Single PSA test predicts prostate cancer risk
Friday, 25 May 2012- – A baseline prostate-specific antigen (PSA) value of less than 1 ng/mL predicted a low risk of prostate cancer over the next 15 years in men younger than 50, none of whom developed intermediate- or high-risk cancer, data from a longitudinal cohort study showed.
During a median follow-up of 16.8 years, six prostate cancers occurred in men with baseline PSA values <1 ng/mL. All of the tumors had low-risk characteristics.
Men with low PSA values at initial measurement had a prostate cancer incidence of <1% at age 55 and <3% at 60, Dr. Christopher Weight reported at the American Urological Association (AUA) meeting held in Atlanta, Georgia, USA.
"Baseline PSA in men younger than 50 can risk-stratify men for prostate cancer," Weight, a urology resident at Mayo Clinic in Rochester, Minnesota, USA, said during an AUA press briefing. "Men with a baseline PSA ≥1 ng/mL have a substantial risk of subsequent biopsy and cancer diagnosis and should be followed annually.
"Men with a baseline PSA <1.0 ng/mL in their 40s appear to be able to safely avoid annual screening until age 55."
Since 2009 the AUA has recommended PSA testing as a screen for prostate cancer beginning at age 40. However, limited population-based screening data exist to characterize trends in prostate cancer in the younger age group and to assess the impact of PSA testing on the trends.
Since 1990 investigators at Mayo Clinic have prospectively followed a random sample of men ages 40 to 49 from Olmsted County, Minn. Each man had PSA assessment, digital rectal exam (DRE), and transurethral ultrasound at entry and then biennially thereafter.
Weight said investigators have examined three principal outcomes during follow-up of the cohort:
• Risk of prostate biopsy and prostate cancer
• Stage and Gleason score of diagnosed cancers
• Long-term prostate cancer risk in men with a baseline PSA <1 ng/mL at age 40 to 49
The initial PSA value was <1 ng/mL in 192 men and ≥1 ng/mL in 76 men. The two groups did not differ significantly with respect to DRE findings or family history of prostate cancer.
The six cases of diagnosed prostate cancer in the low-PSA group translated into an incidence rate of 1.6 per 1,000 patient-years. In contrast, men who had an initial PSA measurement ≥1 ng/mL subsequently developed 12 cancers, resulting in an incidence rate of 8.3 per 1,000 patient years.
Two men with higher baseline PSA values developed intermediate- to high-risk cancers as compared with none in the low-PSA group. The median time to cancer diagnosis was 14.6 years in the low-PSA group and 10.3 years in men with higher baseline PSA measurements.
Analysis of receiver operating characteristic area under the curve (AUC) showed that a baseline PSA cutoff of 1 ng/mL was associated with an AUC of 0.73 for diagnosis of any prostate cancer and an AUC of 0.85 for diagnosis of Gleason score ≥7 cancers.
The 1 ng/mL cutoff also proved useful for stratifying patients into low- and high-risk subgroups for prostate biopsy, said Weight.
The press briefing occurred within the context of the U.S. Preventive Services Task Force final recommendation against PSA testing as a prostate cancer screen. Weight alluded to the recommendation in his comments.
"There is danger in completely throwing out the PSA, and I think that is illustrated here," he said. "There are very few tests that, with one single blood test, can produce an area under the curve this high."
"I think we can potentially use PSA a little differently than we have in the past and eliminate some of the overdiagnosis and overtreatment by risk-stratifying better, which will also help maintain the benefits of PSA screening."
Press briefing moderator Scott Eggener said the risk-stratification strategy suggested by the results offers a reasonable means to balance the benefits and potential harms of PSA screening.
"The goal of any screening test -- or many things we do in medicine -- is to try to identify the cohort of patients who are most likely to benefit and give them the test; hopefully we will then screen, diagnose early, treat early, and cure the cancer," said Eggener, of the University of Chicago.
"The other goal is to identify people who don't need intensive screening or don't need screening at all, and then stand back a little bit to maximize the benefit and minimize the harm."
Source: C. Weight, et al., "Young men (ages 40-49) with a single baseline PSA below 1.0 ng/mL are at very low 10-15 year risk of prostate cancer," American Urological Association Annual Meeting, 2012; Atlanta, Georgia; Abstract 1219; MedPage Today