Fred Hutch, Seattle Cancer Care Alliance, and UW Medicine Complete Restructure of Partnership

Learn More
News Releases

PSA testing alone cannot fully explain drop in deaths from prostate cancer, says Hutch researcher

The PSA cancer-screening test alone probably cannot explain the recent decline in deaths from prostate cancer, according to investigators at the Fred Hutchinson Cancer Research Center and the National Cancer Institute.

In a paper to appear in tomorrow's Journal of the National Cancer Institute, lead researcher Dr. Ruth Etzioni, a biostatistician at the Hutchinson Center, concludes that while prostate-specific-antigen, or PSA, blood testing may have played a partial role in the declining prostate-cancer death rate observed since the early 1990s, it is likely not the only cause.

"Our results suggest that we should interpret the population trends with caution," says Etzioni, an associate member of the Center's Public Health Sciences Division. "The average man on the street should not read too much into the declining mortality numbers. What we're seeing is probably not all due to PSA testing. There are, most likely, other things going on." Other factors that may contribute to the drop include changing patterns of treatment and treatment-related deaths, or misclassification of cause of death.

Etzioni's paper is one of three in this week's issue of the JNCI to address the role of PSA testing in reducing the prostate-cancer death rate. The others are by Drs. Benjamin Hankey and Eric Feuer of the National Cancer Institute, who also collaborated with Etzioni on her paper.

Used widely in this country since 1988, the PSA test measures the amount of prostate-specific antigen in the blood. This enzyme, produced by the prostate gland, can signal the presence of cancer when the level is significantly elevated.

Until quite recently, prostate-cancer mortality had long been on the rise. Between 1973 and 1991, for example, the age-adjusted death rate increased from 21 to 27 men per 100,000. However, in the early 1990s, the death rate started heading south, dropping to 24 men per 100,000 between 1992 and 1996.

As the declining mortality rate dovetailed with the increase in PSA screening, the natural assumption was that the dip in deaths could be attributed to the test.

To test this hypothesis, Etzioni collaborated with researchers at the National Cancer Institute to construct a computer model that would estimate the number of deaths prevented by PSA testing between 1988 and 1994, and project what the mortality would have been in the absence of such testing. The model, a sort of intelligent number-crunching device, used data from Medicare records to translate information at the level of the individual into population-level statistics.

Factors taken into account included the likelihood of annual PSA testing; the chances of cancer being detected by the test; the increase in life expectancy as a result of the test; and the diagnostic "lead time," or time by which diagnosis could be hastened by screening. Without PSA screening, for example, a man might be diagnosed with prostate cancer at age 68, but with screening doctors might catch the cancer at age 64, creating a lead time of four years.

Lead time is an important indicator of a cancer's aggressiveness. Short lead times generally are linked to faster-growing tumors and long lead times typically are associated with non-aggressive cancer that can go undetected for years.

Assuming that PSA testing significantly improves life expectancy, Etzioni and colleagues estimate that for PSA testing to wholly explain the observed mortality decline, the average lead time in the population would have to be quite short -- three years or less. Prostate cancer, however, typically is associated with a longer lead time, as the disease often goes undetected for many years prior to diagnosis.

While two previous studies suggested an average lead time ranging from three to five years, the results were based on subgroups of men who tend to have shorter lead times than the population as a whole, Etzioni says.

"Even if PSA testing improves life expectancy in a clinically meaningful way, we conclude that it is unlikely that the declines in mortality observed from 1992 through 1994 are solely due to PSA testing. From what we know about prostate-cancer progression, mortality declines of the order observed would likely take longer to become apparent," she says.

"It is likely that some other factors are playing a role in these declines, such as treatment innovations for advanced disease. The results do not rule out that PSA testing may be playing a partial role, but they strongly suggest that such screening is not the sole cause of the observed declines."

What does this mean for the man on the street? Should he still get an annual PSA exam starting at age 40 or 50, depending on his individual risk profile?

"This study does not and cannot address this question," Etzioni says. "Since we assume that PSA testing works for the individual in a clinically meaningful way, we do not address whether PSA testing is efficacious. Rather, we assume optimistically that it is, and ask instead how long it would take for this to impact the overall death rate from prostate cancer. We look forward to the results from clinical trials for more definitive answers about the efficacy of PSA testing."


# # #


The Fred Hutchinson Cancer Research Center is an independent, nonprofit research institution dedicated to the development and advancement of biomedical technology to eliminate cancer and other potentially fatal diseases. Recognized internationally for its pioneering work in bone-marrow transplantation, the Center's four scientific divisions collaborate to form a unique environment for conducting basic and applied science. One of 35 National Cancer Institute-designated comprehensive cancer centers in the nation, it is the only one in the Northwest. For more information, visit the Center's Web site at <>

Fred Hutchinson Cancer Research Center

Kristen Woodward
(206) 667-5095


Until 4 p.m. EDT (1 p.m. PDT) TODAY (Tuesday, June 15, 1999


There are no tags on this page. A list of tags will appear here once there are.