Can We Reassure Elderly Men with Normal PSA Levels?

Abstract & Commentary

Synopsis: Of 2950 men age 62 to 91 years with consistent PSA levels of 4.0 ng/mL or less over 7 years, prostate cancer was diagnosed by biopsy in 449 (15.2%). Sixty-seven of these cancers (14.9%) were high-grade tumors (Gleason score of 7 or higher). Normal PSA levels in elderly men should not be cause for reassurance that cancer does not exist. A rise in PSA level at any range should be a cause for concern about cancer.

Source: Thompson IM, et al. N Engl J Med. 2004;350: 2239-2246.

Over the past 2 decades, the serum PSA level has been increasingly used to screen for prostate cancer, considered an accompaniment to digital rectal exam. Despite the lack of solid evidence that PSA screening saves lives, its use has become widespread.1 Once PSA screening became widespread in the United States, the rate of death from prostate cancer declined (50-70 percent decline between 1986 and 1999 among men 50 years and older).2 Even though improvements in treatment may partially explain the decline, there is widespread acceptance of the test among primary care physicians and patients.

PSA levels rise with prostate disease, both benign and malignant. The cutoff for normal has been established at 4.0 ng/mL to detect the majority of cancers while avoiding the high cost of false-positive tests. Even levels above 4.0 ng/mL turn out to be benign disease (prostatic hypertrophy or chronic prostatitis) more often than cancer. Ninety percent of men aged 50 to 92 years have PSA levels of 4.0 ng/mL or less, so consideration for prostate biopsies would only come up in 10% of screened men.3 Following a mildly elevated PSA level is done commonly to avoid the use of an invasive and uncomfortable biopsy.

How much can we reassure men with normal PSA levels (4.0 ng/mL or less)? The data from this large study show that such reassurance must be done with caution. The multicenter Prostate Cancer Prevention Trial gave Thompson and colleagues 2950 men with consistently normal PSA levels over seven years to undergo a prostate biopsy. This remarkable data set may never be repeated. Overall, in these men aged 62 to 91 years, 15.2% had prostate cancer on biopsy. There was no stratification with age, but there was by PSA level. When the PSA level was 0.5 ng/mL or less, the prevalence of prostate cancer was 6.6%. When the level was 0.6 to 1.0 ng/mL, the prevalence of cancer was 10.1%. When the level was 1.1-2.0 ng/mL, the prevalence of cancer was 17.0%. When the level was 2.1- 3.0 ng/mL, the prevalence was 23.9%. When the level was 3.1-4.0 ng/mL, the prevalence was 26.9%. The prevalence of high-grade tumors increased with PSA level, being 12.5% of the cancers with a PSA level of 0.5 ng/mL or less, and 25.0% of cancers with a PSA level of 3.1-4.0 ng/mL.

Comment by Joseph E. Scherger, MD, MPH

This study caused me much alarm. I recall the reassurance that I have given hundreds of men with normal PSA levels. My words will now be one of reassurance mixed with caution. A look at this data made me think the normal cutoff for PSA should be lowered, at least to 3.0 ng/mL. The editorial in the same issue by Carter reversed my thinking, and I highly recommend reading it.3 Carter was able to use previous data to show that these results would be expected in a population of 90% of men in this age range. He argues effectively that the cutoff range for PSA should not be changed from the current level of 4.0 ng/mL. The new information for me to use in clinical practice is the tracking of "PSA velocity" in the normal age range. PSA velocity refers to the rate of rise of PSA levels, which has been shown to correlate directly with the risk of cancer.4 Carter’s work suggests that a rise of 0.75 ng/mL or more in a year would indicate a significant risk for cancer. This study was based on small number men (38).

So, what do we tell men about their PSA levels and when should screening be done? The answers to these questions are not definitive and may never be. Hopefully soon we will have better biomarkers for prostate cancer than the current PSA test. Meanwhile, I will provide cautious reassurance to men with normal levels, and will repeat them yearly to look for a significant change.

Dr. Scherger, Clinical Professor, University of California, San Diego, is Associate Editor of Internal Medicine Alert.


1. Smith DS, et al. JAMA. 1996;276:1309-1315.

2. Chu KC, et al. Cancer. 2003;97:1507-1516.

3. Carter HB. N Engl J Med. 2004;350:2292-2294.

4. Carter HB, et al. JAMA. 1992;267:2215-2220.