SAN FRANCISCO—Three studies presented at the 2006 Prostate Cancer Symposium and discussed at a press briefing highlight the usefulness, as well as the limitations, of PSA velocity (PSAV) in detecting prostate cancer.
"Age-adjusted PSA velocity could improve cancer detection," said Judd W. Moul, MD, chief of urological surgery at Duke University Medical Center and lead author of a study that investigated the usefulness of age-normalized PSAV as a screening tool (abstract 1). Dr. Moul and his colleagues hypothesized that since younger men will live with prostate cancer for a longer time, the PSAV rates used to detect disease should be lower in these men. The researchers assessed 145,593 PSA values from 11,347 men with and without biopsy-proven prostate cancer who underwent PSA testing at Duke University Medical Center between January 1988 and February 2005. Setting the PSAV threshold at 0.25 ng/mL/yr for men between the ages of 40 to 59 years led to prostate cancer detection in an additional 37 men (35% detection rate), compared with the traditional PSAV cutoff of 0.75 ng/mL/yr (19% detection rate). Sensitivity was 51.9% for the lower threshold vs 26.5% for the traditional value.
Setting the PSAV threshold at 0.50 ng/mL/yr for men between the ages of 60 and 69 resulted in detection of prostate cancer in an additional 18 men (57% detection rate vs 25% with the traditional threshold). Sensitivity was 39.8% for the lower threshold vs 30.6% for the traditional threshold. Specificity decreased slightly with use of the lower threshold, from 93.5% to 84.1% in the younger age group and from 90.7% to 77.3% for the older age group. In men age 70 and older, the traditional PSAV threshold remained an accurate predictor of prostate cancer.
"Finding that PSAV is more effective when it is age adjusted is very important, especially since the oldest of the Baby Boomers turn 60 this year, and the youngest Boomers are now considering screening," Dr. Moul said. He recommended the following age-adjusted PSAV cut-offs: 0.40 ng/mL/yr for age 40-59; 0.60 for age 60-69; and 0.75 for age 70 and older, because they balance sensitivity, specificity, and positive predictive values.
Katja Fall, MD, PhD, of Sweden's Karolinska Institute, presented results from the Scandinavian Prostate Cancer Group Study IV (abstract 2). Attempting to determine criteria that would identify risk of aggressive disease, these researchers analyzed PSA levels in 297 men with localized prostate cancer (T1b-c, T2) who were randomly assigned to watchful waiting. With a mean follow-up of 7.8 years, 18 (6%) were alive but had developed distant metastases, and 37 (12%) had died from prostate cancer. Both initial PSA level and PSAV during the first 2 years after diagnosis were strongly correlated with risk of death and metastases: The risk of aggressive disease increased by about 5% for each 2 ng/mL/yr increase in initial PSA: 21% of men with fast-rising PSA developed metastatic disease or died within 6 years of diagnosis vs 8% of men with slow-rising PSA. However, the researchers were not able to determine a specific cut-off point for initial PSA or PSAV, with combined high sensitivity and specificity, that predicted outcome. "In view of the observed low accuracy of initial PSA and PSAV as predictors of outcome, additional decision-making tools are needed to aid in choosing treatment for patients with newly diagnosed prostate cancer," Dr. Fall said.
Scott Eggener, MD, of Memorial Sloan-Kettering Cancer Center, and colleagues, evaluated the relationship between PSAV and detection of prostate cancer or prostatitis on biopsy (abstract 4). They used linear regression to determine PSAV in the year prior to biopsy in 1,851 men who had an initial prostate biopsy between 1991 and 2001. One-fourth of these men were diagnosed with prostate cancer upon initial biopsy, and 6% were diagnosed with prostatitis. While men with a modest increase in PSAV of less than 1 ng/mL/yr had an increased risk of prostate cancer detected on initial biopsy, larger increases were also associated with prostatitis. An increase of more than 4 ng/mL/yr was just as likely to indicate prostatitis as prostate cancer.
