Since measuring prostate-specific antigen (PSA) levels became routine in the late 1980’s, the incidence of death from prostate cancer has declined significantly. Although part of this decline may be due to improved treatment methods, at least some of it probably resulted from the ability to detect prostate cancer at an earlier stage through PSA screening.

Initial research suggested that the risk of prostate cancer increased significantly in men with a PSA level above 4.0 ng/ml (nanograms per milliliter), making this level a reasonable threshold for “normal” PSA. But the risk of prostate cancer for men with PSA levels equal to or below 4.0 has never been well researched. And recent studies suggest that a sizable number of men with PSA levels between 2.5 and 4.0 ng/ml turn out to have prostate cancer.

A new study in the May 27, 2004 issue of the New England Journal of Medicine looked at the prevalence of prostate cancer among a group of men with PSA levels at or below 4.0 ng/ml and found that 15% had prostate cancer, despite “normal” PSA levels.

About the Study

This analysis included 2950 men between the ages of 62 and 91 who were all part of the Prostate Cancer Prevention Trial. As part of this seven-year study, the men had annual PSA measurements and rectal examinations. Men included in this analysis:

  • Never had a PSA level more than 4.0 ng/ml
  • Never had an abnormal digital rectal examination
  • Underwent a prostate biopsy after being in the study for seven years

The researchers compared PSA levels and cancer status based on prostate biopsies for these 2950 men.

The Findings

Fifteen percent (449) of the men had prostate cancer on their end-of-the-study biopsy. The average PSA value was 1.8 ng/ml among the men with cancer and 1.3 ng/ml among the men without cancer.

As the following table shows, the prevalence of cancer increased with increasing PSA levels, from 7% among men with PSA levels of 0.5 ng/ml or less, to 27% among men with PSA levels of 3.1 to 4.0 ng/ml. The severity of cancer also increased with increasing PSA levels. Overall, 15% of the men with cancer had high-grade cancers, characterized by a Gleason score of seven or greater.

PSA Level (ng/ml)Men with Prostate CancerMen with High-Grade Prostate Cancer
< 0.5 32 (7%)4/32 (12.5%)
0.6-1.080 (10%)8/80 (10%)
1.1-2.0170 (17%)20/170 (12%)
2.1-3.0115 (24%)22/115 (19%)
3.1-4.052 (27%)13/52 (25%)

How Does This Affect You?

This study shows that that prostate cancer—even advanced high-grade prostate cancer—can be present in men with low PSA levels, suggesting that a PSA value less than 4.0 ng/ml is not completely reassuring that a man is free of prostate cancer. Does this mean that the PSA test is useless? No. There is still a direct correlation between PSA level and prostate cancer, likewise a PSA level that increases from one year to the next, may be a sign that something is not right.

There has been some debate about whether or not the PSA threshold for obtaining a biopsy should be lowered to diagnose cancer that would otherwise go undetected. But even though some men with PSA levels turn out to have cancer at or below 4.0 ng/ml, it’s not at all clear that men who are treated when their cancers are detected at such low levels fair any better than those who are not treated until their PSA levels are higher. Furthermore, lowering the threshold for abnormal PSA will have the unfortunate consequence of many more biopsies being done on perfectly normal prostate glands.

Research is currently underway to find a more accurate way of detecting prostate cancer with or and without PSA. But until a better way exists, current PSA testing remains the best screening method. The American Cancer Society recommends that all men at average risk without serious health conditions have yearly PSA screenings beginning at age 50, and five years earlier if they have a family history of prostate cancer or are of African American descent.