Prostate cancer test still not foolproof
By Dr. V. Upender Rao
Published September 11, 2006
Prostate cancer is one of the most frequently diagnosed cancers in American men, second only to superficial skin cancer. It is estimated that one in six American men will develop prostate cancer in his lifetime.
Since the introduction of PSA (prostate specific antigen) as a screening test in the 1990s, the incidence of prostate cancer has risen by 300 percent. There is, however, no proportionate increase in the death rate. As a matter of fact, the mortality rate for prostate cancer has not changed since 1994.
This means that a large number of cancers diagnosed by a PSA screen-prompted biopsy carry little clinical relevance. These cancers meet the microscopic criteria set by pathologist for the diagnosis of cancer but do not behave like cancer.
These men live with the diagnosis of prostate cancer, which never becomes clinically significant, and eventually die due to a small risk of complications of diagnostic- and treatment-related procedures or unrelated cause.
This is exemplified by a study by W.A. Sakr, D.J. Grignon and J.D. Crissman published in 1994. They examined the prostate glands of men undergoing autopsy for various causes of death unrelated to prostate cancer.
They found that 2 percent of men in their 30s and 30 percent of men in their 40s had some focus of cancer. By age 70, almost 50 percent had abnormal microscopic findings consistent with cancer. These were latent cancers that did not have much clinical significance and would be diagnosed only by a PSA-prompted biopsy.
The concern among urologic oncologists is that a simple PSA test may have led to overdiagnosis and possible overtreatment of prostate cancer.
In order to fine-tune the PSA and to increase its predictive value, they initially designated a PSA level of 4ng/dl as a cutoff for biopsy. However, 50 percent of these men had PSA increases over the following four years to more than 4 ng/dl, and 13 percent had prostate cancer on biopsy.
Another study of 2,950 men showed that 15 percent with a PSA of less than 4ng/dl have cancer on biopsy. In this study, 10 percent of the cancers were seen with a PSA of less than 1, 17 percent with a PSA of less than 2, 23 percent with a PSA of less than 3 and 23 percent with less than 4ng/dl of a measured PSA value.
So they lowered the PSA cutoff level to 2.6ng/dl in an effort to increase the sensitivity, which unfortunately will pick up more of both significant and insignificant cancers.
Researchers therefore introduced a number of variations on the basic PSA test in an effort to increase its predictive value. These include PSA volume, PSA velocity, TZ-PSA (transitional zone PSA), free PSA and free PSA to total PSA ratio.
None of these, however, is reliable enough to negate the continuing possibility of overdiagnosis and overtreatment of clinically irrelevant prostate cancers.
Newer biological markers and molecular footprint analysis of biopsy specimen of prostate cancer are underway.
PSA was first isolated in 1970 and was characterized as a protein involved in the synthesis of seminal fluid nine years later in 1979. In 1980, its detection in the serum as a simple blood test was perfected and its utility as a possible screening test was identified.
However, it was soon realized that it was not a specific marker for prostate cancer since it could be influenced by nonmalignant conditions of the prostate such as inflammation, benign enlargement, urologic manipulation, ejaculation and treatment with finasteride.
Thirty-six years after its isolation, PSA is still not a precise tool for the detection of clinically relevant prostate cancer. It is, however, widely used by physicians and hospitals and recommended by well-meaning patient advocacy groups.
V. Upender Rao MD, FACP, practices at the Cancer and Blood Disease Center in Lecanto.