1. Home>
2. Specialties>
3. General Medicine>
4. Summary and Comment

Another Salvo in the PSA Controversy

Should the PSA cutoff for biopsy be lowered to 2.5 ng/mL? These authors argue otherwise.

Many men with prostate-specific antigen (PSA) levels lower than the traditional cutoff (4 ng/mL) for biopsy have occult prostate cancer (Journal Watch Jun 8 2004), leading some authorities to argue that the threshold should be lowered. Dartmouth researchers used PSA data from the National Health and Nutrition Examination Survey (2001–2002), and other data on prostate cancer treatment and mortality rates, to estimate the effects of lowering the PSA cutoff to 2.5 ng/mL.

At the 4 ng/mL cutoff, about 1.5 million American men (age range, 40–69) would be labeled as having abnormal PSA levels. If the cutoff were lowered to 2.5 ng/mL, an additional 1.8 million men would be labeled as having abnormal levels. If those 1.8 million men underwent biopsies, 450,000 would be diagnosed with prostate cancer. If all 450,000 underwent radical prostatectomy, 180,000 would become impotent, 40,000 would have at least moderate incontinence, and 1000 would die from the surgery.

Among men aged 50 to 59, 11% have PSA levels higher than 2.5 ng/mL, but only 0.3% are likely to die from prostate cancer in the next 10 years (according to statistics from the pre-PSA era). Among men aged 60 to 69, these figures are 17% and 0.9%, respectively.

Comment: These researchers note that only a tiny fraction of men with PSA levels higher than 2.5 ng/mL will die from prostate cancer during the next 10 years, but large numbers of these men would be subjected to morbidity associated with surgery. The authors, who clearly are skeptical about PSA screening, conclude that lowering the PSA threshold "would be a mistake."

— Allan S. Brett, MD

Published in Journal Watch General Medicine August 26, 2005

Citation(s):

Welch HG et al. Prostate-specific antigen levels in the United States: Implications of various definitions for abnormal. J Natl Cancer Inst 2005 Aug 3; 97:1132-7.

• Original article (Subscription may be required)
• Medline abstract (Free)