Long-term prostate-specific antigen contamination in the Spanish arm of the European Randomized Study of Screening...

Recently, the European Randomized Study of Screening for Prostate Cancer achieved a reduction in prostate cancer mortality by measuring serum prostate-specific antigen (PSA) levels. These results were not reproduced in the Spanish arm of European Randomized Study of Screening for Prostate Cancer. PSA contamination (opportunistic measurements outside the study) could decrease the study's contrasting power if performed in the control arm. We have calculated the long-term rate of PSA contamination and its effect on performing prostate biopsy and detecting cancer.

A total of 4,276 men were randomised (2,415 to the screening arm, 1,861 to the control arm) in the Spanish section of the European Randomized Study of Screening for Prostate Cancer. PSA measurements were not scheduled in the control arm. Sextant prostate biopsy was indicated if PSA levels were ≥3 ng/mL. All PSA readings performed outside the study were labelled as "PSA contamination". We calculated the rates of PSA contamination, biopsy implementation and cancer detection.

The median age and follow-up time were 57 and 15.1 years, respectively. A total of 2,511 men underwent at least one PSA reading outside the study. PSA contamination at 5, 10 and 15 years was 22.0%, 47.1% and 66.3% in the screening arm, respectively, and 20.8%, 43.2% and 58.6% in the control arm, respectively (P<.0001). The biopsy rate at 5, 10 and 15 years was 19.3%, 22.6% and 24.1% (screening), respectively, and 1.0%, 3.6% and 7.1% (control), respectively (P<.0001). The PC detection rate was 6.7% (screening) and 4.3% (control; P=.0006).

Although the cumulative PSA contamination was pronounced in the 2 study arms, the rate of prostate biopsies was low in the control arm. We therefore believe that the effect of PSA contamination on the study's statistical power should be limited.