Urology
Volume 73, Issue 5, Supplement , Pages S11-S20, May 2009

Critical Appraisal of Prostate-specific Antigen in Prostate Cancer Screening: 20 Years Later

  • Kenneth J. Pienta

      Affiliations

    • Department of Internal Medicine, University of Michigan Comprehensive Cancer Center, Ann Arbor, Michigan, USA
    • Department of Urology, University of Michigan Comprehensive Cancer Center, Ann Arbor, Michigan
    • Corresponding Author InformationReprint requests: Kenneth J. Pienta, M.D., Department of Internal Medicine, University of Michigan Comprehensive Cancer Center, 7308 CCC, 1500 East Medical Center Drive, Ann Arbor, MI 48109

Received 2 February 2009; accepted 20 February 2009.

Article Outline

Prostate-specific antigen (PSA) is secreted by all types of prostate epithelial cells and has been used for 2 decades as a biologic marker for prostate cancer (PCa). Since the implementation of PSA screening in the United States, the detection of PCa has increased, accompanied by a decrease in the incidence of high-grade cancer and PCa-specific mortality rates. It has been suggested that these decreases have resulted from the enhanced detection of PCa while still curable. These data have been the impetus for early detection programs, which have recommended the initiation of screening as early as 40 years of age. Despite widespread use, PSA screening remains controversial, principally because of the lack of evidence from randomized controlled trials demonstrating a mortality benefit that could outweigh the concerns of the costs of overdiagnosis and overtreatment. Two ongoing, randomized controlled trials are examining whether screening reduces the risk of PCa-related mortality, and the results of these studies are expected soon. Although it has its limitations, PSA still remains the best-studied marker for the detection of PCa.

 

Prostate-specific antigen (PSA) was approved by the U.S. Food and Drug Administration in 1986 for monitoring disease status after definitive treatment in men with prostate cancer (PCa).1 The clinical utility of PSA determination has subsequently expanded to help identify men with PCa and, most recently, data have been published indicating that PSA can identify men at risk of developing PCa in the future2, 3, 4 (see also the report by Fleshner and Lawrentschuk in this supplement). PSA screening remains controversial despite its widespread use in the United States, and its role has continued to evolve. The aim of the present report was to describe the characterization of PSA, provide a brief historical overview of its emergence as an important biomarker for PCa, and to review the impact of PSA screening on current clinical practice.

Back to Article Outline

History of PSA and its use in PCa 

PSA Characterization 

First identified in seminal fluid in 1966, PSA was originally called γ-seminoprotein and was used for the identification of semen in cases of sexual assault.5 Its relevance as a tumor marker was established in 1979, when Wang et al.6 detected a prostate antigen that came specifically from normal, benign hypertrophic and malignant prostatic tissue samples, but not from other human tissues (hence, prostate-“specific” antigen).6 Importantly, at discovery, this novel prostate antigen was distinguished from prostatic acid phosphatase (PAP), which was discovered in 1938 and used for PCa detection,7 and thus represented a novel PCa marker. Since then, other PSA isoforms have been identified and extensively studied.

PSA is prostate specific, not cancer specific. An increased serum PSA level can be associated with nonmalignant conditions such as benign prostatic hyperplasia (BPH), infection, or chronic inflammation.8, 9 Cancer cells produce lower levels of PSA than do BPH cells but release a greater amount of PSA into the blood, likely reflecting the disordered cellular and glandular architecture of PCa.8

The measurement of specific forms of PSA (eg, free, complex) can help differentiate between PCa and benign prostatic conditions.1 Complexed PSA is the predominant form in serum for patients with BPH or PCa, but the median free-to-total PSA ratio has been shown to be lower (0.18) in men with untreated PCa than in those with BPH (0.28; P < .0001).10 Free PSA is composed of 3 isoforms: BPH-PSA, inactive PSA, and proenzyme-PSA (pro-PSA).1 BPH-PSA and inactive PSA are more abundant than pro-PSA in benign prostatic tissue, whereas in prostatic tissue from men with PCa, the proportion of pro-PSA isoforms is 1.3-1.4 times greater than that of BPH-PSA and inactive PSA.1, 11, 12, 13 In addition, pro-PSA might preferentially identify more aggressive forms of PCa. A large, retrospective analysis provided preliminary evidence that greater pro-PSA/free PSA ratios are associated with more aggressive PCa (Gleason score ≥7 and/or extracapsular tumor extension).14 The assessment of different measures or forms of PSA might more accurately establish why total PSA is generally elevated in men with PCa.

Discovery and Use of PSA Levels in PCa Diagnosis 

The utility of PSA as a marker of PCa has evolved significantly over time. The evolution is described in this section in temporal organization. Since its isolation from prostatic tissue in 1979,6 the use of serum PSA levels in PCa diagnosis and detection has evolved dramatically. Before the measurement of PSA, healthcare practitioners relied on digital rectal examination (DRE), PAP measurement, and transrectal ultrasonography (TRUS) to screen and diagnose PCa.15, 16, 17, 18, 19 DRE had only an estimated 1%-2% cancer detection rate in self-referred screening populations when used as the primary detection method.19 Furthermore, many men are uncomfortable with DRE because they are embarrassed or fearful that the examination will be painful.20 Another drawback of DRE is that in 48%-85% of men whose disease is detected using this method, PCa is already extraprostatic and nonorgan-confined at diagnosis.19 DRE is, in particular, a statistically significant predictor of high-grade disease.21, 22 DRE is discussed in more depth as the trials are reviewed.

Figure 1 highlights the important milestones in the development of clinical utility for PSA. In 1987, Stamey et al.23 evaluated the potential utility of serum PSA and PAP (still used then for PCa diagnosis) in PCa detection and determined that PSA was a more sensitive marker of PCa than PAP. Additionally, it was found that the PSA level increased directly with clinical stage and decreased to undetectable levels after prostatectomy, indicating that it could be useful for detecting residual disease after treatment.23 The use of PAP in PCa diagnosis was rapidly abandoned in favor of PSA.

Subsequently, 2 landmark studies demonstrated the value of PSA for detecting PCa in clinical practice.2, 24 In the first, Cooner et al.2 evaluated 1807 men aged 50-89 years from a referral population of urologic patients, using TRUS, DRE, and PSA determination. The overall PCa detection rate was 15% (compared with the 1%-2% historically seen with DRE alone). Of the 263 patients with PCa, 72% had both suspicious DRE findings and elevated PSA levels. In contrast, 13% of the patients had abnormal DRE findings only, and 16% had elevated PSA levels only. The investigators thus recommended combined use of PSA measurement and DRE, suggesting that the addition of PSA testing to DRE could provide important information regarding the risk of cancer.2 This was 1 of the initial landmark trials that, along with many other trials, used a PSA level of ≥4.0 ng/mL as a threshold for initiating biopsies in PCa diagnosis, a recommendation reflecting, in part, the PSA assay used in the trial (Tandem-R assay, Hybritech, San Diego, CA),2 which reported the “normal” range as 0-3.99 ng/mL.19 Values of ≥4 ng/mL were considered the upper level of normal and indicative of possible PCa, thus warranting confirmation of disease through biopsy. Values of ≥4 ng/mL were widely adopted in clinical practice on the basis of the use of this cutoff in clinical trials, despite the lack of evidence about whether this value represented the optimal balance between sensitivity and specificity. Gann et al.3 reported the first evaluation of the risk of PCa using a baseline PSA level, emphasizing the limitations of focusing on an absolute PSA cutoff for PCa diagnosis. In a case-control study of men enrolled in the Physicians' Health Study, they showed that a PSA cutoff of 4.0 ng/mL had high sensitivity and specificity for the detection of life-threatening PCa, but they also showed that PSA levels well below the traditional cutpoint of 4.0 ng/mL were associated with a significant risk of PCa that increased incrementally with increases in the PSA level.3

The second landmark study, of 1653 healthy volunteers aged ≥50 years with no history of PCa or prostatitis, compared PSA, DRE, and TRUS for the detection of PCa.24 Men with two serum PSA values of ≥4.0 ng/mL underwent DRE and TRUS, with biopsies performed if either the DRE or TRUS findings were abnormal. Of the men undergoing biopsy, PCa was detected in 22% of men with a PSA level of 4.0-9.9 ng/mL and in 67% of those with a PSA level of ≥10 ng/mL. The investigators showed that among the 37 men with cancer, DRE alone would have missed 32% and TRUS alone would have missed 43%. Serum PSA measurement detected PCa more accurately than either DRE or TRUS. Of the 2-test combinations, PSA plus DRE offered the lowest error rate.24 Together, these 2 studies demonstrated that PSA and DRE could be used in a complementary fashion to determine the need for TRUS and biopsy and that PSA was more accurate in the detection of PCa than DRE or TRUS alone. Although PSA measurement is the best single diagnostic tool, Figure 2 shows that a family history of PCa and the DRE findings led to substantial differences in the predictive value beyond the use of the PSA level alone for the individual patient. The average out-of-sample area under the curve (AUC) for the prediction given by this risk equation combining PSA, DRE, and family history was 70.2% (SD 0.57%) in a study by Thompson et al.21 (see later in this section). In another recent study on predicting PCa, the receiver operating characteristic analyses showed an AUC for a risk calculator based on that used in the Prostate Cancer Prevention Trial (PCPT) (ie, reflecting age, race, PSA, DRE, family history, and negative biopsy history) of 66.7%, significantly greater statistically than the AUC for PSA alone of 61.9% (P < .001; Fig. 3).25

  • View full-size image.
  • Figure 2. 

    Risk of prostate cancer (PCa) as function of prostate-specific antigen (PSA) level, digital rectal examination (DRE) result, and family history of PCa for men who had not previously undergone prostate biopsy. Vertical lines indicate pointwise 95% confidence intervals for risk at each PSA level. DRE+, abnormal DRE findings suggestive of PCa; DRE–, normal DRE findings; FAM HIST+, family history of PCa; FAM HIST–, no family history of PCa. Reprinted, with permission, from Thompson et al.21

  • View full-size image.
  • Figure 3. 

    Receiver operating characteristic curves for prostate-specific antigen (PSA), Prostate Cancer Prevention Trial (PCPT) risk calculator, and novel logistic-regression based model for predicting prostate cancer (PCa). Reprinted, with permission, from Hernandez et al.25

With the demonstration in the late 1980s and early 1990s of the utility of PSA in PCa detection,26, 27, 28 large-scale studies, including 1 led by investigators from the American Cancer Society National Prostate Cancer Detection Project, initiated PSA screening, along with DRE and TRUS, for the detection of PCa.16, 29 These studies examined both men who volunteered for screening and populations seen within a general urologic practice.

Since 1986, the Food and Drug Administration has approved several PSA tests for monitoring the occurrence and possible recurrence of PCa.30 PSA screening has been widely adopted in the United States, despite the lack of level 1 evidence demonstrating its effectiveness in reducing PCa mortality.31, 32 Although no large-scale randomized studies have proved that PSA testing affects PCa mortality, an observational study has suggested that PSA screening reduces mortality.33 Data that point to the benefits of PSA in screening are reviewed in the following paragraphs.

Subsequent to the incorporation into large-scale studies of PSA measures to detect PCa, investigators began researching additional possible PSA-related measures. Two of these studies, published in 1992 and 1993, suggested that kinetic PSA measures (eg, PSA velocity and PSA doubling time) might improve the specificity of diagnostic testing and help identify patients with more aggressive tumors.34, 35, 36 PSA velocity has not been shown to increase the diagnostic or predictive accuracy of a simple PSA level test; the utility of kinetic PSA measures in clinical practice is unclear, and research continues on their implications.37

A discussion of adjusted PSA levels for age and race is beyond the scope of this report and has been discussed in previously published studies.38, 39 Oesterling et al.38 studied the correlation of age and PSA level and, using the PSA distribution, they proposed an age-specific PSA threshold for recommending prostate biopsy of 2.5, 3.5, 4.5, and 6.5 ng/mL for men aged 40-49, 50-59, 60-69, and 70-79 years, respectively. Age-specific and race-specific PSA reference ranges have not gained wide acceptance in primary care practice. The data on different thresholds in black men have been inconsistent,40 and the PCPT did not show that age altered PCa risk within the age range of that study.21

By the late 1990s, several of the key findings from the initial PSA screening studies had begun to emerge.29 One key study examined a lower PSA threshold of 2.5 ng/mL for PCa screening.41, 42 The Washington University PSA-3 PCa screening study, conducted from May 1995 to October 1996, screened men aged ≥50 years with PSA measurement and DRE at 6-month intervals. Using a PSA range with a lower cutoff point for biopsy (>2.5 ng/mL but <4 ng/mL) resulted in a large majority of detected tumors with favorable histologic features, with about 80% of tumors organ confined (compared with about 70% when a PSA cutoff of 4.0 ng/mL was used) and did not substantially increase the detection of clinically unimportant cancer.41

The PCPT investigated the prevalence of PCa among men who had a PSA level of ≤4.0 ng/mL. Of the 18 882 men enrolled in the total PCPT (a PSA level of ≤3 was an entrance criterion), 9459 were randomly assigned to receive placebo. Thompson et al.21 examined the end-of-study biopsy results in relation to PSA level (taken within 90 days of biopsy) in 2950 of these men, aged 62-91 years, who had never had a PSA level >4.0 ng/mL or abnormal DRE findings and had not previously undergone biopsy or transurethral resection during the 7-year study itself. Overall, 15.2% of the participants included in the analysis had PCa on biopsy. The prevalence of PCa for men with a PSA level of ≤0.5 ng/mL was 6.6%, and the prevalence of PCa increased in association with increasing PSA levels. For example, a PSA level of 1.1-2.0 ng/mL resulted in a prevalence of PCa of 17%, and the prevalence reached 26.9% in men with a PSA value of 3.1-4.0 ng/mL. Figure 4 shows the prevalence of PCa associated with increasing PSA cutoff values. These findings indicate that PCa can be diagnosed at any PSA level, however low.43 In another separate analysis of data from the PCPT, the variables that predicted for PCa included higher PSA level, a positive family history of PCa, and abnormal DRE results. Neither age at biopsy nor PSA velocity contributed independent prognostic information in the PCPT.21

Another key finding from the PCPT with respect to the diagnostic ability of PSA was the effect of 5-α-reductase inhibitors (5-ARIs) on the PSA level. Previous studies had shown that administration of the 5-ARIs dutasteride and finasteride decreased serum PSA levels by a mean/median of approximately 50% by 12 months of treatment in men with BPH.44, 45 The specificities and sensitivities for detecting PCa and 2 categories of higher grade disease (Gleason score ≥7 and Gleason score ≥8) were compared. This analysis of end-of-study biopsies from the placebo and finasteride groups showed that the matched PSA cutoffs in the finasteride group were more sensitive and specific for PCa detection. A PSA value of 1.6 ng/mL in the finasteride group achieved a specificity >90% for the detection of PCa vs no PCa. Similar values were obtained for the detection of high-grade (Gleason score ≥7) PCa. Finasteride statistically significantly increased the AUC of PSA for detecting PCa and high-grade disease; the AUC was 0.757 in the finasteride group and 0.681 in the placebo group (P < .001; and P = .003 and P = .071 for Gleason score 7-10 and Gleason score 8-10 disease, respectively). On the basis of this evidence, it was proposed that because 5-ARIs reduce BPH symptoms and treatment causes a decrease in PSA levels, 5-ARI treatment could be used to enhance the detection of PCa.46 5-ARIs might preferentially suppress PSA secretion from benign tissue, making the PSA from cancerous tissue more “visible.” This PCPT analysis showed that finasteride can enhance the performance of PSA for detecting overall and high-grade PCa in the general population. Figure 5 shows the receiver operating characteristic curves of the finasteride- and placebo-treated patients for PSA detection of PCa and high-grade PCa.46 This finding represents an important pharmacologic enhancement of the diagnostic ability of PSA to detect PCa and the risk of PCa. The effect of the dual 5-ARI dutasteride on enhancing PCa detection in an analogous manner is currently being studied in the Reduction by Dutasteride of Prostate Cancer Events (REDUCE) trial, which is evaluating the effect of dutasteride on PCa incidence in a high-risk population (PSA level ≥2.5 and prostate volume ≤80 cm3).47

  • View full-size image.
  • Figure 5. 

    Receiver operating characteristic (ROC) curves for prostate-specific antigen (PSA) detection of all prostate cancer (PCa) and high-grade PCa. (Left) ROC curves for all prostate cancer; (Middle) ROC curves for Gleason score ≥ 7 PCa; (Right) ROC curves for Gleason score ≥8 PCa. Solid line indicates placebo group; dashed line indicates finasteride group. For difference between placebo and finasteride groups, P < .001 for all PCa, P = .003 for Gleason score ≥7 PCa, and P = .071 for Gleason score ≥8 PCa. Reprinted, with permission, from Thompson et al.46

The integration of PSA testing into urologic practice has provided an opportunity to detect PCa more often at a localized and, thus, curable, stage.19, 48, 49 It has also improved the evaluation of the disease extent after diagnosis. Additionally, PSA has provided a method for monitoring the effectiveness of PCa treatment. In light of these benefits, current European Association of Urology guidelines and the 2007 American Urological Association (AUA) guidelines recommend measurement of PSA levels for the diagnosis and staging of PCa.32, 50 The 2007 AUA guidelines acknowledge the importance of PSA measurement as a screening tool and recommend the incorporation of PSA into risk stratification for patients with localized disease.32 The National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (Prostate Cancer Early Detection) advocate the use of PSA as a screening tool and detail how to use it in middle-age men to define long-term risk and, therefore, define a repeat screening policy.51

Evidence has shown that PSA screening might reduce mortality, even though we have no randomized trials. In the Tyrol Prostate Cancer Demonstration Project, 86.6% of eligible men aged 45-75 years were tested for PSA at least once after 1993 in Tyrol, Austria. Cancer deaths in Tyrol in 2005 had decreased by 54% (95% confidence interval 34%-69%) compared with a 29% reduction reported for the rest of Austria (95% confidence interval 22%-35%; P = .001). Widespread PSA testing and treatment with curative intent resulted in a reduction in the PCa mortality rates. This reduction in PCa mortality was most probably due to early detection, consequent downstaging, and effective treatment of PCa.33 Numerous studies have confirmed that a greater PSA level is associated with a larger tumor volume and greater clinical stage, pathologic stage, and Gleason score. PSA is now used in several multivariate models that predict PCa stage and grade.37

Limitations of PSA Testing 

Despite its proven diagnostic ability, PSA testing has limitations as a biologic marker. One important limitation is that a dichotomous assessment is not very useful for distinguishing between nonaggressive and aggressive PCa.52 The PCPT demonstrated that PSA was better at detecting high-grade tumors than low-grade tumors in the placebo arm.46 However, recent data have suggested that additional measurement of specific PSA isoforms and the determination of PSA kinetics might help identify patients with more aggressive disease.14, 52 These data are controversial, however, and have not been proved as uniformly as PSA testing. Researchers are actively investigating additional PCa markers to add specificity to PSA testing. A relatively high rate of false-positive results associated with testing is another limitation of PSA screening. Ranges vary, but approximately 48% of men with solely an elevated PSA level are found to have PCa.53 PSA levels are highly variable and can be influenced by a number of factors, such as ejaculation and prostatic manipulation (eg, catheterization, cystoscopy, prostatic massage), although some uncertainty exists regarding the magnitude of the effect of ejaculation on PSA levels and the mechanisms underlying this effect.1, 54, 55 Patient physiologic factors (eg, weight and carbohydrate intake, insulin resistance, metabolic syndrome) also appear to increase or decrease PSA levels.56

Back to Article Outline

PCa screening: Too much or too little? 

Rationale for Screening 

As with other cancers, the rationale for screening for PCa is the detection of cancer at an early stage, when it is more likely to be curable; PSA testing is also used to predict the likelihood of PCa development.57 Recent epidemiologic studies have suggested that PSA facilitates the early detection of PCa but have not documented an effect on reducing mortality. The incidence of detected PCa increased sharply in the United States with the initiation of PSA screening in several large-scale trials in the late 1980s to early 1990s, although the disease incidence has since decreased and somewhat stabilized.58, 59 The increase was accompanied by a decrease in the incidence of high-grade PCa.59 Since then, many organizations, including the AUA and the National Comprehensive Cancer Network, have issued recommendations for screening guidelines.32, 51 More than 90% of cases of PCa are currently detected in the local or regional stage, thus increasing the likelihood of survival. The 10-year relative survival rate for men diagnosed with PCa of all grades in the United States is 91% and 15-year survival rate is 76%.60 Furthermore, since the approval of PSA testing in the United States, the PCa mortality rates in white men have declined to levels below those observed before the diagnostic use of PSA testing.61 Using mortality data from the World Health Organization, a recent study concluded that trends in PCa mortality rates in the United States and 6 other highly developed countries suggest that PSA screening might be effective in reducing PCa mortality.62

PCa Screening Remains Controversial 

Despite the widespread adoption of PCa screening in the United States, it remains controversial. Factors contributing to this controversy include an absence of international consensus about routine screening, a lack of conclusive evidence demonstrating the effectiveness of PSA screening in reducing the mortality due to PCa, and issues relating to overdiagnosis and overtreatment.63, 64 Another area of controversy concerns the steps to be taken in patients screened for PSA and found to have noninvasive PCa. A consensus is lacking on the best treatment options vs watchful waiting.32, 65 In addition, opinions vary and a seeming consensus is lacking among professional peer medical associations such as the American Academy of Family Physicians, American College of Physicians-American Society of Internal Medicine, AUA, and Canadian Task Force on Preventive Health Care with respect to both the merits and the methods of screening (Table 1).66 The question is being addressed by 2 large randomized controlled trials: 1 in Europe, the European Randomized study of Screening for Prostate Cancer (ERSPC), and 1 in the United States, the Prostate, Lung, Colon, and Ovary (PLCO) trial. These studies aim to compare PCa mortality rates in men offered PCa screening against those of men not offered screening.65, 67, 68 The ERSPC trial was initiated in 1993 and includes 8 countries. A total of 267 994 men, of whom the core group was aged 55-69 years, were randomized to the screening (n = 126 219) and control (n = 141 775) groups. A protocol change, effected in 1997 in Rotterdam and shortly afterward at most other sites, replaced DRE and TRUS with PSA levels of 3-4 ng/mL as a biopsy indication, with an additional limited study on PSA cutoff levels of 2-2.9 ng/mL. The primary endpoint of the ERSPC is PCa-specific mortality. Other endpoints include morbidity due to PCa (including the occurrence of metastatic disease) and quality of life. National registry information is being used to evaluate non-PCa-related mortality. From an interim analysis in 2006, the Data Monitoring Committee recommended continuation of the ERSPC. The study is powered to demonstrate a difference of 25% with 10 years of follow-up, and the first complete evaluation is planned once the follow-up data for 2008 have been completed.68 In the Rotterdam cohort of ERSPC, the number of PCa cases detected and the distribution of clinical stage and Gleason score at biopsy confirmed that a relatively large proportion of potentially curable cancers can be found in the low PSA ranges. The PSA velocity did not appear to be a useful screening tool for the identification of these cases.69

Table 1. Guidelines for PCa screening in general population issued by professional medical organizations and government agencies
OrganizationRecommended Guidelines
American Academy of Family Physicians (AAFP)No published standards or guidelines for low-risk patients
American College of Physicians-American Society of Internal Medicine (ACP-ASIM)Physicians should describe potential benefits and known harms of screening, diagnosis, and treatment; listen to patient's concerns, then individualize decision to screen
American Cancer Society (ACS)⁎ and American Urological Association (AUA)Offer annual DRE and PSA screening, beginning at age 50, to men who have ≥10-y life expectancy and to younger men with defined risk factors, including men with first-degree relative who has PCa and black men
American Medical AssociationProvide information regarding the risks and potential benefits of prostate screening
Canadian Task Force on Preventive Health Care (CTFPHC) and U.S. PreventiveDRE and PSA tests not recommended for general population; recommendation against screening for men ≥75 years old
Services Task Force (USPSTF)

DRE, digital rectal examination; PSA, prostate-specific antigen; PCa, prostate cancer.

Reprinted, with permission, from Zoorob et al.66 © American Academy of Family Physicians.

⁎Information should be provided to men about benefits and limitations of testing so that informed decision about testing can be made with clinician's assistance.

The PLCO trial is a randomized controlled trial conducted at 10 centers within the United States, with all participants being followed up for ≥13 years.65, 67 Randomization for the PLCO was conducted from September 1992 to July 2001. Eligible subjects were aged 55-74 years at enrollment and reported no personal history of prostate, lung, colorectal, or ovarian cancer; no current treatment for cancer other than basal or squamous cell skin cancer; no previous surgical removal of the entire prostate or colon or 1 lung; no participation in another cancer screening or primary prevention trial; and no use of finasteride within the previous 6 months. Starting in April 1995, men reporting >1 PSA blood test or any lower gastrointestinal procedure in the previous 3 years were also excluded. The PLCO randomized 76 705 male subjects.67 The primary endpoint of the PLCO is cause-specific mortality for each of the cancers evaluated. Secondary endpoints include cancer incidence, stage shift, and case survival data; the biologic and/or basic prognostic characteristics of the cancers will be correlated with mortality to determine the individual predictive values of these intermediate endpoints.65 The results of the ERSPC and PLCO are due to be reported at the AUA Annual Meeting in 2009, along with a combined analysis. These randomized controlled trials should establish what effect PSA screening has on PCa morbidity and mortality.

Avoiding Underdiagnosis and Overdiagnosis 

A key challenge for physicians is to avoid both underdiagnosis and overdiagnosis of PCa.1 Ideally, cancer screening for diagnostic purposes should reduce the number of individuals who are diagnosed too late, when their disease is incurable (ie, underdiagnosis).57 Screening should increase the number of individuals diagnosed at an early stage who might otherwise have progressed to metastatic disease. The goal is to find these individuals without also including individuals who do not have PCa but, for other reasons, have an increase in PSA (false-positive results) or who would never progress to high-grade or metastatic PCa (nonaggressive cancer). False-positive PSA test results can frequently lead to otherwise unnecessary biopsies, which often cause pain and discomfort. Other issues associated with overdiagnosis and overtreatment include the increased cost (for repeated screenings, biopsy, and treatment), the unnecessary treatment of nonaggressive PCa (including the potential morbidity associated with therapy), a reduction in the patient's quality of life from the effects of treatment, and increased anxiety in the patient and his family.63, 70, 71 Potentially more problematic than false-positive PSA test results is overdiagnosis: the detection of nonaggressive forms of PCa that are considered clinically irrelevant from a treatment perspective (thus obviating the need for biopsy, for example).72, 73 One method of addressing the issue of detecting more clinically relevant (aggressive) PCa in screening paradigms has been to factor in additional markers, along with the PSA values, although this remains controversial.74

When considering the issues surrounding whether to implement universal screening programs, when to begin screening, and the optimal PSA cutoff points, it is, however, important to note that underdiagnosis of clinically relevant PCa (eg, with PSA levels of 4-10 ng/mL) appears more frequently than overdiagnosis and reduces the chances of successful intervention.75, 76 Regions in which PSA testing was most prevalent had a correspondingly lower proportion of men who presented with metastatic disease and a lower PCa mortality rate. PSA screening has been associated with a 70% reduction in the proportion of men who presented with metastatic disease at diagnosis.1 However, >30% of patients treated for localized disease required a secondary intervention for recurrence,1 and these metastases are clinically silent and beyond the means of current detection methods such as bone scans, computed tomography, and magnetic resonance imaging. It is not clear whether curative intervention in men with clinically silent metastasized cancer results in decreased mortality, but it does expose patients to the morbidity of treatment. Lowering the PSA threshold to 2.5 ng/mL would more than double the number of “abnormal” PCa cases—an additional 1.8 million men—if all men aged 40-69 years in the United States were screened (and would also incur a large increase in the false-positive rate).31 Studies have suggested that PCa is overdiagnosed 30%-50% of the time. Other analyses have suggested that we might be treating ≥20 men to keep 1 man from dying of PCa.74 The continuing challenge is to identify which men have disease that will be cured with treatment.

Cost-effectiveness Analyses 

The views are conflicting on the “value for money” that PCa screening with PSA might represent. The first-year costs of treating localized PCa are highly dependent on the initial choice of treatment and the PSA level at diagnosis. In addition, the first-year costs can be associated with the stage at presentation, with a greater stage associated with greater costs.77 One cost-effectiveness analysis, using favorable screening assumptions, determined the marginal cost-effectiveness of screening men aged 65 years using PSA measurement and DRE to be between $12 500 and $15 000 per life-year saved.78, 79, 80 Although the model was based on 1995 dollars, in the United States, from 1993 to 2002, the average resource cost of screening with PSA has decreased by $20.64, which would lead us to believe that the cost is lower now.64 However, in an update to the cost-effectiveness model made in 2002, the additional cost per year per life saved was $12 000–$15 000.81 No recent modeling for the cost of PCa screening (2009) is available; however, these data will likely be published once a mortality benefit has been established.

To the extent that early detection and treatment are effective, savings accrue from avoiding the later costs associated with cancer progression. To estimate the health outcomes of 1-time screening for each of 3 age groups (50-59, 60-69, and 70-79 years), 1 study tracked a hypothetical cohort of 100 000 men, using a Markov process with 6-month cycles. The estimated discounted cost-effectiveness of screening compared with no screening in the 3 age groups of 100 000 men was calculated. The discounted maximal average number of days of life saved per person screened was 11 days for patients 50-59 years old, 7 days for patients 60-69 years old, and 3 days for patients 70-79 years old. The estimates of the discounted average cost per person screened were $216 at 50-59 years, $387 at 60-69 years, and $532 at 70-79 years of age. The costs of the initial test, which are primarily for PSA measurement, are a small component of the overall average cost compared with the major costs that accrue for follow-up after suspicious test results and for treatment of patients proven to have cancer. If the optimistic estimates of treatment benefit are correct, the cost per year of life saved as a result of prevalence screening with DRE and PSA measurement would be $12 491 at 50-59 years, $18 769 at 60-69 years, and $65 909 at 70-79 years of age. Such estimates largely fall within the range of derived cost-effectiveness ratios for many commonly accepted medical practices, including cancer screening.

The aggregate morbidity and mortality that can be attributed to PCa are certainly sufficient to justify a search for effective and efficient strategies for early detection. The analysis of the cost-effectiveness of screening suggests that, given a combination of favorable assumptions, early detection efforts that use PSA measurement might well be cost-effective, at least for men in their 50s and 60s.79

Active Surveillance 

Active surveillance aims to reduce overtreatment of PCa by closely monitoring patients with localized PCa in lieu of immediate therapy. This evolving practice is based on the premise that some, but not all, patients with primary PCa will eventually benefit from active treatment. According to the AUA, the goals of an active surveillance program are “to provide definitive treatment for men with localized cancers that are likely to progress” and “to reduce the risk of treatment-related complications for men with cancers that are not likely to progress.” Although the ideal active surveillance regimen has not been defined, it might include periodic PSA measurement and physical examination or periodic repeat biopsies. The AUA guidelines consider patients with lower risk tumors (in the AUA guidelines, low risk is Gleason score ≤6, PSA level ≤10 ng/mL, clinical stage T1c or T2a; intermediate-risk is PSA >10-20 ng/mL, Gleason score 7, clinical stage T2b, but not qualifying for high risk), particularly those with a shorter life expectancy (eg, older patients), as suitable candidates for active surveillance because their tumors are more likely to remain clinically insignificant during the next 10-15 years after diagnosis (ie, during their lifetime). The AUA guidelines recommend presenting active surveillance as a treatment alternative for high-risk patients (PSA >20 ng/mL, Gleason score 8-10, clinical stage T2c) with clinically localized PCa, with the caveat that the risk of progression is high.32

Back to Article Outline

Conclusions 

The role of PSA measurement in PCa risk assessment has evolved dramatically since its discovery nearly 40 years ago. Since Food and Drug Administration approval of PSA testing in 1986, multiple studies have shown its utility in diagnosing PCa. The PCPT demonstrated the predictive ability of PSA for diagnosing the current risk of PCa, and several studies have shown that PSA also has the capability to assess an individual's risk for developing future PCa. (These studies are discussed at length in a subsequent report by Fleshner in this supplement.) The results of the large-scale, randomized trials (ERSPC and PLCO) designed to document the efficacy of screening programs in reducing mortality are eagerly awaited. (Note: at press time, two reports from these studies were published in the New England Journal of Medicine and are now very important topics in the urology community. A thorough discussion of the results could not be addressed in this article; however, the reader is directed to the following publications for the primary results.82, 83) Strategies such as active surveillance to limit the potential risks of overdiagnosis are being developed and incorporated into practice. However, serum PSA measurement remains the best widely available screening tool for PCa, and a rationale and evidence exists for initiating PSA-based screening in men. Despite its limitations, PSA measurement continues to drive PCa diagnosis and monitoring.

Back to Article Outline

References 

  1. Loeb S, Catalona WJ. What to do with an abnormal PSA test. Oncologist. 2008;13:299–305
  2. Cooner WH, Mosley BR, Rutherford CL, et al. Prostate cancer detection in a clinical urological practice by ultrasonography, digital rectal examination and prostate specific antigen. J Urol. 1990;143:1146–1152
  3. Gann PH, Hennekens CH, Stampfer MJ. A prospective evaluation of plasma prostate-specific antigen for detection of prostatic cancer. JAMA. 1995;25:289–294
  4. Ulmert D, Cronin AM, Bjork T, et al. Prostate-specific antigen at or before age 50 as a predictor of advanced prostate cancer diagnosed up to 25 years later: a case-control study. BMC Med. 2008;6:6–13
  5. Hara M, Inoue T, Koyanagi Y. Preparation and immunoelectrophoretic assessment of antisera to human seminal plasma. Nippon Hoigaku Zasshi. 1996;20:365
  6. Wang MC, Valenzuela LA, Murphy GP, et al. Purification of a human prostate specific antigen: 1979 [classical article]. J Urol. 2002;167:1226–1230
  7. Gutman AB, Gutman EB. An “acid” phosphatase occurring in the serum of patients with metastasizing carcinoma of the prostate gland. J Clin Invest. 1938;17:473–478
  8. Bensalah K, Lotan Y, Karam JA, et al. New circulating biomarkers for prostate cancer. Prostate Cancer Prostatic Dis. 2008;11:112–120
  9. Scardino PT. The Gordon Wilson lecture: natural history and treatment of early stage prostate cancer. Trans Am Clin Climatol Assoc. 2000;111:201–241
  10. Christensson A, Bjork T, Nilsson O, et al. Serum prostate specific antigen complexed to α1-antichymotrypsin as an indicator of prostate cancer. J Urol. 1993;150:100–105
  11. Mikolajczyk SD, Millar LS, Wang TJ, et al. “BPSA,” a specific molecular form of free prostate-specific antigen, is found predominantly in the transition zone of patients with nodular benign prostatic hyperplasia. Urology. 2000;55:41–45
  12. Mikolajczyk SD, Millar LS, Wang TJ, et al. A precursor form of prostate-specific antigen is more highly elevated in prostate cancer compared with benign transition zone prostate tissue. Cancer Res. 2000;60:756–759
  13. Mikolajczyk SD, Catalona WJ, Evans CL, et al. Proenzyme forms of prostate-specific antigen in serum improve the detection of prostate cancer. Clin Chem. 2004;50:1017–1025
  14. Catalona WJ, Bartsch G, Rittenhouse HG, et al. Serum pro-prostate specific antigen preferentially detects aggressive prostate cancers in men with 2 to 4 ng/ml prostate specific antigen. J Urol. 2004;171(6 Pt. 1):2239–2244
  15. Lee F, Torp-Pedersen ST, Siders DB, et al. Transrectal ultrasound in the diagnosis and staging of prostatic carcinoma. Radiology. 1989;170(3 Pt. 1):609–615
  16. Mettlin C, Murphy GP, Lee F, et al. Investigators of the American Cancer Society-National Prostate Cancer Detection Project Characteristics of prostate cancers detected in a multimodality early detection program. Cancer. 1993;72:1701–1708
  17. Littrup PJ, Lee F, Mettlin C. Prostate cancer screening: current trends and future implications. CA Cancer J Clin. 1992;42:198–211
  18. Lowe FC, Trauzzi SJ. Prostatic acid phosphatase in 1993: its limited clinical utility. Urol Clin North Am. 1993;20:589–595
  19. Makarov DV, Carter HB. The discovery of prostate specific antigen as a biomarker for the early detection of adenocarcinoma of the prostate. J Urol. 2006;176(6 Pt. 1):2383–2385
  20. Romero FR, Romero KR, Brenny FT, et al. Reasons why patients reject digital rectal examination when screening for prostate cancer. Arch Esp Urol. 2008;61:759–765
  21. Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk: results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2006;98:529–534
  22. Borden LS, Wright JL, Kim J, et al. An abnormal digital rectal examination is an independent predictor of Gleason ≥7 prostate cancer in men undergoing initial prostate biopsy: a prospective study of 790 men. BJU Int. 2007;99:559–563
  23. Stamey TA, Yang N, Hay AR, et al. Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med. 1987;317:909–916
  24. Catalona WJ, Smith DS, Ratliff TL, et al. Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med. 1991;324:1156–1161
  25. Hernandez DJ, Han M, Humphreys EB, et al. Predicting the outcome of prostate biopsy: comparison of a novel logistic regression-based model, the prostate cancer risk calculator, and prostate-specific antigen level alone. BJU Int. 2009;103:609–614
  26. Brawer MK, Beatie J, Wener MH, et al. Screening for prostatic carcinoma with prostate specific antigen: results of the second year. J Urol. 1993;150:106–109
  27. Brawer MK, Chetner MP, Beatie J, et al. Screening for prostatic carcinoma with prostate specific antigen. J Urol. 1992;147(3 Pt. 2):841–845
  28. Terris MK, Stamey TA. Utilization of polyclonal serum prostate specific antigen levels in screening for prostate cancer: a comparison with corresponding monoclonal values. Br J Urol. 1994;73:61–64
  29. Hernández J, Thompson IM. Prostate-specific antigen: a review of the validation of the most commonly used cancer biomarker. Cancer. 2004;101:894–904
  30. Henkel J. Prostate cancer: new tests create treatment dilemmas. FDA Consumer. December 1994;http://www.fda.gov/fdac/reprints/prostate.htmlUpdated 1997. Accessed December 16, 2008
  31. Welch HG, Schwartz LM, Woloshin S. Prostate-specific antigen levels in the United States: implications of various definitions for abnormal. J Natl Cancer Inst. 2005;97:1132–1137
  32. Thompson I, Thrasher JB, Aus G, et al. Guideline for the management of clinically localized prostate cancer: 2007 update. J Urol. 2007;177:2106–2131
  33. Bartsch G, Horninger W, Klocker H, et al. Tyrol Prostate Cancer Demonstration Project: early detection, treatment, outcome, incidence and mortality. BJU Int. 2008;101:809–816
  34. Carter HB, Morrell CH, Pearson JD, et al. Estimation of prostatic growth using serial prostate-specific antigen measurements in men with and without prostate disease. Cancer Res. 1992;52:3323–3328
  35. D'Amico AV, Hanks GE. Linear regressive analysis using prostate-specific antigen doubling time for predicting tumor biology and clinical outcome in prostate cancer. Cancer. 1993;72:2638–2643
  36. Carter HB, Pearson JD. PSA velocity for the diagnosis of early prostate cancer: a new concept. Urol Clin North Am. 1993;20:665–670
  37. Lilja H, Ulmert D, Vickers AJ. Prostate-specific antigen and prostate cancer: prediction, detection and monitoring. Nat Rev Cancer. 2008;8:268–278
  38. Oesterling JE, Jacobsen SJ, Chute CG, et al. Serum prostate-specific antigen in a community-based population of healthy men: establishment of age-specific reference ranges. JAMA. 1993;270:860–864
  39. Morgan TO, Jacobsen SJ, McCarthy WF, et al. Age-specific reference ranges for prostate-specific antigen in black men. N Engl J Med. 1996;335:304–310
  40. Evans S, Metcalfe C, Ibrahim F, et al. Differences in prostate cancer prognosis: a systematic review and metal-analysis. Int J Cancer. 2008;123:430–435
  41. Catalona WJ, Smith DS, Ornstein DK. Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination: enhancement of specificity with free PSA measurements. JAMA. 1997;277:1452–1455
  42. Punglia RS, D'Amico AV, Catalona WJ, et al. Effect of verification bias on screening for prostate cancer by measurement of prostate-specific antigen. N Engl J Med. 2003;349:335–342
  43. Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level ≤4.0 ng per milliliter. N Engl J Med. 2004;350:2239–2246
  44. Andriole GL, Roehrborn C, Schulman C, et al. Effect of dutasteride on the detection of prostate cancer in men with benign prostatic hyperplasia. Urology. 2004;64:537–541
  45. Guess HA, Heyse JF, Gormley GJ, et al. Effect of finasteride on serum PSA concentration in men with benign prostatic hyperplasia: results from the North American phase III clinical trial. Urol Clin North Am. 1993;20:627–636
  46. Thompson IM, Chi C, Ankerst DP, et al. Effect of finasteride on the sensitivity of PSA for detecting prostate cancer. J Natl Cancer Inst. 2006;98:1128–1133
  47. Andriole G, Bostwick D, Brawley O, et al. Chemoprevention of prostate cancer in men at high risk: rationale and design of the reduction by dutasteride of prostate cancer events (REDUCE) trial. J Urol. 2004;172(4 Pt. 1):1314–1317
  48. Oesterling JE. Prostate-specific antigen and diagnosing early malignancies of the prostate. J Cell Biochem Suppl. 1992;16H:31–43
  49. el-Shirbiny AM. Prostatic specific antigen. Adv Clin Chem. 1994;31:99–133
  50. Heidenreich A, Aus G, Bolla M, et al. EAU guidelines on prostate cancer. Eur Urol. 2008;53:68–80
  51. National Comprehensive Cancer Network. NCCN clinical practice guidelines in oncology: prostate cancer early detection. V.2 http://www.nccn.org2007;Updated May 10, 2007. Accessed December 24, 2008
  52. Carter HB, Ferrucci L, Kettermann A, et al. Detection of life-threatening prostate cancer with prostate-specific antigen velocity during a window of curability. J Natl Cancer Inst. 2006;98:1521–1527
  53. Ashley T. Using predictive value, sensitivity and specificity to interpret laboratory tests: PSA for the diagnosis of prostate cancer. J Insur Med. 2005;37:261–263
  54. Klein LT, Lowe FC. The effects of prostatic manipulation on prostate-specific antigen levels. Urol Clin North Am. 1997;24:293–297
  55. Erdogan K, Gurdal M, Tekin A, et al. The effect of urethral catheterisation on serum prostate-specific antigen levels in male patients with acute urinary retention. Yonsei Med J. 2003;44:676–678
  56. Parekh N, Lin Y, Marcella S, et al. Associations of lifestyle and physiologic factors with prostate-specific antigen concentrations: evidence from the National Health and Nutrition Examination Survey (2001-2004). Cancer Epidemiol Biomarkers Prev. 2008;17:2467–2472
  57. Bangma CH, Roemeling S, Schröder FH. Overdiagnosis and overtreatment of early detected prostate cancer. World J Urol. 2007;25:3–9
  58. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008. CA Cancer J Clin. 2008;58:71–96
  59. Thompson IM, Ankerst DP. Prostate-specific antigen in the early detection of prostate cancer. Can Med Assoc J. 2007;176:1853–1858
  60. American Cancer Society. Cancer facts and figures 2008. http://www.cancer.org/docroot/stt/stt_0.aspAccessed December 9, 2008
  61. Tarone RE, Chu KC, Brawley OW. Implications of stage-specific survival rates in assessing recent declines in prostate cancer mortality rates. Epidemiology. 2000;11:167–170
  62. Bouchardy C, Fioretta G, Rapiti E, et al. Recent trends in prostate cancer mortality show a continuous decrease in several countries. Int J Cancer. 2008;123:421–429
  63. Clements A, Watson E, Rai T, et al. The PSA testing dilemma: GPs' reports of consultations with asymptomatic men: a qualitative study. BMC Fam Pract. 2007;8:35–41
  64. Ekwueme DU, Stroud LA, Chen Y. Cost analysis of screening for, diagnosing, and staging prostate cancer based on a systematic review of published studies. Prev Chronic Dis. 2007;4:A100–A116
  65. Prorok PC, Andriole GL, Bresalier RS, et al. Design of the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial. Control Clin Trials. 2000;21(suppl):273S–309S
  66. Zoorob R, Anderson R, Cefalu C, Sidani M. Cancer screening guidelines. Am Fam Physician. 2001;63:1101–1112
  67. Andriole GL, Levin DL, Crawford ED, et al. Prostate cancer screening in the Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial: findings from the initial screening round of a randomized trial. J Natl Cancer Inst. 2005;97:433–438
  68. Schröder FH. Screening for prostate cancer (PC)—an update on recent findings of the European Randomised Study of Screening for Prostate Cancer (ERSPC). Urol Oncol. 2008;26:533–541
  69. Roobol MJ, Roobol DW, Schröder FH. Is additional testing necessary in men with prostate-specific antigen levels of 1.0 ng/mL or less in a population-based screening setting? ((ERSPC, section Rotterdam)). Urology. 2005;65:343–346
  70. U.S. Preventive Services Task Force recommendation statement. Screening for prostate cancer. Ann Intern Med. 2008;149:185–191
  71. Lin K, Lipsitz R, Miller T, et al. Benefits and harms of prostate-specific antigen screening for prostate cancer: an evidence update for the U.S. Preventive Services Task Force. Ann Intern Med. 2008;149:192–199
  72. Davidov O, Zelen M. Overdiagnosis in early detection programs. Biostatistics. 2004;5:603–613
  73. Etzioni R, Penson DF, Legler JM, et al. Overdiagnosis due to prostate-specific antigen screening: lessons from U.S. prostate cancer incidence trends. J Natl Cancer Inst. 2002;94:981–990
  74. Taichman RS, Loberg RD, Mehra R, et al. The evolving biology and treatment of prostate cancer. J Clin Invest. 2007;117:2351–2361
  75. Graif T, Loeb S, Roehl KA, et al. Under diagnosis and over diagnosis of prostate cancer. J Urol. 2007;178:88–92
  76. Pelzer AE, Bektic J, Akkad T, et al. Under diagnosis and over diagnosis of prostate cancer in a screening population with serum PSA 2 to 10 ng/ml. J Urol. 2007;178:93–97
  77. Penson DF, Schonfeld WH, Flanders SC, et al. Relationship of first-year costs of treating localized prostate cancer to initial choice of therapy and stage at diagnosis: results from the CaPSURE database. Urology. 2001;57:499–503
  78. Barry MJ, Fleming C, Coley CM, et al. Should Medicare provide reimbursement for prostate-specific antigen testing for early detection of prostate cancer? (Part IV: estimating the risks and benefits of an early detection program). Urology. 1995;46:445–461
  79. Coley CM, Barry MJ, Fleming C, et al. Early detection of prostate cancer (Part II: estimating the risks, benefits, and costs). Ann Intern Med. 1997;126:468–479
  80. Crawford ED, Abrahamsson PA. PSA-based screening for prostate cancer: how does it compare with other cancer screening tests?. Eur Urol. 2008;54:262–273
  81. Harris R, Lohr KN. Screening for prostate cancer: an update of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med. 2002;137:917–929
  82. Andriole GL, Grubb RL, Buys SS, et al. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med. 2009;360:1310–1319
  83. Schröder FH, Hugosson J, Roobol MJ, et al. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med. 2009;360:1320–1328

 K. J. Pienta, M.D., is a consultant to GlaxoSmithKline.

PII: S0090-4295(09)00237-4

doi:10.1016/j.urology.2009.02.016

Refers to erratum:

Urology
Volume 73, Issue 5, Supplement , Pages S11-S20, May 2009

Article Tools

* Image Usage & Resolution