American Urological Association (AUA) Guideline on prostate cancer detection: process and rationale

Authors

  • H. Ballentine Carter

    Corresponding author
    1. The Johns Hopkins University School of Medicine, Department of Urology, The James Buchanan Brady Urological Institute, Johns Hopkins Hospital, Baltimore, MD, USA
    • Correspondence: H. Ballentine Carter, Department of Urology, Marburg 145, Johns Hopkins Hospital, 600N Wolfe Street, Baltimore, MD 21287-2101, USA.

      e-mail: hcarter@jhmi.edu

    Search for more papers by this author

Abstract

To review the process and rationale for the American Urological Association (AUA) guideline on prostate cancer detection. The AUA guideline on detection of prostate cancer involved a systematic literature review of >300 studies that evaluated outcomes important to patients (prostate cancer, incidence/mortality, health-related quality of life, diagnostic accuracy and harms of testing). A multidisciplinary panel interpreted the evidence and formulated statements to assist the urologist and the asymptomatic average-risk man in decision-making about prostate cancer detection. Other than prostate-specific antigen (PSA)-based prostate cancer screening, there was no evidence to address the outcomes of interest to patients. The strongest evidence that benefits may outweigh harms was in men aged 55–69 years undergoing PSA-based screening. This led the panel to recommend shared decision-making for these men at average risk, but recommend against routine screening for other age groups at average risk. Further, to reduce the harms associated with screening (false positive tests, over diagnosis, over treatment), the panel recommended against annual screening for those who choose to be screened. A panel under the auspices of the AUA recommended shared decision-making for the average risk asymptomatic man aged 55–69 years considering PSA-based screening for prostate cancer detection.

Abbreviations
ERSPC

European Randomized Study of Screening for Prostate Cancer

PLCO

the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial

The Process

The AUA has been a leader in the development of evidence-based guidelines for clinical practice [1]. A nine-step process over a ≈2-year period is required for current guideline development that follows the Institute of Medicine principles for creating a trustworthy guideline [2]. In 2011, the Practice Guidelines Committee of the AUA identified the topic of ‘Prostate Cancer Detection’ as a focused and addressable topic for which a guideline could reduce unnecessary practice variation and improve the quality of patient care.

A multidisciplinary panel consisting of expertise from radiation and medical oncology, cancer epidemiology, general internal medicine, health policy, and urology defined the focus of the guideline in 2011, as a review of the evidence for detection of prostate cancer among asymptomatic men at average risk for prostate cancer (where symptoms imply the presence of symptoms suggestive of locally advanced or metastatic disease). The panel identified four a priori outcomes of interest relevant to the diagnosis or screening of prostate cancer:

  1. Incidence/mortality.
  2. Health-related quality of life.
  3. Diagnostic test accuracy.
  4. Harms of testing.

A systematic literature review spanning 1995 to 2013 included search terms focused on prostate cancer detection with DRE, serum biomarkers (PSA, PSA isoforms, PSA kinetics, free PSA, complexed PSA, proPSA, prostate health index, PSA velocity, PSA-doubling time), urine biomarkers (prostate cancer antigen 3, TMPRSS2:ERG fusion), imaging (TRUS, MRI, magnetic resonance spectroscopy, MR-TRUS fusion), genetics (single nucleotide polymorphisms), shared decision-making and prostate biopsy. The methodology team identified only PSA as a test that evaluated the a priori outcomes of interest. Thus, the framework and purpose for the guideline was an evidence review of PSA-based prostate cancer screening, interpretation of the evidence by the panel, and formulation of guideline statements to assist the physician and patient facing choices about prostate cancer detection to reduce prostate cancer mortality. Evidence interpretation was from a clinical perspective of the patient/physician interaction considering patient values and preferences, rather than a public health perspective.

The methodology team and panel rated the quality of the evidence as A (high), B (moderate), C (low) based on the strength of the evidence and certainty of the results for >300 published studies that addressed the topic of interest. In the process of formulating the statements for the guideline, the evidence quality was linked to the type of statement [3]. The panel used three types of statements for the Prostate Cancer Detection guideline: (i) Standard, (ii) Recommendation, and (iii) Option.

A ‘standard’ and a ‘recommendation’ are directive statements suggesting that the benefit of taking or not taking an action is greater or less than the harm; with a standard supported by evidence of higher quality (A or B) as compared with a recommendation (C). An ‘option’ is a non-directive statement where the balance of benefit to harm is equal or not clearly supported by evidence of high or low quality (A, B, or C).

The guideline addressed the average risk asymptomatic man without risk factors including a family history of prostate cancer or African-American race. The panel chose to categorise statements by age (<40, 40–54, 55–69, ≥70 years), as the ratio of benefit to harm with PSA-based prostate cancer screening is age dependent, and the available evidence sources addressed specific age distributions.

Several caveats are worth mentioning before discussing the statements and the underlying evidence base supporting the statements. First, the AUA guideline was not a reaction to the USA Preventive Services Task Force grade D for PSA-based prostate cancer screening that discouraged the use of the service [4]. The Task Force statement was published in July 2012, 1 year after the AUA guideline panel began deliberations. Second, the present guideline differs from the PSA Best Practice Statement published in 2009, in that the Best Practice Statement addressed the use of PSA for detection, risk stratification, and management; did not include a systematic literature review; and was based on clinical experience and expert opinion [5]. In contrast, the present guideline was based on evidence rather than values, opinions, or clinical experience [6].

Guideline Statements: Rationale and Comments

Guideline Statement 1: Recommend Against PSA Screening in Men Aged <40 Years. (Recommendation; Evidence Strength Grade C)

The panel concluded that PSA screening in this age group was not appropriate for three reasons: (i) low prevalence of prostate cancer [7], (ii) absence of evidence for any benefit of screening given that randomised trials did not assess this age group, and (iii) the same harms of screening, i.e. biopsy risks and treatment-related harms, as in any age group.

Guideline Statement 2: Routine Screening in Men Aged 40–54 Years at Average Risk Is Not Recommended (Recommendation; Evidence Strength Grade C)

This statement is likely the most contentious and elicited the most controversy from reviewers. When urologists treat men in this age group for high-grade cancer that was detected because a routine PSA test was elevated or rising, it reinforces what seems intuitive – testing in this age group must be beneficial. The panel did not explicitly recommend that screening in this age group be actively discouraged and recognised that some men in this age group, particularly those at higher than average risk due to family history or race [8], may benefit from screening. For these men, decisions regarding screening should be individualised.

The two largest randomised trials of PSA-based screening [9, 10] did not include men aged 40–54 years and thus could not provide evidence of benefit, but the harms of screening remain high including biopsy related complications and those associated with treatment. One estimate of the benefit of screening in this age group was 0.004 prostate cancer deaths averted per 1000 men screened over a decade [11]. In addition, a recommendation against routine screening in men aged 40–54 years is not a recommendation against screening per se, but a benefit to harm assessment of beginning screening before the age of 55 years vs initiating at the age of 55 years for those who wish to be screened.

It has been documented that a single PSA test before the age of 50 years can risk stratify men with respect to the probability of detection of aggressive prostate cancer decades later [12]. But this finding does not answer the question the panel addressed – whether or not the benefits of screening in this age group outweigh the harms. The low prevalence of lethal prostate cancer in this age group [13], longer lead times among younger men as compared with older men [14], and the longer time during which side-effects of treatment would be realised for a younger man compared with an older man, all make it less likely that benefits outweigh harms.

Guideline Statement 3: Recommend Shared Decision-making for Men Aged 55–69 Years That Are Considering PSA Screening and Proceeding Based on a Man's Personal Values and Preferences

The highest level of evidence for the benefits of screening was in this age group, where an estimate of one death averted per 1000 men invited for screening was found in the European Randomized Study of Screening for Prostate Cancer (ERSPC) over a decade (relative risk reduction of 21%) [15]. Recognising that benefits could be larger with longer follow-up beyond a decade, the panel concluded that in men aged 55–69 years benefits could outweigh harms, and thus a decision to screen or not should consider a man's personal values and preferences (e.g. shared decision-making).

The panel weighted the quality of the evidence from the ERSPC higher than that of the Prostate, Lung, Colorectal, and Ovarian cancer screening trial (PLCO) because the ERSPC [10] addressed the question of interest – screening vs little or no screening, as compared to the PLCO [9] in which the rate of background screening in the USA population prevented such a comparison.

The Swedish (Göteborg) screening trial was designed separately and published independently [16], but the results were included in the ERSPC publications [10, 15]. With follow-up to 14 years among men aged 50–64 years, the relative reduction in prostate cancer death was 44% with an absolute reduction of four deaths per 1000 men screened [16]. It has been argued that as compared with the ERSPC overall, the younger age of men, longer follow-up, and shorter screening intervals of 2 years as compared with 4 years, was responsible for the greater reduction in prostate cancer mortality. An alternative explanation is that deaths from prostate cancer in Sweden are higher than in other countries that were part of the ERSPC [17], resulting in a greater benefit of screening. Of note, the Finnish section of the ERSPC, the largest trial making up the ERSPC, recently reported no significant reduction in cancer-specific mortality at 12 years of follow-up [18]. Higher pre-screening rates in Finland as compared with Sweden may explain these findings in part [18]. Nevertheless, the panel was not guided by the results of any single country making up the ERSPC, but rather the overall results.

Given the modest cancer-specific mortality reduction in the ERSPC [15], the potential for greater benefit with longer follow-up, as well as the potential for harm (false positive tests, prostate biopsy complications, over diagnosis and treatment-related side-effects) [4], the panel emphasised shared decision-making as a prerequisite to PSA-based screening – a discussion that should include the likelihood that screening would be beneficial based on age and health state. Further, the panel discouraged screening in settings where shared decision-making would be unlikely to occur, i.e. health fairs.

Guideline Statement 4: To Reduce the Harms of Screening, a Routine Screening Interval of ≥2 Years May Be Preferred Over Annual Screening in Those Men Who Have Participated in Shared Decision-making and Decided on Screening. As Compared with Annual Screening, It Is Expected That Screening Intervals of 2 Years Preserve Most of the Benefits and Reduce Over Diagnosis and False Positives. (Option; Evidence Strength Grade C)

Annual testing has been the norm in the USA since the onset of widespread PSA-based screening. Carter et al. [19] suggested that biennial screening for most men would reduce the burden of screening (false positive tests, unnecessary biopsies) while maintaining the identification of curable disease. Modelled simulations based on national and trial data suggest that as compared with annual screening, biennial screening will preserve benefits (deaths prevented) while reducing false positive tests and over diagnosis rates [20]. The results from randomised screening trials also support the concept that a longer than annual screening interval is unlikely to reduce benefits. For example, in PLCO [9], the control group underwent opportunistic screening with an average of tests biennially; whereas the intervention group underwent PSA screening at annual intervals. The absence of a difference in mortality between the groups suggests that annual screening will not improve health outcomes as compared with every 2 years. Further, in the only trial to show a benefit of PSA-based screening [10, 15], re-screening intervals were 2–4 years. Thus, the body of evidence suggests that re-screening intervals longer than yearly will reduce harms but not benefits.

Data support the concept that re-screening intervals can be individualised based on PSA levels. In the Baltimore Longitudinal Study of Aging, PSA levels below the median for age in men in their 40s and 50s were predictive of a lower risk of prostate cancer 2–3 decades later as compared with PSA levels above the median for age [21]. Subsequent studies have confirmed that a re-screening interval of 4 years is not likely to miss a curable prostate cancer among men with a PSA level <1.0 ng/mL [22, 23].

Guideline Statement 5: Screening as a Routine Is Not Recommended in Men Aged ≥70 Years, or Any Man with a <10–15 Years Life Expectancy. (Recommendation; Evidence Strength Grade C)

Discontinuing screening at an age of 70 years may be too early for some healthy men with more than a 10 to 15-year life expectancy. Thus, one could argue for a flexible approach that accounts for age and health state. However, as a routine, the evidence suggests that screening in this age group will be associated with greater harm than benefits. First, over diagnosis rates increase with age and are especially high among older men with low-risk disease [24]. Second, a subset analysis from a large randomised screening trial [15] showed no reduction in prostate cancer deaths for men aged ≥70 years. Third, the competing risks of mortality in this age group make it less likely that a benefit from treatment of localised disease will be realised [25-27].

For those healthy men that decide on screening in this age group after shared decision-making, a higher threshold for biopsy referral [20] and discontinuing screening in those with lower PSA levels (i.e. <3.0 ng/mL) [28] may reduce the harms of over diagnosis and over treatment.

Conclusions

Undoubtedly, some will interpret the AUA guideline as an abandonment of PSA-based prostate cancer screening and conclude that if followed; the guideline would return us to the pre-PSA situation in which most men were diagnosed with advanced disease. I think that the evidence suggests otherwise; and that a more targeted screening approach will reduce harms and maintain benefits. Regardless of one's views, it is ironic that the message from investigators that have produced the only evidence for a screening benefit is one of caution until the evidence is clear [10, 15]; while in contrast the approach in the USA has been to encourage mass screenings at health fairs often sponsored by health systems that benefit directly from the diagnosis and treatment of prostate cancer. If nothing else, perhaps the AUA guideline will put an end to this practice until more is known about the benefits and harms of PSA-based screening [29].

Conflict of Interest

None declared.

Ancillary