European Urology

Loeb and colleagues present an interesting and timely review of prostate-specific antigen (PSA) testing in younger men, titled “Baseline Prostate-Specific Antigen Testing at a Young Age”[1]. In the article, younger is defined as ≤60 yr. Intuitively, it makes sense that PSA may be a better tumor marker in younger men. Serum levels are less likely to be confounded by coexistent benign prostatic hyperplasia, and as has been described in many autopsy series, many such men already harbor small tumors. These features underlie the presumption that younger men with the highest serum PSA values are at highest risk for a later diagnosis of important prostate cancer (PCa). Moreover, compared with older men, younger men have more risk of death from a diagnosis of PCa and the most to gain from treatment. The Scandinavian Prostate Cancer Group-4 study showed an overall benefit for treatment of PCa, but in stratified analysis this benefit was limited to men <65 yr [2].

Despite these considerations, PSA testing has generally occurred in older rather than younger patients. Two recent publications show the magnitude of this disconnect: In a large health system in the northwestern United States, <35% of patients aged 40–54 yr had testing, compared with nearly 50% of patients aged 55–74 yr and 25% of patients aged >75 yr [3]; Drazer et al showed that <45% of men in their 40s and 50s, compared with >45% of men in their 70s, were tested [4]. Moreover, few organizations have advocated PSA testing in younger men. While the American Urological Association and the National Comprehensive Cancer Network in the United States now recommend starting discussions about PSA testing in men at the age of 40 yr [5], the European Association of Urology guidelines and others from a variety of primary care organizations do not specify 40 yr as the age at which to start PCa early detection efforts in average-risk men [6]. Such caution stems from concerns about both overdiagnosis of cancers that might not be harmful to patients and the lack of randomized data on which to base recommendations. The three large randomized trials in the literature—the European Randomized Study of Screening for Prostate Cancer (ERSPC) trial [7]; the Prostate, Lung, Colorectal and Ovary (PLCO) trial [8]; and the Goteborg trial [9]—did not address PSA testing among younger men. In the ERSPC study, the core age range of men was 55–69 yr, with the mean and median age of participants just >60 yr. In the PLCO trial, the age range was 55–74 yr, with nearly 70% of patients >60 yr. In the Goteborg study, the age range of participants was lower than in the ERSPC and PLCO studies, 50–64 yr. The median age at first invitation to screening was 56 yr. None of the studies addressed PSA testing among men <50 yr. Thus, all recommendations about PSA testing in younger age groups rely on observational data.

As Loeb and colleagues explain, these observational data are of mixed quality for answering the question of when to offer men PSA testing. Prospective cohort studies in which PCa cases and mortality were ascertained independently of PSA testing provide substantial support for the concept of measuring PSA at younger ages. In these studies, serum from participants was stored, this serum was later checked for PSA levels, and the PCa status of the participants was ascertained based on clinically diagnosed PCa. This prospective cohort design eliminates the selection bias that occurs when clinical screening data are analyzed. In contrast, studies in which the serum PSA level has been used to justify the need for, and intensity of, further PSA screening introduce a significant selection bias. Men with higher PSA levels are more likely to continue PSA testing and ultimately are more likely to have a prostate biopsy than men who have lower PSA levels. These men may have earlier-stage cancers but may not have a mortality benefit from this extra PSA screening and cancer detection. Despite these limitations, observational data generally support the concept of PSA testing for younger men. Validation of the concept, however, will require either randomized trials or rigorous analysis and the application of sophisticated modeling techniques to determine the consequences of extending PSA testing to younger cohorts of men. Such techniques with comparisons among multiple models have been used successfully to estimate the lead time and overdiagnosis of PCa associated with PSA screening in the population aged 54–80 yr [10].

Beyond enhanced and earlier cancer detection, the overall clinical implications of PSA testing in younger men need to be considered. Potentially, such testing will result in the discovery of many “low-risk” cancers in younger men. Are there high-quality data on how to best treat younger patients diagnosed with PCa? For example, can active surveillance be applied to 40-yr-old men with low-risk disease? Would functional outcomes of radical prostatectomy or radiation therapy be different among men in their 40s compared with the more typical patient population of men in their 60s and 70s? In most surgical and radiation treatment series, men <50 yr were usually not diagnosed by screening, making this population very different from the patients in whom PCa would be detected if widespread PSA testing were brought into the population of men aged 40–50 yr. Finally, what are the implications of labeling a man with an above-average PSA as “high-risk” for PCa? While such patients would receive PSA testing of increased frequency, the impact of this testing on them is unclear.

Younger men may have much to gain from PSA testing, but the risks compared with benefits of this approach are unclear. Continued work on defining best PSA testing practices is needed. As research into this important topic continues, continued efforts to find markers independent of PSA will be vitally important.

Conflicts of interest

Gerald L. Andriole received support from Amarex LLC, Amgen, Augmenix, Bayer, Bristol Myers Squibb Company, Cambridge Endo, Caris, Envisioneering Medical, GlaxoSmithKline, Janssen Biotech, Inc., Myriad Genetics, Steba Biotech, Ortho-Clinical Diagnostics, and Viking Medical. Seth A. Strope has nothing to disclose.