Prostate cancer is the most commonly diagnosed cancer in Canadian men and their third-highest cause of cancer death 1. According to the Canadian Cancer Society, Canadian men carry a 14.3% lifetime chance of diagnosis and a 3.7% risk of dying from the disease 2.The discovery of prostate-specific antigen (PSA) in 1970 and subsequent development and approval of measurement assays for the compound have been the subject of immense hope that it could serve as a screening marker for the disease 3,4. However, the definitive answer as to PSA’s usefulness as a screening and detection mechanism for prostate cancer has been elusive and subject to much controversy. Indeed, a recent commentary in the New England Journal of Medicine labeled the PSA-screening debate “the controversy that refuses to die” 5.
PSA is a kallikrein-like serine protease produced by the columnar epithelial cells surrounding the ducts of the prostate and periurethral glands. It is secreted into the prostatic ducts and makes up part of the seminal fluid. Studies looking at the characteristics of PSA assays have estimated the positive predictive value anywhere between 30% and 54%; these values are actually higher than the 20% positive predictive value observed with mammographic screening for breast cancer 6.
The use of PSA levels to detect the presence of prostate cancer has proliferated, with most men in the United States and many more in Canada having undergone this test. One paper reports that 95% of American urologists and 78% of primary-care physicians have undergone one themselves 5; another estimates that 72% of all men in Canada have had a screening PSA test 7. PSA testing has been said to be nearly as common in men, or even more so, as measurements of serum cholesterol 8.
Some experts in the field have spoken out against what they see as overreliance on a substandard test with poor diagnostic accuracy leading to unnecessary treatment, with attendant high rates of morbidity. Richard Ablin, who discovered PSA 9, and his colleagues have noted that PSA is prostate specific, but not actually cancer specific. Many non-malignant conditions, including simple prostate growth, can elevate PSA levels 7. Indeed, prostate cancer cells produce less PSA than normal prostate cells 6. Given that between 45% and 80% of men between the ages of 50 and 75 years have histological evidence of prostate cancer 10, though not necessarily clinical disease, Ablin characterizes the diagnostic ability of PSA screening as “slightly better than that of flipping a coin” 7.
Whether by detection of the cancer or by chance, many men who undergo biopsy or another diagnostic intervention will have small amounts of low-grade cancer diagnosed that would never end up being clinically significant in their lifetime 11. In short, the only reason these men could ever know they have prostate cancer is that they or their doctor decided to look for it 12; if left alone, their disease would never trouble them at all over their entire lives, much less kill them. Following diagnosis, these men end up receiving unnecessary treatment that subjects them to risks of urinary incontinence of up to 30% and impotence of 30%–60% 13, in addition to the stress and anxiety that comes from a cancer diagnosis.
In light of these questions, health authorities are divided in their approach to recommendations on PSA screening for detection of prostate cancer. In 1994, the Canadian Task Force on Preventive Health Care recommended against the use of PSA screening for prostate cancer and has not revised its recommendation since 14. The Toward Optimized Practice Program of Alberta, however, stresses that PSA testing is “critical in management of men with prostate cancer” and that physicians should discuss the risks and benefits of PSA testing with their patients and offer it to well-informed men who desire testing 15. The American Urological Association and American Cancer Society agree with this approach, recommending that doctors offer screening at age 50 years after discussing risks and benefits 16. Conversely, the U.S. Preventive Services Task Force 17, the World Health Organization 18, and the National Health Service in the United Kingdom 19 do not recommend PSA screening.
Several recent, large, randomized controlled trials have tried to evaluate whether screening for prostate cancer with PSA confers a mortality benefit. One is the large, multicenter Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) from the United States 16. This trial enrolled over 75,000 participants between the ages of 55 and 74 years and randomized them to screening and a control group. These men were offered annual PSA screening for 6 years and received a digital rectal examination (DRE) as well for the first 4 screening examinations. Biopsies were performed for PSA values greater than 4.0 mg/L (ng/mL) or after an abnormal DRE. Results reported after 7 years did not show a statistically significant benefit or harm from screening; the reported rate ratio of 1.13 (95% CI: 0.75–1.70), however, seemed to point to an increase in death in the screening group compared with the controls. Preliminary results from 10 years of follow-up agreed with these findings, reporting a rate ratio of 1.11 (95% CI: 0.83–1.50). However, relatively short follow-up and significant contamination from men in the control group receiving PSA tests has led to criticism of this trial 20.
The European Randomized Study of Screening for Prostate Cancer (ERSPC) ran concurrently with the PLCO trial 21. This study spanned several centers in multiple European countries and was significantly larger than PLCO, with over 160,000 men aged 55–69 years enrolled. These men were screened every 4 years with PSA and optional DRE. The criteria for biopsy were not uniform across all centers, though most used a cutoff of 3.0 mg/L (ng/mL) (see paper for details). After a median follow-up of 9 years, a statistically significant rate ratio of 0.80 (95% CI: 0.65–0.98) was reported for prostate cancer mortality in the screening group. It was noted, however, that the absolute difference in prostate cancer death between the screening and control groups was 0.71 deaths per 1000 men, meaning that 1410 men would need to be screened and 48 treated to prevent a single death. The study has also been criticized for being designed to show a 25% relative mortality benefit, but reporting a 20% benefit, and for non-uniformity of screening techniques across sites 7.
A separately reported analysis from 1 of the ERSPC sites reported a more pronounced benefit for screening 22. The Swedish arm enrolled 20,000 men aged 50–64 years from the town of Göteborg, beginning screening at a younger age than in the PLCO trial. They were screened with PSA tests every other year for a median follow-up of 14 years, and further investigation was initiated if the PSA reading was elevated (initially >3.4 mg/L [ng/mL], then adjusted to 3.0 mg/L [ng/mL] to match the other ERSPC sites). This study reported a relative risk of 0.56 (95% CI: 0.39–0.82) from prostate cancer screening, corresponding to an absolute risk reduction from 0.90 deaths per thousand in the control group to 0.50 per thousand in the screening group. This means that 293 men needed to be screened and 12 treated to prevent 1 prostate cancer death. Interestingly, the men in this study were recruited through a government population registry and randomized prior to being enrolled in the study. Thus, this study was able to examine the effect of non-compliance in a population of men offered screening. The Göteborg group reported that 76% of men invited for screening had at least 1 PSA measurement, slightly less than in the PLCO trial (85%) and the ERSCP trial (82%). The Göteborg study did, however, have a significantly smaller number of patients enrolled in it than the ERSPC and PLCO trials—in fact, it was just 1 arm of the ERSPC—which may limit its applicability.
A study in Quebec followed over 45,000 men between ages 45 and 80 years for a median follow-up of 8 years 23. Patients in the screening arm received annual PSA testing with a DRE at first visit. The threshold for further investigation was 3.0 mg/L (ng/mL). They reported a relative risk of 0.385 (95% CI: 0.207–0.719) for prostate cancer death in the screening arm compared with the control group, the largest reduction reported for any of the studies discussed here. This study has been criticized, however, for low response rates to their invitation for screening, improper randomization 24 and intention-to-treat procedures due to crossover between the 2 study groups, and the lack of reporting on contamination of the screening group 25.
A systematic review and meta-analysis published in the British Medical Journal 26 concluded that PSA screening does not reduce mortality from prostate cancer or all-cause mortality in men with the disease. However, a review and commentary from the Annals of Internal Medicine 27 argued that several of the 6 randomized controlled trials included in the review (which included several of the aforementioned studies) used short follow-up times and were susceptible to strong bias through contamination of control groups, potentially leading to an understatement of the protective effect of screening. The review also praises the design of the Göteborg study and posits that that study’s result may actually be more applicable to practice than that of the meta-analysis.
Some trials concluded that there are no mortality benefits from PSA screening, while others have disagreed. However, even if a mortality benefit exists from screening, is it worth the cost? As mentioned before, extant treatments for prostate cancer carry significant risks of morbidity. If screening does save lives, the benefits of longevity must be balanced with the risks of treatment. The magnitude of the measured benefit from screening varied greatly in the trials, with the number needed to treat ranging from 48 in the complete ERSPC to 12 in the Göteborg study. The latter result compares favorably with those for mammography, for which a Cochrane review showed that 10 women needed to be treated to prevent 1 death 28. Studies of fecal occult blood testing for colorectal cancer have estimated that 1173 people need to be screened to prevent 1 death 29, a number close to those reported in the PSA-screening studies. If the results of the Göteborg studies are correct, PSA screening seems to compare quite favorably with these other, well-established cancer-screening methods.
Dr. Barry’s aphorism seems truer than ever—whatever the merits, the PSA controversy simply refuses to die. The discussions around the topic facilitated by Current Oncology on the Cancer Knowledge Network are sure to be heated, with many passionate voices on both sides of the aisle. We hope that this article has provided a relevant snapshot of the current data on the role of PSA in prostate cancer screening.
The literature on PSA and its applicability to prostate cancer practice is vast, and only a small sampling has been presented here. Armed with a solid basis of the present state of the science, the reader should continue to explore the topic and participate in the lively debates at the Cancer Knowledge Network. As oncologists, we all share the goal of providing our patients with the most accurate, up-to-date information and treatment options that will maximize their longevity and quality of life. Your participation in discussions about PSA screening will help all cancer-care professionals achieve that goal.
- Marrett LD, De P, Airia P, Dryer D; Steering Committee of Canadian Cancer Statistics 2008. Cancer in Canada in 2008. CMAJ 2008;179:1163–70. [PMID: 19015567]
- Canadian Cancer Society. Cancer Statistics 2009. Toronto, ON: The Society; 2009.
- Haythorn MR, Ablin RJ. Prostate-specific antigen (PSA) testing across the disease continuum. Prostate Cancer Commun 2009;25:3–9.
- Sirovich BE, Schwartz LM, Woloshin S. Screening men for prostate and colorectal cancer in the United States: does practice reflect the evidence? JAMA 2003;289:1414–20. [PMID: 12636464]
- Barry MJ. Screening for prostate cancer: the controversy that refuses to die. N Engl J Med 2009;360:1351–4. [PMID: 19297564]
- Kantoff PW, Carroll PR, D’Amico AV, eds. Prostate Cancer: Principles and Practice. Philadelphia, PA: Lippincott Williams & Wilkins; 2002: 195–206.
- Ablin RJ, Haythorn MR. Screening for prostate cancer: Controversy? What controversy? Curr Oncol 2009;16:1–2. [PMID: 19526077]
- Boyle P. Screening for prostate cancer: have you had your cholesterol measured? BJU Int 2003;92:191–99. [PMID: 12887466]
- Rao AR, Motiwala HG, Karim OM. The discovery of prostate-specific antigen. BJU Int 2008;101:5–10. [PMID: 17760888]
- Sakr WA, Grignon DJ, Haas GP, Heilbrun LK, Pontes JE, Crissman JD. Age and racial distribution of prostatic intraepithelial neoplasia. Eur Urol 1996;30:138–44. [PMID: 8875194]
- Klotz L. Active surveillance for prostate cancer. Curr Urol Rep 2010;11:165–71. [PMID: 20425623]
- Welch HG, Schwartz LM, Woloshin S. Prostate-specific antigen levels in the United State: implications of various definitions for abnormal. J Natl Cancer Inst 2005;97:1132–7. [PMID: 16077071]
- Paker C, Muston D, Melia J, Moss S, Dearnaley D. A model of the natural history of screen-detected prostate cancer, and the effect of radical treatment on overall survival. Br J Cancer 2006;94:1361–8. [PMID: 16641912]
- Feightner JW. Screening for prostate cancer. In: Canadian Task Force on the Periodic Health Examination. Canadian Guide to Clinical Preventive Health Care. Ottawa, ON: Health Canada; 1994: 812–23.
- Toward Optimized Practice Program. Prostate Cancer: Guideline for the Use of PSA and the Early Diagnosis Edmonton, Alberta: TOP; 1997.
- Andriole GL, Crawford ED, Grubb RL 3rd, Buys SS, Chia D, Church TR, et al; PLCO Project Team. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med 2009;360:1310–9. [PMID: 19297565]
- Lin K, Lipsitz R, Miller T, Janakiraman S; U.S Preventive Services Task Force. 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;149:192–9. [PMID: 18678846]
- World Health Organization. Screening for prostate cancer. World Health Organization Web site. http://www.who.int/cancer/detection/prostatecancer/en/index.html. c2011. Accessed January 31, 2011.
- Wint A, Persad R, Bahl A. New Guidelines on PSA Testing in Primary Care. Bristol, UK: Avon, Somerset, and Wiltshire Cancer Services; 2009.
- Catalona WJ. Prostate-cancer screening. N Engl J Med 2009;361:202. [PMID: 19587348]
- Schroder FH, Hugosson J, Roobol MJ, Tammela TLJ, Ciatto S, Nelen V, et al; ERSPC Investigators. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med 2009;360:1320–8. [PMID: 19297566]
- Hugosson J, Carlsson S, Aus G, Bergdahl S, Khatami A, Lodding P, et al. Mortality results from the Göteborg randomised population-based prostate-cancer screening trial. Lancet Oncol 2010;11:725–32. [PMID: 20598634]
- Labrie F, Candas B, Cusan L, Gomez JL, Bélanger A, Brousseau G, et al. Screening decreases prostate cancer mortality: 11-year follow-up of the 1988 Quebec prospective randomized controlled trial. Prostate 2004;59:311–8. [PMID: 15042607]
- Elwood M. A misleading paper on prostate cancer screening. Prostate 2004;61:372. [PMID: 15503318]
- Pinsky PF. Results of a randomized controlled trial of prostate cancer screening. Prostate 2004;61:371. [PMID: 15503321]
- Djulbegovic M, Beyth RJ, Neuberger MM, Stoffs TL, Vieweg J, Djulbegovic B, et al. Screening for prostate cancer: systematic review and meta-analysis of randomised controlled trials. BMJ 2010;341:c4543. [PMID: 20843937]
- Luckmann R. Review: prostate cancer screening using PSA does not decrease mortality. Ann Intern Med 2011;154:JC1–2. [PMID: 21242356]
- Gøtzsche PC, Nielsen M. Screening for breast cancer with mammography. Cochrane Database Syst Rev 2011;(1):CD001877. [PMID: 21249649]
- Towler B, Irwig L, Glasziou P, Kewenter J, Weller D, Silagy C. A systematic review of the effects of screening for colorectal cancer using the faecal occult blood test, hemoccult. BMJ 1998;317:559–65. [PMID: 9721111]