As every urologist is aware, cancer of the prostate is a major public health threat. The American Cancer Society estimates that 189,000 American men will have been diagnosed with prostate cancer and that 30,200 will have died of the disease in 2002.1 It is the most common cancer diagnosed in men and the second leading cause of death from cancer in the United States, exceeded only by lung cancer.
Until 1988, only a small proportion of new cases were detected early enough to benefit from curative therapy, and mortality had not declined during the previous 20 years.1 The development of serum prostate-specific antigen (PSA) testing, however, offered an opportunity to improve early detection. The apparent benefits of PSA compared with either a digital rectal examination or transrectal ultrasonography led to an enthusiasm to advise all asymptomatic men in the high-risk age group to undergo a PSA test.2 With so many men having or dying of prostate cancer, and with advanced disease still rarely curable, why would anyone argue against undergoing a PSA test?
Unfortunately, finding prostate cancer when it is localized is necessary but insufficient proof that routine screening should be recommended. The most important requirements are that advanced disease at diagnosis and disease-specific mortality must also decline. Proponents of PSA screening argue that both have occurred. Metastatic disease has been almost eliminated, accounting for less than 6% of new prostate cancer cases.3 In addition, Surveillance, Epidemiology, and End Results data from the National Cancer Institute showed a 17.6% fall in prostate cancer mortality in 1993,4 which continued for the next 3 years, several years after PSA testing began.
On the other hand, critics of screening also raise several valid arguments. First, a reduction in mortality from PSA screening by 1993 would be highly improbable given the low mortality rate at 5 years after diagnosis with conservative therapy.5 A more feasible explanation for the change in mortality is the earlier use of hormonal therapy after increases in PSA levels before clinical symptoms appeared or metastases were detected. Furthermore, mortality has also decreased in regions of the world in which PSA screening has not been prevalent.6
Therefore, because the impact of routine PSA screening on mortality is still uncertain, several questions remain. First, is it reasonable to conclude that mortality will definitely be reduced with longer follow-up? Second, and most importantly, what message should be delivered to the public and how should this be accomplished? Reviewing data from screening for other diseases may provide an answer to the first question.
After years of global acceptance, there is now doubt over whether breast cancer screening confers as much benefit, if any, as once thought, despite an improvement in early diagnosis. This comes after investigators at the Nordic Cochrane Center controversially stated that “screening for breast cancer with mammography is unjustified.” In a systematic review of eight randomized trials of screening mammography, significant flaws were identified in six of the eight identified trials, with the remaining two trials showing no effect of screening on breast cancer mortality.7 Further inconsistencies were highlighted by Black et al.,8 who analyzed 12 screening trials for colon, breast, and lung cancer. They concluded that 3 of 7 influential mammography studies were unintentionally flawed by inaccurate determination of whether participants died of cancer or unrelated causes.8 Subsequently, the U.S. Preventive Services Task Force downgraded its assessment of the evidence of mammography from “good” to “fair” and lowered its recommendation from “A” (strongly recommended) to “B” (simply recommended).
The results of prospective studies of neuroblastoma screening in infants also have demonstrated that improving early diagnosis may not affect mortality. Investigators from Germany and Canada conducted clinical trials to determine whether screening infants for neuroblastoma, through a simple urine test, would reduce the mortality caused by the disease. The investigators found no difference in mortality between the screened and unscreened groups, but they did find substantial overdiagnosis in the screened cohort.9, 10 Although screening identified a higher proportion of localized cancers, it has been reported that at least 60% of neuroblastoma cases regress spontaneously, implying that many children are subjected to unnecessary and avoidable diagnostic procedures and therapy.11
Even though there is no evidence that prostate cancer spontaneously regresses, it is well recognized that the vast majority of men with histologic prostate cancer are not at risk of suffering from, or dying of, the disease. Currently, there is no guarantee that screening will detect only “life-threatening cancer” and miss those that are not. Despite differences in the biology of these three cancers, the examples with breast cancer and neuroblastoma demonstrate the potential error of prematurely concluding that finding a higher proportion of early-stage prostate cancers by PSA testing will definitely result in lowering mortality from this disease. Making such a conclusion may possibly result in significant overdiagnosis and unnecessary treatment and harm.
These results make answering the second question quite challenging: what to do while waiting for definitive evidence that screening lowers mortality? One solution is to create an informed consent form. Patients could read the form, ask questions as needed, and then sign it to have the PSA test performed.
Many advantages of informed consent have been reported. In a review of 153 published studies evaluating the effectiveness of interventions to improve patient compliance, Roter et al.12 found that better-informed patients were significantly more likely to comply with treatment. Furthermore, with proper counseling, patients have less anxiety and more satisfaction if they are allowed to take an active part in their cancer management.13, 14 Perhaps most importantly, however, it is possible to improve outcomes if patients make informed decisions regarding their own medical care.15, 16
Another advantage of using an informed consent form is the potential to avoid medicolegal issues. Some individuals have already filed a “failure to diagnose” claim against their physicians because they were diagnosed with an advanced prostate cancer and state that their doctor never discussed, performed, nor offered them a PSA test. Even though organizations such as the American Medical Association, American College of Physicians, and American Task Force for Preventative Disease have recommended against PSA testing of asymptomatic men,17 a consent form would provide added protection to both patient and physician should a cancer subsequently be detected.
Similarly, other men who specifically ask their physician for the PSA test would benefit from reading a consent form because they may have been led to believe that screening and subsequent treatment is simple, effective, and does not cause harm. Moreover, studies have shown that many individuals have misconceptions about the purpose of screening, accuracy of screening tests, and magnitude of benefit.18 Regarding breast cancer, for every 1000 women screened biennially throughout 12 years, one disease-specific death is avoided.7 It would be interesting to assess an individual’s choice for having a PSA test, assuming a biopsy rate of approximately 15% and a survival benefit of similar or smaller magnitude as that with breast cancer screening. A signed consent form would provide protection for men by ensuring they are informed about the PSA screening controversy, have realistic expectations about the probability of benefit and potential harm, and have a choice of whether to undergo the test.
Another way that a written consent form protects patients is by maintaining their autonomy should they decide they do not want to undergo tests that still have not shown clear benefit. In many physician offices and clinics in the United States, a routine PSA test is now included as part of a general checkup or insurance physical without it ever being discussed with the patient. This situation could have harmful ramifications for patients by potentially making them less insurable if the test is abnormal or creating unwanted anxiety knowing they may have cancer even if a biopsy proves negative. Surely every person has a right to participate in the decision to undergo medical intervention, particularly when the benefit is uncertain.
Many doctors might argue that obtaining informed consent would be too time-consuming, because it would restrict their ability to deal with other more pressing medical problems confronting their patients. However, is that really a valid reason to simply obtain the test without first discussing it or ignoring the test entirely? In fact, a consent form could provide the information to patients more rapidly than a physician and then the physician would only need a short period, if any, to answer any questions that arise.
If an informed consent form is the best approach until definitive proof of benefit is available, the last question is what should be included. The best approach would be to outline the positive and negative aspects of being tested and treated with an estimate of the probability of benefit and the probability of harm, both from the ensuing biopsy and possible treatment if cancer is detected. Without understanding these probabilities, patients may have difficulty assessing the trade-off they would face by undergoing or not undergoing the test. This information could be developed by a workshop of experts and could lead to a useful document throughout the United States.
We believe that notwithstanding the progress made in early detection, uncertainty remains about whether performing a PSA test on asymptomatic men in the high-risk age group will reduce the risk of dying of prostate cancer. Valid arguments can be made to recommend for or against this test. Until studies in progress determine whether a benefit exists, routine use of an informed consent form seems the only ethical and rational approach to deal with the problem.
FULL TEXT
