A Step Backward: The ACPM Recommendations on Prostate Cancer Screening

William J. Catalona, MD; Stacy Loeb, MD


April 21, 2008

After reviewing the literature prior to July 2007, The American College of Preventive Medicine (ACPM) concluded that there is insufficient evidence to recommend routine prostate cancer screening with prostate-specific antigen (PSA) testing and digital rectal examination (DRE).[1] Rather, they recommended that clinicians have an annual discussion with their patients about the potential risks and benefits of prostate cancer screening and the limitations of the evidence currently available. Furthermore, routine testing is not even recommended for high-risk groups, such as men with a positive family history and African-American men, and if the patient defers to the physician concerning the decision to screen, testing should not be offered.

Several other organizations have adopted a similar position, including the American Academy of Family Physicians, US Preventive Services Task Force, and American College of Physicians.[2] These groups remain unconvinced that prostate cancer screening is worthwhile, despite the overwhelming emerging evidence that screening saves lives.

The rationale for the ACPM position involved a discussion of 2 randomized trials as well as case-control association studies that did not show a benefit.

They also expressed concern about false-negative and false-positive results, as well as the possible adverse psychological effects of screening on patients. Furthermore, they dismissed the use of various PSA-based adjunctive parameters to improve the accuracy of screening, because they found no evidence of a mortality benefit from using these variables.

The ACPM emphasized the high financial cost associated with the implementation of a national screening program and alluded to the additional costs of treatment in subsequent years. However, they did not consider the favorable cost trade-offs of treating early prostate cancer vs advanced-stage disease.

Despite an acknowledgment that screening results in earlier cancer detection and has the potential to decrease mortality rates, they emphasize that it also has the potential to increase treatment-related morbidity. They point out the benefits of screening are unproven and may not be realized due to the growing prevalence of indolent tumors.

They cite studies showing that watchful waiting is inferior to active treatment, especially for men younger than 65 years old. Yet they deny that early detection through screening combined with effective treatment can decrease mortality rates. Overall, the ACPM position is that more men may potentially be harmed than helped.

The reality is that until there is an effective way to prevent prostate cancer altogether or to cure advanced-stage disease, the only practical strategy for reducing death from prostate cancer is to diagnose it early and treat it effectively. Because prostate cancer arises silently and typically passes into an incurable stage without any symptoms, the only way to detect it early is through proactive screening.

Prostate cancer screening is widespread throughout the world. Approximately 57% of US men older than age 50 and 87% of male physicians in this age range have had a PSA test.[3] However, physicians and patients alike need guidance about how best to implement screening.

Organizations, such as the ACPM, periodically review what they consider the highest level evidence on the subject, which remains very limited. However, additional consideration has not been given to the overwhelming body of emerging evidence that screening with PSA and DRE saves lives. Accordingly, they have issued a misguided recommendation that there is insufficient information to recommend for or against PSA screening.

In the meantime, physicians and patients who are concerned about preventing death from prostate cancer have proceeded with screening based on the imperfect, but growing, body of evidence currently available. An increasing number of clinicians are using PSA velocity, PSA density, PSA isoforms, and age-adjusted PSA values to reduce the false-positive and false-negative rates. Indeed, numerous studies have shown that these parameters help to decrease the number of unnecessary biopsies and also to identify men with the most aggressive tumors so that they can receive timely treatment.[4,5,6]

In contrast to the ACPM, the American Urological Association recommends offering PSA screening beginning at age 50 in men with at least a 10-year life expectancy.[7] African-American men and those with a first-degree relative with prostate cancer are urged to consider screening at an earlier age. The American Cancer Society makes similar recommendations.[8]

In our view, the National Comprehensive Cancer Center (NCCN) Guidelines on the Early Detection of Prostate Cancer are the most progressive, and provide the most helpful guidance to physicians and patients.[9] Unlike many other guidelines, they incorporate much of the emerging information, including some that has not yet been validated in a randomized controlled clinical trial.

The NCCN guidelines start from the premise that the patient has made a decision to seek early prostate cancer detection. They recommend beginning screening at age 40. The baseline PSA level, race, and family history are then used to determine the subsequent screening intervals. They recommend considering a biopsy for men with a total PSA level > 2.5 ng/mL, after further consideration of the PSA velocity, PSA density, and percent free PSA. They explain how these parameters can be used to lessen the possibility of confounding from benign prostatic hyperplasia. Furthermore, they describe how repeating PSA determinations with or without a trial of antibiotic therapy, as well as consideration of variability between different PSA assays, can reduce the likelihood of confounding from prostatitis or differences in PSA assay standardization. Moreover, they provide advice about whether or not repeat biopsies are needed and how to deal with the findings of high-grade prostatic intraepithelial neoplasia or atypical glands suspicious for carcinoma on an initial biopsy.

What is the important emerging information that the ACPM found unconvincing? There has been a 75% reduction in the rate of metastatic disease at diagnosis since the beginning of the PSA era.[10] The relative 5-year survival rate of patients with prostate cancer has increased from approximately 75% before the PSA era to 99% in the PSA era (a larger increase than for any other cancer site).[11] Prostate cancer death rates have decreased 35% in the United States in the PSA screening era, and statistical modeling studies suggest that 45% to 70% of this decrease is due to PSA screening.

There is high-quality epidemiologic data from 30 population-based US cancer registries, representing 65% of the US population.[12] These data showed that in regions where PSA testing was more prevalent, there was less late-stage disease and a lower prostate cancer mortality rate.

On a global scale, the World Health Organization data similarly show that the age-adjusted prostate cancer mortality rates have decreased in countries where PSA screening is widespread and have remained stable or increased in countries where it is not widely used. Particularly striking contrasts have been reported in Austria and among the Scandinavian countries.[13,14]

PSA testing provides a powerful marker for prostate cancer risk and aggressiveness, including the risk of dying from prostate cancer. Generally, the higher the PSA and the more rapidly it is increasing, the greater the risk of high-grade disease and the prostate cancer-specific mortality.[4]

Indeed, longitudinal studies with PSA measurements and PSA velocity from men in their 30s and 40s have linked this with the risk of subsequent prostate cancer diagnosis and cancer-specific survival many years later.[5,15,16] Moreover, prostate cancers treated at lower PSA levels have better outcomes than those treated at higher PSA levels.[17]

Reports of over-diagnosis are exaggerated, and prostate cancer continues to be diagnosed more often too late than too early. Most PSA-detected cancers have the histopathologic features of significant cancers. In addition, there are numerous parameters that are useful in selectively identifying clinically significant cancers, including the Gleason grade and extent of cancer in the biopsy specimens, as well as PSA density and PSA velocity. If screening detected only harmless cancers, treating them could not produce the striking decline in prostate cancer mortality rates that have occurred.

There is evidence from a randomized trial that convincingly demonstrates that treatment of early stage disease is effective. Similar findings were reported using Medicare data, wherein treatment of early prostate cancer reduced prostatecancer-specific mortality. One of the widely publicized negative case-control studies[18] does not have adequate follow-up, and the researchers did not examine prostate cancer-specific mortality. Policy decisions regarding screening should not be decided one way or the other on the basis of such studies. In contrast, a recently reported case-control study[19] showing a 62% reduction in prostate cancer-specific mortality in men younger than the age of 65 has not received wide publicity or attention in deciding policy.

Regardless of the ACPM recommendations, the prostate cancer screening train has left the station. Should we stop screening to prevent side effects? Should we go back to a 20% rate of metastatic disease at diagnosis and a 35% higher mortality rate? Or, rather, should we continue to strive to improve the accuracy of screening and the quality of treatment?

The senior author recommends that annual screening should be initiated at age 40 with informed patient management, to avoid under-diagnosis. Aggressive cancers can be identified through the combined use of PSA-based parameters and biopsy findings. Rather than eliminating screening and the possibility for early diagnosis in healthy men, we should combat the risks of over-diagnosis through careful patient selection and the provision of high-quality treatment. Clearly, avoiding 27,000 prostate cancer deaths per year is worth the tradeoffs.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.