Prostate Cancer Screening Can Lead to Overtreatment

July 26, 2010— -- Many men being treated aggressively for low-grade prostate cancer -- particularly if it was detected during prostate-specific antigen (PSA) screening -- are unlikely to benefit from the intervention, a new study suggests.

Men with screen-detected cancer and PSA levels below 4 ng/mL were less likely to have high-grade tumors, disease outside the prostate, or tumors larger than 0.5 cubic centimeters.

Yet, they were nearly 1.5 times more likely to undergo radical prostatectomy than men who had prostate cancer that was not detected by screening, Yu-Hsuan Shao of the Cancer Institute of New Jersey in New Brunswick and colleagues reported in the July 26 issue of Archives of Internal Medicine. They were also more likely to have radiation therapy.

Although the relative five-year survival for prostate cancer increased from 69 percent in 1975 to nearly 99 percent in 2003, concerns about overdiagnosis and overtreatment have arisen.

Because little is known about the risk profile of men whose PSA is below 4 ng/mL, Shao and colleagues looked at data from the Surveillance, Epidemiology, and End Results (SEER) database, which includes approximately one-quarter of the U.S. population.

Among 123,934 men who received a diagnosis of prostate cancer from 2004 to 2006, 14 percent had PSA values below 4 ng/mL, 73.5 percent were between 4.1 and 20 ng/mL, and 12.5 percent had levels above 20 ng/mL.

Approximately 54 percent of patients with PSA levels below 4 ng/mL had low-risk cancers, with risk being determined according to clinical stage, PSA level, and Gleason score, compared with 48 percent of those whose PSA levels were between 4.1 and 10 ng/mL.

A total of 44 percent of men whose PSA was lower than 4 ng/mL had a radical prostatectomy, as did 38 percent of those whose levels were 4.1 to 10 ng/mL and 24 percent of those whose values were between 10.1 and 20 ng/mL.

Radiation was given to 33 percent, 40 percent, and 41.3 percent of the three groups, respectively.

"Despite their lower risk of having clinically significant disease, treatment rates for men with PSA values of 4.0 ng/mL or lower were comparable to those of men presenting with PSA values between 4.0 and 20.0 ng/mL," wrote Shao and colleagues.

In an invited commentary that accompanied the study, Dr. Richard M. Hoffman of the University of New Mexico in Albuquerque and Steven B. Zeliadt of the University of Washington in Seattle, argued in favor of active surveillance for patients at low risk.

They defined active surveillance as monitoring by PSA testing and digital rectal exam every three to six months and biopsy every one to two years, and recommended it as an "acceptable alternative" for men with a PSA level of 10 ng/mL or lower, a Gleason score of six or lower, and clinical disease stage T1c or T2a.

"Prostate-specific antigen testing has led to an epidemic of prostate cancer, but a substantial proportion of PSA-detected cancers will never be clinically significant, and continuing to aggressively treat most men with low-risk cancers will certainly do more harm than good," Hoffman and Zeliadt wrote.

A different view was offered by Dr. William J. Catalona of Northwestern University Medical School in Chicago, who pointed out several failings in the study, including the fact that information was not available on how rapidly patients' PSA levels were rising (PSA velocity), how large the prostate was in relation to the PSA level (PSA density), the percentage of free PSA (lower with aggressive cancers), or the number of biopsy cores positive for cancer.

"All of these are important indicators of tumor aggressiveness and determinants of treatment selection," Catalona, who developed the PSA screening test, wrote in an e-mail to MedPage Today and ABC News.

"The authors and the editorial comment strongly advocate active surveillance," Catalona continued. "However, with active surveillance protocols, the 'good news' comes early and the 'bad news' comes late."

It has already become clear that 30 percent to 50 percent of patients abandon active surveillance for treatment within five years, and for these patients active surveillance has meant delayed cancer treatment," he said.

"There is no doubt that with active surveillance, some patients will die of prostate cancer unnecessarily. In a sense, widespread adoption of active surveillance in all patients who appear to have low-risk disease would be to say it is okay for some men to die of prostate cancer so that others may avoid treatment -- a debatable choice," Catalona warned.