DRAFT: This module has unpublished changes.

Prostate Specific Antigen Test: a review of

prostate cancer screening on cause-specific death



A recommendation by the Federal Advisory Group on downgrading the use of Prostate Specific Antigen (PSA) test was recently proposed. The panel recommends cessation of such screening as part of preventative exams for the general population of older males due to its inconsistent efficacy results, feasibility, and potential overdiagnosis. Prostate cancer (PC) is one of the leading causes of death from malignant disease among men in the developed world.1 Early stage diagnosis by high PSA test results followed with curative treatment is the current, leading strategy to improve survival in PC. Simultaneously, high PSA test results are often due to diagnoses other than cancer generating costly biopsies, as well as unnecessary anxiety in the patients.  

Analyzing what is known regarding the PSA test as a preventative screen for PC is useful for several reasons. First, definitive evidence of PSA testing as a way to significantly reduce prostate-cancer mortality is needed to justify screening costs and potentially subjecting patients to unnecessary angst. Secondly, if PSA testing was to continue, a unanimous PSA level for detecting cancer needs to be established. Finally, once an ideal PSA level is determined, an algorithm of PC screening needs to be created for scaling up and adapting successful interventions to new contexts.


Literature review methodology

Six leading articles on PSA screening and prostate cancer were provided by Dr. Rich Feeley for this literature review.


Main findings

With the exception of Etzioni et al.’s study regarding PC overdiagnosis and Hugosson et al.’s study suggesting free and total PSA measurement, prostate cancer mortality was the primary end-point.  Table 1 provides the summary findings of the reviewed articles listed by publishing date. Four of the reviewed articles supported PSA screening while the remaining two articles provided evidence against PSA screening, highlighted in blue and red respectively in Table 1.



Time of Study


Study Type

Study Population

Selection Criteria

Statistical Results

Study Outcome

Labrie, 19991

November 1988 - December 1996

Quebec, Canada

- Randomized

- Prospective

46,193 men aged 45-80

Registered in electoral roll of Quebec City & Metropolitan area

The 8-year PC death IR is 67.1% lower in men of the screened group (p=0.02)

PSA screening started at age 50 is highly efficient to identify men who are at high risk or having PC

Etzioni, 20022



- Hypothetical cohort


2,000,000 men aged 60-84

US census data

The PC overdiagnosis rates were 28.8% for white men and 43.8% in black men

- PSA screening resulted in considerable overdiagnosis rates of PC

- Only a minority of PC cases found at autopsy would have been detected by PSA testing

Hugosson, 20033

January 1995

Göteborg, Sweden

- Randomized

- Population-based

20,000 men aged 50-64

Random sampling from population register

Sensitivity and specificity of PSA screen can be increased by incorporating f/tPSA with new tPSA threshold of <4ng/ml

PSA screening effectively detects early-stage low-grade PC

Labrie, 20044

November 1988 - 1999

Quebec, Canada

- Randomized

- Prospective

46,486 men aged 45-80

Registered in electoral roll of Quebec City & Metropolitan area

The 8-year PC death IR is 62% lower in group screened for PC vs group receiving standard medical practice

Results are in agreement with continuous decrease of PC mortality observed in North America

Lu-Yao, 20085


Seattle-Puget Sound and Connecticut, USA

- Ecological

- Cohort

215,521 men aged 65-79 (Seattle = 94,900, Connecticut = 120,621)

Male Medicare beneficiaries

There was no significant difference in PC mortality between Seattle and Connecticut cohorts (RR=1.02; 95% CI 0.96-1.09)

PC screening and intensive treatment were not associated with lower PC mortality in men aged 65+

Hugosson, 20106

December 1994 - 2008

Göteborg, Sweden

- Randomized

- Population-based

20,000 men aged 50-64

Random sampling from population register

The absolute CI reduction of death from PC at 14 years was 0.40% (95% CI 0.17-0.64) in the screening group vs control group

Benefit of PC screening compares favorably to other cancer screening programs

Table 1: Summary findings of selected PSA studies


  • Follow-up treatment after diagnosis. It is important to note that the treatment options available between studies were not the same. Treatment options include radical prostatectomy, radiation therapy, hormonal therapy, and active surveillance. Since PC mortality is the primary end-point, PC treatment after diagnosis is just as important as the early detection. It is interesting to note that the four studies supporting PSA screening were done in Canada and Sweden where health care are publicly-funded and government-funded respectively; whereas the two studies opposing PSA screening were done in the USA where health care is largely owned and operated by the private sector. Since treatment options are not equally effective and some treatment options may be financially restricting to patients, PC mortality will be affected by the available treatments after diagnosis.     
  • Age group of study population. Different definitions of “older male population” were used between studies ranging from 45 to 84 years of age. Depending on the time of PC diagnosis, age could affect the decision on the course of action taken by the patient. The four articles supporting PSA screening used study groups aged 45-80 and 50-64 years of age; whereas the two articles opposing PSA screening used older study groups aged 60-84 and 65-79 years of age. Early PC detection in “younger” men may affect the aggressiveness of the treatment option taken after diagnosis or affect the success of the treatment and/or recovery. Furthermore, it is more likely that the PC detected in “younger” men are still localized compared to the PC detected in older patients. It is well recognized that prevention of PC death is achieved by treatment of the localized disease. Perhaps PSA screening should be strongly recommended as preventative exam for men aged 45-60 to detect any early stages of PC and become optional for men aged 60+ due to the general slow growth of PC.7 With this protocol, PSA screening would be in effect during the time when aggressive treatment is a more likely option. Further studies should be conducted to confirm the ideal age range for early stage PC detection. 
  • Cost feasibility. Taking the PC screening algorithm proposed by Labrie et al., the estimated cost of the screen is $2,665 per cancer at first visit.1 This estimated cost may very well have inflated since its publishing in 1999; however, this cost was relatively low compared to the costs estimated for cervical and breast cancer diagnosed by screening ($10,000 and $30,000 respectively) which would have inflated with time as well. Men should be offered PC screening like women are offered breast and cervical cancer screening to provide best chance of cancer survival.
  • Ideal PSA level for cancer. From the four studies supporting PSA screening, the PSA level used for further testing were not unanimous. PSA levels varying from 2.5-3.4 ng/ml were used as the upper limit of normal levels in the studies done by Labrie, et al. and Hugosson, et al. Further studies need to be conducted to determine the ideal PSA level that maximize correct PC diagnosis. Incorporating Hugosson et al.’s method of measuring free PSA and total PSA ratio as the standard screening protocol could increase diagnostic accuracy. A unanimous agreement needs to be established to maximize sensitivity and specificity before scaling up PSA screening.
  • Overdiagnosis. Several of the studies identified concern of PC overdiagnosis as an issue. Etzioni et al.’s study illustrated that there is a considerable overdiagnosis among PSA-detected cases causing unnecessary anxiety in patients and incurring excessive costs. However, it should be noted that the data used for this study was derived from USA’s SEER registry. In the United States, the fee-for-service system can provide powerful incentives for physicians to provide more treatments and tests – including unnecessary ones. This conflict of interest could explain the observed trends seen in the study. Nevertheless, overdiagnosis should not be overlooked and efforts should be made to improve accuracy of PSA screening. A thorough explanation should be provided with PSA screening to reduce unnecessary angst in patients. Perhaps studies should be conducted to see how men respond and cope emotionally to PSA testing and its results.


Risk status is important even if active surveillance is the only action patients choose to take after diagnosis and PSA screening should be continually provided to men. However, for such screening to be financially justifiable and effective in reducing PC mortality, several issues need to be addressed: 1) a unanimous establishment of PSA level for PC diagnosis, 2) a unanimous establishment of screening age, and 3) treatment options must be as readily available as the screening.




1 Labrie, F., Candas, B., Dupont, A., Cusan, L., Gomez, J., Suburu, R.E., Diamond, P., Levesque, J., Belanger, A. (1999). Screening Decreases Prostate Cancer Death: First Analysis of the 1988 Quebec Prospective Randomized Controlled Trial. The Prostate, 38, 83-91.


2 Etzioni, R., Penson, D.F., Legler, J.M., Tommaso, D., Boer, R., Gann, P.H., Feuer, E.J. (2002). Overdiagnosis Due to Prostate-Specific Antigen Screening: Lessons From U.S. Prostate Cancer Incidence Trends. Journal of the National Cancer Institute, 94(13), 981-990.


3 Hugosson, J., Aus, G., Bergdahl, S., Fernlund, P., Frosing, R., Lodding, P., Pihl, C., Lilja, H. (2003). Population-based screening for prostate cancer by measuring free and total serum prostate-specific antigen in Sweden. BJU International, 92, 39-43.


4 Labrie, F., Candas, B., Cusan, L., Gomez, J.L., Belanger, A., Brousseau, G., Chevrette, E., Levesque, J. (2004). Screening Decreases Prostate Cancer Mortality: 11-Year Follow-Up of the 1988 Quebec Prospective Randomized Controlled Trial. The Prostate, 59, 311-318.


5 Lu-Yao, G., Albertsen, P.C., Stanford, J.L., Stukel, T.A., Walker-Corkery, E., Barry, M.J. (2008). Screening, Treatment, and Prostate Cancer Mortality in the Seattle Area and Connecticut: Fifteen-year Follow-up. Journal of General Internal Medicine, 23(11), 1809-1814.


6 Hugosson, J., Carlsson, S., Aus, G., Bergdahl, S., Khatami, A., Lodding, P., Pihl, C., Stranne, J., Holmberg, E., Lilja, H. (2010). Mortality results from the Goteborg randomized population-based prostate-cancer screening trial. The Lancet, 11, 725-732.


7 Gronberg, H., Damber, J-E., Jonsson, H., Lenner, P. (1994). Patient age as a prognostic factor in prostate cancer. Journal of Urology, 152, 892-295.

DRAFT: This module has unpublished changes.