To investigate the natural history of prostate-specific antigen (PSA) increase in men with and without prostate cancer to clarify the probability of cancer-related PSA increase.

Between 1986 and 2001, 504 men aged 79 years or younger with baseline PSA levels of 4.0 ng/mL or less and a PSA increase greater than 4.0 ng/mL on consecutive screening were enrolled in this study. The types of PSA increase were classified as “non-cancer-related PSA increase,” “suspicious cancer-related PSA increase,” and “cancer-related PSA increase.” The probability of a “cancer-related PSA increase” was investigated and stratified by baseline PSA levels and elapsed years until the PSA level increased to greater than 4.0 ng/mL.

The probability of a “non-cancer-related increase,” “suspicious cancer-related PSA increase,” and “cancer-related PSA increase” was 57%, 15%, and 28%, respectively. The PSA velocity before the PSA increase was not significantly different between those with and without prostate cancer. A “non-cancer-related PSA increase” was observed in 92% of those with a PSA increase within 2 years of baseline PSA ranges of 2.0 ng/mL or less. Regardless of elapsed years until a PSA increase to greater than 4.0 ng/mL, a “suspicious cancer-related PSA increase” or “cancer-related PSA increase” was observed in almost one half of those with baseline PSA levels of 2.1 to 4.0 ng/mL.

Intensive serial observations should be recommended before undergoing biopsy for those with a PSA increase within 2 years of a baseline PSA range of 0.0 to 2.0 ng/mL. It may be difficult to distinguish between those with and without cancer using only subsequent total PSA measurements for the remaining cases, and prostate biopsy should be recommended at present.