CME

Current status of PSA screening

Early detection of prostate cancer

Tom Pickles, MD, FRCPC, MRCP(UK)

This article has been peer reviewed. Cet article a fait l’objet d’une évaluation externe. Can Fam Physician 2004;50:57-63.

Dr Pickles is Chair of the Genito-Urinary Tumour Group, a radiation oncologist at the British Columbia Cancer Agency, and a Clinical Associate Professor at the University of British Columbia in Vancouver.

Incidence of prostate cancer exceeds that of any other cancer; about 18 000 cases are expected this year in Canada.1 The disease accounts for more than 4000 deaths annually, being second only to lung cancer in terms of cancer mortality, and is also an important cause of morbidity. Prostate-specific antigen (PSA), a tumour marker introduced in the late 1980s, is one of the best oncology markers available. It has been used extensively to screen normal healthy men for the disease.

About 13% of Canadian men aged 50 to 59 and 24% of those aged 60 or older have had PSA tests2; screening has been the most common reason for requests for PSA testing among those with no established diagnosis of cancer.3 Screening is controversial, however, because it has not been shown to reduce mortality (unlike mammography for breast cancer, for example) and because the effectiveness of interventions for prostate cancer has been questioned. Family physicians are usually the first ones approached with requests for PSA testing. As use of PSA testing is controversial, they need to be aware of issues that argue for and against testing. This article reviews current evidence for routine use of PSA screening and suggests an approach to use in discussing testing with patients.

Wilson and Jugner4 defined criteria for effective screening tests in 1968. Accurate tests are a prerequisite, as are efficacy and acceptability of treatment, cost-effectiveness, and resource availability. Emerging evidence supports the efficacy of curative therapeutic options, such as surgery or radiation therapy, for prostate cancer.5 Many more men will die with prostate cancer, however, than from it. A 50-year-old man with a life expectancy of 25 years has a 42% lifetime risk of developing microscopic prostate cancer, a 9.5% chance of having clinical cancer, and a 2.9% chance of dying of it. Treatments are associated with substantial morbidity and reduced quality of life. Initial enthusiasm for widespread testing 10 years ago was replaced with caution in the later 1990s, but as new evidence emerges, rates of PSA screening appear to be increasing once again.

Literature search

MEDLINE was searched using the terms “prostate neoplasms,” “prostate-specific antigen,” and “diagnosis or mass screening.” From a total of 3619 references, 139 articles reported on controlled (37) or randomized (102) trials. Uncontrolled trials were excluded because they are particularly open to bias in patient selection. Cross-matching of randomized trials revealed that there were, in fact, only three trials, and only one of them has reported results. The other articles either referenced these trials or discussed prostate screening in a different context. From the many controlled trials, I was necessarily selective; interested readers are referred to some of the larger reviews cited.

Natural history of prostate cancer

Prostate cancer represents a spectrum of malignancy. In its earliest forms, it appears to arise from prostate intraepithelial neoplasia and is of low grade. Autopsy series show that a high proportion of healthy men have such early tumours. A 40- to 50-year-old man has about a 34% chance of having early occult prostate cancer,6 yet only one in eight men will be diagnosed with cancer during his lifetime. The chance of a man younger than 50 being diagnosed clinically is <0.3%.1

A placebo-controlled study of finasteride chemo­prevention in men older than 55 (with normal results of PSA and digital rectal examination) confirmed a high incidence of occult prostate cancer. Routine sextant biopsies were carried out for 4692 men in the placebo arm; 24% were positive for prostate cancer.7 Incidence rapidly increases with age. Most clinically apparent cases are seen in men in their 60s and 70s. More aggressive cancers are associated with higher PSA levels, Gleason grade, and clinical stage. Distant spread to bones and nodes is the common pattern of dissemination that ultimately leads to death. Hormone therapy with androgen ablation is typically effective for only 2 to 3 years after metastases are diagnosed.

The challenge for prostate screening programs is, therefore, to pick up clinically significant cancers at an age and stage when intervention will be successful, and to avoid overdiagnosis of occult low-grade cancers, particularly in elderly people, who will likely die of something else or old age long before the tumour progresses.

Unsanctioned PSA screening started in Canada in about 1990 and was accompanied by a rapid increase in incidence of prostate cancer. This evened out a few years later when diagnosed patients were taken out of the “population pool” and incidence fell (Figure 1). At the same time, but not necessarily causally, there was a dramatic reduction in incidence of metastatic cancer and advanced tumours. In 2000, more than 80% of patients referred to the British Columbia Cancer Agency had “curable” organ-confined prostate cancer, compared with less than 50% in 1988.8

Figure 1. Age-standardized rates in British Columbia: A) Incidence of prostate cancer normalized to 1980 (similar rates are reported by other provinces). Broken line indicates the underlying trend. B) Mortality rates from prostate cancer during the same period.

A

Data from: BC cancer registry.

B

Data from: BC cancer registry.

Mortality from prostate cancer has been falling since it peaked in about 1990. In British Columbia, the death rate has fallen 15% among all men and 19% among those younger than 65. While it is tempting to ascribe these results to PSA screening, which was introduced at about the same time, it is unlikely that PSA testing caused the mortality reduction. Although PSA testing can detect cancer about 5½ years earlier than it can be detected clinically,9 and even if subsequent localized treatment (such as radical prostatectomy or radiotherapy) had failed, it is unlikely that metastatic disease would develop for another 4 to 6 years. Hormone therapy would be expected on average to work for 2 to 3 years, giving a minimum likely time from PSA-based diagnosis to death of 10 to 15 years.10

Accuracy of the test

There is little doubt that PSA can detect prostate cancer. A study of 22 071 healthy men aged 40 to 82 years who had serum stored at baseline showed that 366 developed prostate cancer during the next 10 years.9 When PSA was measured in the stored samples, 46% of the cancers were detected. For diagnosis within 5 years of sample date, sensitivity rose to 65% and specificity was an impressive 91%. Overall, 87% of aggressive prostate cancers were detected within the first 4 years, compared with only 53% of nonaggressive cancers. Use of age-adjusted rates is important, because benign prostatic hypertrophy, which is increasingly common as men age, also causes PSA levels to rise: 2.5 ng/mL is abnormally high for a 49-year-old; 6 ng/mL is normal for an 80-year-old.

The free PSA ratio, a refinement of the PSA test, can help distinguish benign from malignant sources of PSA. Because PSA binds more avidly to a1-antichymotrypsin when it is from a malignant source, a low PSA ratio suggests malignancy. This is most useful when PSA falls into the “gray” zone, 4 to 10 ng/mL. Normal age-adjusted reference ranges should be part of laboratory reports. If they are not reported, they should be obtained from the laboratory. A possible drawback of being tested is the low positive predictive value of the test (25% to 35%, meaning that 65% to 75% of men testing positive turn out not to have cancer after further investigations) and the anxiety this generates.

Ongoing randomized trials

To date, there are no mature randomized studies, but two large studies are ongoing. The American Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial11 completed recruiting 154 000 men and women in July 2001. Mortality results are expected in 2015. In Europe, more than 150 000 men have been randomized in the European Randomised Study of Screening for Prostate Cancer (ERSPC).10 While mortality results will not be available for another 5 to 10 years, early reports of other end points have been published. The most promising is that incidence of metastatic cancer in the screening arm is only one tenth of that in the control arm12 (0.6% versus 6.7%) and that screen-diagnosed cancers are at an early stage. Both these findings should translate into a survival advantage with sufficient follow up.13

In Canada, a so-called randomized trial was published in 1999.14 In this population-based study, men were identified from electoral rolls and “randomized” to PSA testing in the experimental arm, but not informed they were in a study in the control arm. The design of the study, the method of analysis, and the fact that only 23% of those approached participated led to serious criticism of the results.15 In particular, men in the control arm were exposed to risk of prostate cancer for 3 years longer than those in the screening arm. An intention-to-treat analysis actually suggested an increased, rather than reduced, death rate from prostate cancer.16

Many reviews of PSA screening have been published, but no complete Cochrane systematic review is available. A systematic review published in the United Kingdom carries a balanced summary of the literature to 1996.17 It recommends PSA screening be discouraged.

Observational studies

A “natural experiment” from the United States describes different PSA screening rates, incidence of cancer, and use of curative treatments in Seattle and Connecticut from 1987 to 1990, and follows outcomes during the next decade.18 The cancer detection rate in the Seattle area was five times that in Connecticut; the cumulative incidence of diagnosed prostate cancer was 93% greater; and use of radical prostatectomy and radiation was 5.9- and 2.3-fold higher. The mortality rate over 11 years of follow up, however, was similar in the two regions; with Seattle having slightly higher mortality. Although not a randomized comparison, this study argues strongly that increased diagnosis of prostate cancer by PSA testing and subsequent treatment do not affect mortality over the following decade. It does not exclude subsequent benefit, however. The results also suggest that many of the cancers detected and treated were clinically insignificant.

Epidemiologic studies of worldwide mortality in the PSA era show that mortality rates are now lower in several countries.19 In some countries, the decline began before PSA testing (eg, in Italy in 1988), and in others, it began despite lack of interest in PSA screening (eg, the United Kingdom). Potential explanations include increased (and earlier) use of androgen ablation for metastatic cancers, attribution bias related to death certification, increased use of curative treatment, and a change from high- to lower-fat diets.

In Canada, two similarly designed studies have compared incidence of prostate cancer and mortality in patients in Quebec20 and British Columbia.21 Neither study shows a decline in mortality attributable to PSA within 10 years of commencement of community screening. A study of the Tyrol region in Austria, where PSA testing was made freely available in 1993, showed reductions in mortality throughout Austria, but greater reductions in the Tyrol and adjacent regions.22,23 Whether this reduction resulted from the screening program is uncertain.

National and international recommendations

There is no clear consensus on PSA screening. Some groups, such as the American Urological Association, are in favour; most others, such as the Canadian Cancer Society, are not. There is general consensus that men should be made aware of PSA as a screening test and should make an informed decision about whether to have it. The United States Preventive Services Task Force24 concluded that evidence is insufficient to recommend for or against routine screening for prostate cancer using the PSA test or digital rectal examination. In the United Kingdom, perhaps due to a concerted media campaign, the National Health Service has adopted an “on-demand” PSA screening program.25

What patients want

Whatever the controversy among medical professionals, patients want PSA screening.26 This is unsurprising, given that physicians usually suggest that patients present early to examine for lumps, undergo mammography, and so on. Patients think PSA screening offers the benefits of early diagnosis, of taking responsibility for their own health, of avoiding regret if cancer is subsequently diagnosed, of peace of mind, and of the right to information and access to the test.26 Although a physician’s views from a public health perspective should agree with his or her views toward an individual patient, the wishes and views of that patient should not be ignored.

Suggested approach to screening

Widespread enthusiasm for PSA screening has been replaced with a more cautious, individualized approach. If a man wants to minimize risk of prostate cancer and maximize his chance of living as long as possible, PSA screening might be appropriate. If a man wants to maximize his quality of life, minimize his risk of complications (such as impotence and incontinence), and undergo only medical tests that we know to be beneficial, however, PSA screening is unlikely to be of value to him.27

Patients should be informed that PSA testing can lead to earlier diagnosis and that it has both benefits and drawbacks. Useful minimum information and handouts have been published.28 Resources for this and many other cancer topics for physicians and patients can be found on the US National Cancer Institute website (http://cancer.gov/). Background documents on prostate screening in the UK for patients and physicians can be found on the National Health Services website (http://www. cancerscreening.nhs.uk/prostate/index.html). A Canadian patient information sheet can be downloaded from the BC Cancer Agency website (www.bccancer.bc.ca/HPI/CancerManagement Guidelines/Genitourinary/Prostate/PSAScreening/Recommendations.htm).

The natural course of prostate cancer is long, so patients should be screened only if they have a 10-year life expectancy and are at appreciable risk of cancer. Most authorities that recommend screening include only men between 50 and 70 years, but there could be a good case for extending this to 75 years, as life expectancy at 75 is now 10 years. Optimum screening intervals could depend on PSA velocity and absolute level. Suggested intervals are at age 40 and then 45 to establish a baseline, and then not again until 50. After that, men should be tested every 2 years unless there has been a substantial change in results or levels are raised.29 If PSA is <1 ng/mL, however, it need be checked only every 5 years; if 1 to 2 ng/mL, every 2 years. Such an approach has been predicted to reduce the number of PSA tests by 55% and save $500 to $1000 million annually in the United States.30

Current Canadian recommendations vary by province. Saskatchewan, Nova Scotia, and Quebec have funded PSA programs, but most provinces do not. Whether patients pay for tests or not, the approach should be the same: physicians inform patients of the test, discuss advantages and disadvantages, and take appropriate action (usually referral to a urologist) when abnormal results are detected. Remember that normal ranges are age-specific and that day-to-day variation of up to 25% is common. Unless abnormal results normalize on repeat testing, additional investigation is warranted.

Conclusion

Incidence of prostate cancer increased dramatically when unofficial PSA screening was introduced about 1990. Mortality began declining at the same time, but the decline is not thought to be a consequence of PSA screening. Early results from completed, but unreported, trials of PSA screening indicate that screened patients’ cancers are detected at an earlier stage than unscreened patients’ cancers. Whether this will translate into a survival advantage is uncertain, and we must await results of ongoing studies, which will not be available for another 5 to 10 years. In the meantime, there is consensus that men should be informed of the availability of the test and of its advantages and disadvantages and should make their own decisions about whether to be tested.

Competing interests

None declared

Correspondence to: Dr Tom Pickles, BC Cancer Agency, 600 West 10th Ave, Vancouver, BC V5Z 4E6; telephone (604) 877-6000, extension 2665; fax (604) 708-2101

References

1. Canadian Cancer Society. Canadian cancer statistics, 2002. Toronto, Ont: National Cancer Institute of Canada; 2002. Available at: http://www.cancer.ca. Accessed 2003 October 30.
2. Mercer SL, Goel V, Levy IG, Ashbury FD, Iverson DC, Iscoe NA. Prostate cancer screening in the midst of controversy: Canadian men’s knowledge, beliefs, utilization, and future intentions. Can J Public Health 1997;88(5):327-32.
3. Bunting PS, Goel V, Williams JI, Iscoe NA. Prostate-specific antigen testing in Ontario: reasons for testing patients without diagnosed prostate cancer. CMAJ 1999;160(1):70-5.
4. Wilson J, Jugner G. Principles and practice of screening for disease. Bol Oficina Sanit Panan 1968;65(4):281-393.
5. Holmberg L, Bill-Axelson A, Helgesen F, Salo JO, Folmerz P, Haggman M, et al. A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer. N Engl J Med 2002;347(11):781-9.
6. Sakr WA, Haas GP, Cassin BF, Pontes JE, Crissman JD. The frequency of carcinoma and intraepithelial neoplasia of the prostate in young male patients. J Urol 1993;150(2 Pt 1):379-85.
7. Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG, et al. The influence of finasteride on the development of prostate cancer. N Engl J Med 2003;349(3):215-24.
8. Pickles T, Coldman A, Phillips N. The changing face of prostate cancer in British Columbia 1988-2000. Can J Urol 2002;9(3):1551-7.
9. Gann PH, Hennekens CH, Stampfer MJ. A prospective evaluation of plasma prostate-specific antigen for detection of prostatic cancer. JAMA 1995;273(4):289-94.
10. Auvinen A, Rietbergen JB, Denis LJ, Schroder FH, Prorok PC. Prospective evaluation plan for randomised trials of prostate cancer screening. The International Prostate Cancer Screening Trial Evaluation Group. J Med Screen 1996;3(2):97-104.
11. National Cancer Institute. Prostate, lung, colorectal and ovarian (PLCO) cancer screening trial. Bethesda, Md: National Cancer Institute; 2003. Available at: http://www3.cancer.gov/prevention/plco/. Accessed 2003 October 30.
12. Van der Cruijsen-Koeter I, Roobol M, Damhuis R, De Koning H, Schroeder F. Comparison of screen detected and clinically diagnosed prostate cancer in the European randomized study of screening for prostate cancer. American Urological Association Annual Meeting in Anaheim, Cal. J Urol 2002;4(Suppl):102. Abstract 407.
13. Maattanen L, Auvinen A, Stenman UH, Rannikko S, Tammela T, Aro J, et al. European randomized study of prostate cancer screening: first-year results of the Finnish trial. Br J Cancer 1999;79(7-8):1210-4.
14. Labrie F, Candas B, Dupont A, Cusan L, Gomez JL, Suburu RE, et al. Screening decreases prostate cancer death: first analysis of the 1988 Quebec prospective randomized controlled trial [see comments]. Prostate 1999;38(2):83-91.
15. Boer R, Schroder FH. Quebec randomized controlled trial on prostate cancer screening shows no evidence for mortality reduction. Prostate 1999;40(2):130-4.
16. Alexander FE, Prescott RJ. Reply to Labrie et al. Results of the mortality analysis of the Quebec randomized controlled trial (RCT). Prostate 1999;40(2):135-7.
17. NHS Centre for Reviews and Dissemination. Screening for prostate cancer. York, Engl: NHS Centre for Reviews and Dissemination; 1997. Available at: http://www.york.ac.uk/inst/crd/em22a. Accessed 2003 October 30.
18. Lu-Yao G, Albertsen PC, Stanford JL, Stukel TA, Walker-Corkery ES, Barry MJ. Natural experiment examining impact of aggressive screening and treatment on prostate cancer mortality in two fixed cohorts from Seattle area and Connecticut. BMJ 2002;325(7367):740-6.
19. Oliver SE, May MT, Gunnell D. International trends in prostate-cancer mortality in the “PSA Era.” Int J Cancer 2001;92(6):893-8.
20. Perron L, Moore L, Bairati I, Bernard PM, Meyer F. PSA screening and prostate cancer mortality. CMAJ 2002;166(5):586-91.
21. Coldman A, Phillips N, Pickles T. Trends in prostate cancer incidence and mortality: an analysis of mortality change by screening intensity. CMAJ 2003;168(1):31-5.
22. Bartsch G, Horninger W, Klocker H, Reissigl A, Oberaigner W, Schonitzer D, et al. Prostate cancer mortality after introduction of prostate-specific antigen mass screening in the Federal State of Tyrol, Austria. Urology 2001;58(3):417-24.
23. Boyle P, Horninger W, Oberainger W, Schonitzer D, Severi G, Robertson C, et al. Prostate cancer mortality in Tyrol, Austria. Update to year 2000. American Urological Association Annual Meeting in Orlando, Fla. J Urol 2002;4(Suppl):102. Abst 406.
24. Agency for Healthcare Research and Quality. Screening for prostate cancer. Rockville, Md: Agency for Healthcare Research and Quality; 2000. Available at: http://www.ahrq.gov/clinic/uspstf/uspsprca.htm. Accessed 2003 Dec 8.
25. NHS Cancer Screening Programmes. Prostate cancer risk management. Sheffield, UK: National Health Service; 2002. Available at: http://www.cancerscreening.nhs.uk/prostate. Accessed 2003 October 30.
26. Chapple A, Ziebland S, Shepperd S, Miller R, Herxheimer A, McPherson A. Why men with prostate cancer want wider access to prostate specific antigen testing: qualitative study. BMJ 2002;325(7367):737-41.
27. Chodak GW. Screening for prostate cancer. Eur Urol 1993;24(Suppl 2):3-5.
28. Chan EC, Sulmasy DP. What should men know about prostate-specific antigen screening before giving informed consent? Am J Med 1998;105(4):266-74.
29. Ross KS, Carter HB, Pearson JD, Guess HA. Comparative efficiency of prostate-specific antigen screening strategies for prostate cancer detection. JAMA 2000;284(11):1399-405.
30. Crawford D, Chia D, Andriole GL, Reding D, Gelmann E, Gohagan P, et al. PSA testing interval, reduction in screening intervals: Data from the Prostate, Lung, Colorectal and Ovarian Cancer (PLCO) screening trial. American Urological Association Annual Meeting in Orlando, Fla. J Urol 2002;4(Suppl):99. Abst 397.