surgery in the prediction of biochemical failure at ten years.

Hypothesis: The dissemination of cancer cells to distant sites will determine the patient’s prognosis. The absence of primary CPCs in men undergoing radical prostatectomy for prostate cancer may imply a less aggressive disease and therefore could be utilised as a prognostic factor to predict biochemical failure after surgery.

Methods and patients: A single-centre observational study of a cohort of 285 men who underwent radical prostatectomy as monotherapy for prostate cancer, in whom the number of CPCs prior to treatment was determined, and who were followed up for ten years to determine biochemical failure. A Cox proportional risks with polynomial fractions analysis was used to predict biochemical failure based on the number of primary CPCs detected. A decision curve analysis was performed for the model obtained.

Results: Kaplan–Meier curves for biochemical free survival at ten years was 47.34% (95% CI 38.71–55.48%). It is important to note that in CPC negative men, the ten years Kaplan–Meier biochemical-free survival was 90.35% (95% CI 75.0–96.27) whereas in men who were primary CPC positive, the biochemical free survival rate was 30.00% (95% CI 20.34–40.60%). The Coxs´model to predict biochemical failure using transformed data with a power of minus one for the number of primary CPCs detected, showed a Harrell´s C concordance index of 0.74 and a decision analysis curve showing a net benefit of CPC detection over other risk factors to predict biochemical failure.

Conclusions: The number of primary CPCs detected before surgery permits a good prediction of subsequent biochemical failure in men undergoing radical prostatectomy as monotherapy for prostate cancer. Men negative for primary CPCs have a biochemical-free survival of
over 90% at ten years and should be considered for curative surgery.