Where the New AUA Guidelines Went Wrong

Selected by Pietro Cazzola

AUAThe AUA guidelines are a step in the wrong direction for patient-centered care and may deprive many men of the opportunity to pursue shared decision making about possibly life-saving prostate-specific antigen (PSA) testing. A more forward-looking approach is needed.
In my opinion, the AUA guidelines are based on incomplete data and inaccurate estimates of the benefits and harms of PSA testing. The AUA asked its Guidelines Panel to develop guidelines based on evidence from randomized clinical trials (RCTs) and statistical modeling studies. But, there is not much reliable evidence from the available RCTs, and some are profoundly flawed.  In fact, more than a dozen professional organizations have reviewed the same body of evidence and have come up with vastly divergent guidelines, ranging from the US Preventive Services Task Force recommending against PSA testing for any man to the European Association of Urology recommending a baseline PSA test beginning at age 40–45, with follow-up testing for all men with a > 10-year life expectancy in a setting of shared decision making.
The RCTs were conducted over a limited time period and do not reveal true information about absolute benefits of screening over a lifetime. Using trials’ data to estimate benefits and harms of PSA testing underestimates benefits and exaggerates harms.
In weighing the benefits, the Panel focused solely on prostate cancer death without also considering avoiding suffering from metastatic disease that might not have resulted in a cancer death.  An analogy would be a study of the benefits of wearing seatbelts.  Is the benefit only the deaths prevented or should they also include the catastrophic injuries prevented that did not result in death? However, avoiding metastatic disease shifts the balance of harms and benefits; men diagnosed with more advanced disease ultimately require more treatments and have more side effects. In weighing the harms of PSA testing, the Panel cast a net over a variety of side effects of PSA testing, biopsy, and prostate cancer treatment that range from minor to serious. The harms of a simple blood test should not be linked with the possible harms of biopsy and treatment, and few of these side effects reach the extreme of death from metastatic prostate cancer.
The AUA does not recommend screening men under age 55 who have an average risk of prostate cancer. However, the primary objective of testing men in their 40s is to establish a baseline PSA that helps assess a man’s risk for the subsequent development of life-threatening disease.  Men in their 40s in the top 10% of PSA levels for their age group account for almost half of all prostate cancer deaths up to 30 years later, and those with levels above 1 ng/mL clearly warrant more careful monitoring. A higher baseline PSA in a man in his 40s is a strong risk factor, perhaps stronger than African heritage or a positive family history. Accordingly, it is not possible to make a full assessment of which patients are actually at high risk without having a patient’s baseline PSA from his 40s or early 50s.
The reason the AUA did not recommend screening in average-risk men prior to age 55 is because PSA screening has not been adequately tested in an RCT, and there is little evidence from RCTs concerning benefits or harms of PSA testing in men younger than 55. The meager evidence available suggests that it is beneficial.  But, starting at age 55 is too late. There is no reason to believe that if PSA testing works in men 55 to 69 years old, it would not also work in men ages 45 to 55. Although the fine print in the AUA guidance document explains that the Panel does not recommend against PSA testing for men 40 to 55 years old, the actual guidelines statement uses the language, “we do not recommend.” Rather, it should read, “there is insufficient evidence to recommend for or against early detection in men younger than 55.” Because of this, the AUA Guidelines frequently have been misrepresented, with headlines such as “Urology group stops recommending routine PSA test,” which appeared in USA Today.
The AUA Panel’s suggestion for longer PSA testing intervals needs to be reconciled with the realization that less frequent testing decreases the ability to detect the aggressive cancers that have the shortest preclinical phases.  However, with less frequent testing, there remains the undesirable effect of detecting all of the low-risk cancers (length-time bias), possibly doing more harm than good.
The AUA does not recommend routine PSA testing in men older than 69, despite the fact that 50% of prostate cancer deaths occur in men diagnosed after age 75. Age 70 is too young to stop testing in otherwise healthy men who have a 10–15 years or more life expectancy. Therefore, PSA testing in men over 70 should be performed on an individual basis with shared decision making between physicians and patients.  In the absence of shared decision making, men are more than twice as likely not to undergo PSA testing.  
There has been a 75% reduction in metastatic prostate cancer at diagnosis and more than a 45% decrease in the age-adjusted prostate cancer mortality rate in the United States during the PSA era, largely attributable to PSA testing. Restricting PSA testing too much would significantly compromise these benefits.

William J Catalona, MD