In 2014, an estimated 233,000 men in the United States will be diagnosed with prostate cancer, the majority of whom will be diagnosed by opportunistic PSA screening [1]. Since the PSA test was made commercially available in the late 1980’s, the incidence rate of prostate cancer has increased by roughly 33 % (from 120 to 160 per 100,000 population) and a man’s lifetime risk has more than doubled (from 7.3 % in 1975 to 18 % in 2012) [1]. PSA screening preferentially detects cancers of low-grade and low-stage that may pose a limited threat to a man in terms of longevity and quality-of-life. As a result, the majority of men diagnosed with prostate cancer have low-risk tumors, defined as PSA <10 ng/mL, biopsy Gleason score 2–6, and clinical T2a or less [2].

Estimates from the European Randomised Screening Trial for Prostate Cancer and the U.S. Surveillance, Epidemiology, and End Results (SEER) registry indicate that 25–50 % of cancers detected through screening represent over diagnosis, defined as tumors that would have gone undiagnosed over a man’s lifetime in the absence of screening [3, 4]. In older men, prostate cancer is ubiquitous, as approximately 60–70 % of men dying from non-prostate cancer causes will have histological evidence of prostate cancer in autopsy series [5]. While a 35 % reduction in prostate cancer-specific mortality in the U.S. has been observed since the early 1990’s (of which 45–70 % may be attributable to screening) [6], it is estimated that the vast majority of excess men diagnosed as a consequence of opportunistic PSA testing were over-treated and did not benefit from early detection [7]. The issue of over-treatment is particularly relevant for prostate cancer as all treatments (surgery and radiation therapy) may negatively impact urinary, bowel, and sexual function (even when performed at high-volume hospitals), each is associated with a small but definable risk of treatment-related mortality, and many men will live decades with the sequelae of the treatments they have received [8]. The issues of over diagnosis and over-treatment were largely behind the recent U.S. Preventive Services Task Force’s grade D recommendation for PSA screening, indicating that there is a moderate-to-high certainty that the screening has no net benefit or that the harms outweigh the benefits [9].

While a strong rationale can be made that the majority of men diagnosed with screen-detected prostate cancer do not need immediate radical therapy (particularly for those with life expectancy <10–15 years due to advanced age and/or comorbid medical illness), the vast majority of American males receive some form of definitive local therapy. Data from the combined, linked SEER-Medicare registry of men over 65 years of age suggest that 75–84 % of elderly men diagnosed with localized prostate cancer receive surgery or radiation therapy [7, 10]. Expectant management rates for men with low-risk prostate cancer in the CaPSURE registry are only 6–8 % and these rates appearing to be decreasing over time [11, 12]. Similarly low rates of surveillance among low-risk patients has been reported from high-volume, academic centers [13, 14].

It is likely that both patient- and physician-related factors have contributed to the low utilization of expectant management for prostate cancer in the United States, such as anxiety and loss of control for the patient, and financial disincentives and medicolegal concerns for the physician. One of the major factors that has contributed to the low acceptance of this approach is concern by the patient and physician that the threat a cancer poses to an individual may be underestimated by the PSA, clinical stage, biopsy Gleason score, and other information from the diagnostic biopsy (e.g. number of positive and negative cores, percentage of cancer within the biopsy specimens). Data from conservatively-treated patients from the pre-PSA era have suggested rates of prostate cancer-specific mortality at 15 years of 20–25 %, 65–75 %, and 80–90 % for men with biopsy Gleason score 6, 7, and 8–10 cancers, respectively [15]. As such, expectant management has largely been restricted to those with a high probability of indolent cancer (defined as tumor volume <0.5 cc, absence of Gleason pattern 4, and no evidence of pathologically advanced cancer features such as extraprostatic extension or seminal vesicle invasion) as these features are associated with cancers detected at autopsy and, with exceedingly rare exceptions, all these men are cured by radical therapy [5, 16]. Various clinical criteria and nomograms have been developed to predict which men have a high probability of having indolent prostate cancer for whom expectant management is a safe approach [17–20]. However, fewer than 15 % of patients from screening and surgical series meet pre-treatment criteria for indolent prostate cancer [21–23], and thus likely under-estimate the proportion of men who may be eligible for active surveillance.

For various reasons, the outcomes of conservatively managed patients from the pre-PSA era may not be applicable to contemporary patients to determine who should and should not be considered for expectant management. First, PSA screening is associated with a lead-time of 5–12 years in the diagnosis of prostate cancer [24]. Second, patients treated in the pre-PSA era may not have been appropriately staged at diagnosis as bone scintigraphy and computed tomography (CT) imaging were in their infancy and many of these men may have had metastatic disease at diagnosis. Third, many of these men were diagnosed by TURP (which largely samples the transition zone) and thus index tumors that largely develop in the peripheral zone may not have been adequately sampled. Lastly, substantial changes to the Gleason scoring system have taken place which culminated in a modification of the Gleason scoring system such that many tumors of lower grade in years past would now be classified as high-grade cancers [25, 26].

In summary, the rationale for expectant management of localized prostate cancer is based on the fact that the majority of men with prostate cancer will die with their disease rather than from their disease; estimates from various sources suggest the number of men needed to be treated to prevent one prostate cancer death ranges from 18 to 48 [7, 27, 28]. As treatment may be associated with important impacts on quality-of-life, an observational strategy has the potential to reserve treatment for those who it need and avoid unnecessary treatment for those that do not, thus preserving quality-of-life without compromising survival.

Expectant management of prostate cancer may take one of two forms: watchful waiting and active surveillance. Watchful waiting involves less intensive monitoring with periodic clinical assessment with or without PSA testing and imaging where treatment (usually in the form of androgen deprivation therapy) is typically administered with palliative intent for symptomatic local or systemic progression. Active surveillance involves intensive monitoring with PSA testing and surveillance prostate biopsies at periodic intervals and intervention with curative intent is recommended when evidence of more important cancer is detected (either cancer of higher grade, increasing cancer burden on biopsy or prostate exam, or changes in PSA). It is anticipated that the risks of clinical progression and prostate cancer-specific mortality will be less with an active surveillance strategy compared to watchful waiting (though the probability of receiving radical treatment will be higher). As such, active surveillance has emerged as the preferred observational strategy by clinicians and patients for those with life expectancy >10 years who would otherwise be candidates for definitive local therapy.