Screening for Prostate Cancer

Michael J. Barry

Prostate cancer is a common cause of morbidity and mortality among older men in the United States. The lifetime probability of a man learning that he has prostate cancer in the “prostate-specific antigen (PSA) era” is now about 18%, about double what it was before the PSA era, but the probability of death from prostate cancer is approximately 3%. This ratio of cumulative incidence to mortality indicates that most patients in whom prostate cancer is diagnosed die of something else. In fact, the lifetime risk of being diagnosed with prostate cancer depends on the intensity of screening. In the recent Prostate Cancer Prevention Trial (PCPT), after 7 years of annual digital rectal exams (DREs) and PSA tests followed by a set of biopsies, regardless of DRE and PSA results, about 25% of average-risk older men were diagnosed with prostate cancer.

Patients and clinicians face a great deal of uncertainty in making clinical decisions about the early detection and treatment of prostate cancer. These tumors are exceedingly common and have the potential to cause significant morbidity and mortality. They also have a variable, often indolent course and a higher prevalence in elderly men, whose health is often more limited by their age and other diseases. Most of the prostate cancers found through PSA screening are not destined to cause future morbidity or mortality, yet it is difficult to know for certain which prostate cancers have been “overdiagnosed” at the individual level. The clinician should be aware of the unpredictable natural history of the disease and the limitations of knowledge about the benefits of therapy when considering the use of screening tests for prostate cancer. At long last, clinical trials of PSA screening have shed some light on the trade-off between harms and benefits of PSA screening, but they have not resolved the uncertainty.

EPIDEMIOLOGY AND RISK FACTORS (1, 2, 3 and 4)

The incidence of prostate cancer increases with age. Age-specific prevalence rates derived from autopsy studies are about 15% for men in their sixth decade, 25% for those in their seventh decade, and 40% or higher for men older than 70 years. The incidence of prostate cancer rose dramatically in the United States with increasing use of the PSA test and has now fallen to a new, higher plateau. Mortality from prostate cancer also increased with the introduction of PSA testing and is also now falling (Fig. 126-1). African Americans have a two- to threefold higher incidence of prostate cancer. Having a brother or a father with prostate cancer increases the risk twofold. A diet high in fat (particularly red meat) may also be a risk factor. Obesity may be a risk factor, particularly for more aggressive forms of prostate cancer. Evidence linking prostate cancer with prior vasectomy has been equivocal. No strong evidence has linked prostatitis or benign prostatic hyperplasia (BPH) to the development of prostate cancer. Despite some suggestive epidemiologic data, clinical trials of selenium and vitamin E supplements have not been shown to reduce the risk of prostate cancer. Chemoprevention using the 5-α reductase inhibitors finasteride and dutasteride can decrease the overall incidence of prostate cancers, but may actually increase the risk of high-grade cancers.

NATURAL HISTORY OF DISEASE AND EFFECTIVENESS OF THERAPY (5, 6, 7, 8, 9 and 10)

Clinical Presentation and Course

Early prostate cancers are asymptomatic; older men not infrequently have concurrent lower urinary tract symptoms caused by BPH, but there is no evidence that such symptoms increase the likelihood that a man is harboring prostate cancer. Early prostate cancers may be discovered through screening or diagnosed incidentally when men undergo prostatectomy for BPH. Advanced cancers may present with obstructive symptoms, bleeding, or bone pain due to metastases.

The prognosis of prostate cancer depends as much on the degree of histologic differentiation of the tumor as on the extent of the disease, at least before detectable metastases are
evident. Histologically, prostate cancers are often heterogeneous. Pathologists commonly assign a Gleason score of 1 to 5 to the mostcommon and next-most-common histologic patterns and add the two scores to obtain a Gleason sum ranging from 2 to 10. Cancers with Gleason sums of 2 to 4 are considered well differentiated, 5 to 7 moderately differentiated, and 8 to 10 poorly differentiated (although cancers with sums of 7 in fact behave in an intermediate fashion between moderate and poorly differentiated cancers). Most cancers discovered by screening are Gleason 6 to 7. In recent years, pathologic grading (but not the cancers themselves) has shifted so that few prostate cancers are now graded less than Gleason 6.

Figure 126-1 Incidence and mortality of prostate cancer in the United States. (From Howlader N, Noone AM, Krapcho M, et al., eds. SEER Cancer Statistics Review, 1975-2008. Bethesda, MD: National Cancer Institute. http://seer.cancer.gov/csr/1975_2008/, based on November 2010 SEER data submission, posted to the SEER web site, 2011.)

Clinically localized prostate cancers generally have doubling times measured in years. The prognosis of prostate cancer depends on tumor factors such as grade and patient factors such as age and comorbidity. Table 126-1 provides estimates of the 15-year probability of being alive, dying of other causes, or dying of prostate cancer for different combinations of these factors. These data were generated from men in whom cancer was diagnosed in the “pre-PSA era,” with adjustment for lead time, “grade creep,” and increasing life expectancy in the “PSA era.”

TABLE 126-1 Estimates of 15-Year Outcome Probabilities for Men Diagnosed with Prostate Cancer in the PSA Era, by Age and Gleason Score

Gleason Score <7

Gleason Score 7

Gleason Score >7

Age (y)

15-Year Outcome

Conservative Treatment

Curative Treatment

Conservative Treatment

Curative Treatment

Conservative Treatment

Curative Treatment

55-59

% Alive

% Other death

% PC death

60-64

% Alive

% Other death

% PC death

65-69

% Alive

% Other death

% PC death

70-74

% Alive

% Other death

% PC death

PC death, death attributed to prostate cancer.

Adapted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: British Journal of Cancer, from Parker C, Muston D, Melia J, et al.

Effectiveness of Therapy

The variability in the natural history of this disease complicates an assessment of therapeutic efficacy. Some have argued that early detection and aggressive treatment with radical prostatectomy or radiation therapy, either external beam radiotherapy or implantation of radioactive seeds (interstitial radiotherapy or brachytherapy), may improve survival, particularly for patients with at least a 10-year life expectancy in whom the tumor appears to be localized to the gland. However, practices have evolved in the absence of evidence for a mortality benefit from randomized clinical trials, and these treatments have side effects, including a relatively high risk of sexual dysfunction and urinary incontinence. A randomized trial of radical prostatectomy compared with watchful waiting was conducted among Scandinavian men younger than age 75 years with well- and moderately differentiated prostate cancer. Fifteen-year results showed surgery resulted in a significant reduction in prostate cancer-specific mortality, from about 21% with watchful waiting to 15%, with a similar absolute
decrease in overall mortality. The benefit appeared confined to men younger than 65 years old. However, it is important to recognize that men in the Scandinavian trial were generally diagnosed with a nodule or disease that was otherwise clinically evident, with almost half of participants undergoing surgery showing evidence of extracapsular cancer. The recently published Prostate cancer Intervention Versus Observation Trial (PIVOT) randomized 731 men with prostate cancer largely discovered through PSA screening to radical prostatectomy versus observation. After a median follow-up of 10 years, overall and prostate cancer specific mortality were not significantly lower with surgery. However, for the minority of men with more aggressive cancers, such as with a baseline PSA level >10 ng/mL, surgery significantly reduced overall mortality. An ongoing trial in the United Kingdom has a treatment trial (surgery vs. radiotherapy vs. active monitoring) nested within a screening trial.

If aggressive treatment of clinically localized prostate cancer had fewer side effects, there would be less controversy about PSA screening. However, both surgery and radiation therapy can have lasting side effects, particularly in terms of erectile function and continence. “Active surveillance” rather than immediate treatment for lower risk prostate cancer may be an important strategy to help prevent the conversion of the overdiagnosis associated with PSA screening to overtreatment.