Correspondence
Screening for Prostate Cancer
N Engl J Med 1996; 334:666-668March 7, 1996
- Article
To the Editor:
Woolf's review (Nov. 23 issue)1 of screening for cancer with prostate-specific antigen (PSA) omits important findings. He states that PSA testing could cost billions of dollars but does not mention that treating advanced prostate cancer is far more costly than treating early disease.
Woolf questions the importance of the spread of cancer beyond the prostate gland because extraprostatic extension has been reported in autopsy studies; however, the spread of cancer is associated with bad outcomes.2 Findings in autopsy series are not representative of those in the general population. It cannot be assumed that a tumor with adverse histologic features in a man killed in an accident would be harmless in a man with a longer life expectancy, who would have an increased probability of cancer progression.
Woolf does not mention that the Food and Drug Administration has approved one PSA assay, in conjunction with digital rectal examination, for the early detection of cancer or that the positive predictive value of PSA testing exceeds that of mammography for breast-cancer screening.
Woolf cites studies of highly selected patients that question the value of treating prostate cancer.3-6 These studies give little consideration to the need for treatment to prevent suffering from progressive disease, which should be measured along with metastatic rates and deaths from cancer.
Woolf overstates the complications of radical prostatectomy, citing outdated results obtained before or during the learning curve of contemporary radical prostatectomy. These outcomes do not accurately represent what is available to many patients today. Most of the complications can be effectively treated. The institutional variation in outcomes is not a justification for not screening for early detection of cancer. Early detection remains a goal in the approach to most cancers, even though the benefits have not been formally established for many cancers, including prostate cancer.
Before PSA testing was available, prostate cancer was usually detected too late for cure. Consequently, the mortality and morbidity rates of prostate cancer have continued to rise. Since there are no established modifiable risk factors for prostate cancer, the only practical strategy for improving outcomes is early detection and appropriate treatment. If this strategy is effective, outcomes should begin to improve within the next five years.
A complete assessment of PSA screening, with equal scrutiny of the assumptions and quality of data on both sides, suggests that in appropriately selected men, screening will allow the curative treatment of presymptomatic cancers that otherwise would result in morbidity and mortality.
(My research is supported in part by a grant from Hybritech Inc., San Diego, California.)
6 ReferencesWilliam J. Catalona, M.D.
Washington University School of Medicine, St. Louis, MO 631051
Woolf SH . Screening for prostate cancer with prostate-specific antigen -- an examination of the evidence. N Engl J Med 1995;333:1401-1405
Full Text | Web of Science | Medline2
Wingo PA ,Tong T ,Bolden S . Cancer statistics, 1995. CA Cancer J Clin 1995;45:8-30
CrossRef | Web of Science | Medline3
Johansson JE . Expectant management of early stage prostatic cancer: Swedish experience. J Urol 1994;152:1753-1756
Web of Science | Medline4
Chodak GW ,Thisted RA ,Gerber GS , et al. Results of conservative management of clinically localized prostate cancer. N Engl J Med 1994;330:242-248
Full Text | Web of Science | Medline5
Krahn MD ,Mahoney JE ,Eckman MH ,Trachtenberg J ,Pauker SG ,Detsky AS . Screening for prostate cancer: a decision analytic view. JAMA 1994;272:773-780
CrossRef | Web of Science | Medline6
Fleming C ,Wasson JH ,Albertsen PC ,Barry MJ ,Wennberg JE . A decision analysis of alternative treatment strategies for clinically localized prostate cancer. JAMA 1993;269:2650-2658
CrossRef | Web of Science | Medline
To the Editor:
Woolf's article on PSA screening is well researched and well reasoned. However, his proposal to have physicians fully explain the PSA-screening controversy to patients, assess their preferences, and then let them decide is unrealistic. As a generalist physician who is supposed to put screening recommendations into practice, I am painfully aware of how many issues I am expected to address during a “routine” history taking and physical examination. A partial list includes screening for depression, alcohol abuse, tobacco use, incontinence, sexual dysfunction, physical or emotional abuse, and the desirability of advanced directives. If any of these are issues, I would then need to devote more time to addressing them.
In clinical practice, time is crucial. Woolf's proposal to explain the PSA-screening controversy to patients takes time, and PSA screening is not the only gray area in terms of cancer detection. Shouldn't I also discuss with patients the pros and cons of undergoing mammography before the age of 50, screening sigmoidoscopy, and even testing for occult blood? Neither fee-for-service nor capitated insurance seems eager to reimburse me for implementing each of these recommendations. So I am left with a long list of recommendations and precious little time. The medical community must consider generalists when making recommendations that they are supposed to implement.
William A. Hensel, M.D.
Memorial Hospital, Greensboro, NC 27401To the Editor:
I think we are all confused about what to do in regard to screening for prostate cancer. Our patients recognize our confusion in our mixed messages on screening and treatment recommendations. Dr. Woolf advises that “physicians should neither recommend nor discourage PSA testing without, first, ensuring that patients have complete information about potential benefits and risks, and second, determining their personal preferences.” Unfortunately, this approach does not help me in terms of giving advice about the PSA test to a healthy 52-year-old man with a normal digital rectal examination.
Currently, no one can differentiate an early prostate cancer that may be a latent cancer from one that is aggressive and clinically important. Until such a thing is possible, I feel a medical and ethical obligation to search for early prostate cancers. There seems to be quite a difference between telling a man who is 80 years old that he has prostate cancer that will probably be unimportant in his lifetime and telling a man who is 52 that he has prostate cancer and that it may or may not be important in his lifetime.
Until someone can clearly tell me the difference between the two situations, I will follow the American Cancer Society's recommendation to perform a digital rectal examination of the prostate and PSA screening in every man over the age of 50. Then the patient and I can decide when we can stop, since, as Dr. Woolf suggests, “men with a life expectancy of less than 10 years should be advised that screening and treatment are unlikely to be helpful.”
Randy Stevens, M.D.
Union Hospital, Terre Haute, IN 47807To the Editor:
Dr. Woolf fails to identify one of the most serious harms resulting from screening for prostate cancer. When we detect prostate cancer early, but it ultimately does nothing to change a patient's outcome, we harm that patient. Screening, by definition, involves healthy people. A man undergoing PSA screening cannot have symptom relief as a result of screening. The discovery of prostate cancer would mean he is no longer healthy. If his ultimate outcome is not improved by changing his status from healthy to sick, he has been harmed. He has had time spent as a healthy man taken away from him. Since three of every five prostate cancers detected in a screening program are not organ-confined1 and are therefore not curable with current therapies, these men are harmed by PSA screening. If they suffer adverse effects from the diagnostic workup or treatment of their prostate cancer, the harm is multiplied. . . .
Hence, it is certain that at least some men are harmed by PSA screening. Applying the first ethical principle of medicine — primum non nocere (first, do no harm) — leads one to the conclusion that PSA screening should not be generally advised before benefit is proved. It is appropriate for a physician opposed to PSA screening to avoid the topic when patients do not request the test. Why should we even indirectly encourage testing known to harm some and not proved to offer a greater likelihood of benefit than harm? Outside of randomized trials of PSA screening, such discussion may well do more harm than good.
1 ReferencesBrian Budenholzer, M.D.
Group Health Northwest, Spokane, WA 99204-02041
Catalona WJ ,Richie JP ,Ahmann FR , et al. Comparison of digital rectal examination and serum prostate specific antigen in the early detection of prostate cancer: results of a multicenter clinical trial of 6,630 men. J Urol 1994;151:1283-1290
Web of Science | Medline
Author/Editor Response
Dr. Woolf replies:
To the Editor: Several of Dr. Catalona's arguments miss the point. That PSA screening has a higher positive predictive value than mammography says little about its effectiveness. The positive predictive value of a test is proportional to the prevalence of the disease being tested for, which is higher for prostate cancer than for breast cancer.1 The higher positive predictive value may simply mean that prostate cancer is more common in men than is breast cancer in women, not that PSA testing is more accurate than mammography. Moreover, even if PSA screening is more likely to detect a true positive result, the benefit to the patient is only speculative. Clinical trials demonstrate that mammography lowers mortality, but no such evidence exists for PSA screening. Similarly, approval by the Food and Drug Administration of the PSA assay simply means that it can detect prostate cancer.2 Tests need not improve outcomes to win such approval.
I agree that postmortem evidence of extracapsular penetration or undifferentiated cells does not eliminate the increased risk of progression associated with these findings. Autopsy data remind us, however, that adverse histopathological findings are not a guarantee of symptomatic disease. Autopsy studies of men in their 70s and 80s show that men with these findings may live long lives and die of other diseases, free of prostate-related symptoms. Therefore, the histologic findings in tumors detected by PSA screening cannot be cited as “evidence of cure.”
Although Dr. Catalona is correct that isolating mortality from morbidity results in an underestimate of potential benefits, the morbidity of screening and treatment must also be considered in judging harms. The data in Table 2 of my article, which he claims are “outdated,” were taken largely from studies published in 1993 to 1995. In my article I acknowledge that PSA screening may be cost effective, but we currently lack the data to confirm that it is. Cost effectiveness cannot be determined by simply comparing treatment costs for early and advanced disease.
As a family physician, I share Dr. Hensel's concern about the lack of time for PSA-assay counseling, but patients have too much at stake for physicians to withhold information on these grounds. Counseling of this intensity is less necessary for the other tests listed by Dr. Hensel, because in those cases benefit is more certain and is less dependent on patient preferences. With practice, PSA-assay counseling requires only a few minutes, especially if the patient is also given educational materials. Primary care physicians spend far more time on less useful activities.3 A few minutes of shared decision making can spare patients unwise choices and reduce costs.4
Dr. Budenholzer believes that physicians opposed to PSA screening need not tell patients about the test. Certainly, physicians should not make all tests available to patients, especially those that are ineffective or harmful. In fairness, however, this cannot be said for PSA screening; without better data, no one truly knows whether PSA screening helps or harms patients. With such uncertainty about a leading cause of death, honest disclosure of the options seems appropriate. In settings in which a uniform policy to test or not test is to be implemented, recommendations against routine screening5 should apply.
5 ReferencesSteven H. Woolf, M.D., M.P.H.
Fairfax Family Practice Center, Fairfax, VA 220331
Ries LAG, Miller BA, Hankey BF, Kosary CL, Harras A, Edwards BK, eds. SEER cancer statistics review, 1973–1991: tables and graphs. Bethesda, Md.: National Cancer Institute, 1994. (NIH publication no. 94-2789.)
2
Advisory Panel Chairpersons Meeting of the Medical Devices Advisory Committee 3
Luckmann R ,Melville SK . Periodic health evaluation of adults: a survey of family physicians. J Fam Pract 1995;40:547-554
Web of Science | Medline4
Wagner EH ,Barrett P ,Barry MJ ,Barlow W ,Fowler FJ Jr . The effect of a shared decisionmaking program on rates of surgery for benign prostatic hyperplasia: pilot results. Med Care 1995;33:765-770
CrossRef | Web of Science | Medline5
Preventive Services Task Force. Guide to clinical preventive services: an assessment of the effectiveness of 169 interventions. Baltimore: Williams & Wilkins, 1995.
- Citing Articles (4)
Citing Articles
1
Judd W. Moul. (2005) A discussion of general prostate cancer screening versus targeted diagnosis in younger men. Current Prostate Reports 3:3, 99-111
CrossRef2
A. M. Kulikov. (2005) Genetically modified organisms and risks of their introduction. Russian Journal of Plant Physiology 52:1, 99-111
CrossRef3
Judd W. Moul. (2000) Screening for prostate cancer in african americans. Current Urology Reports 1:1, 57-64
CrossRef4
(2000) EDITORIAL COMMENT. The Journal of Urology 163:1, 148-149
CrossRef
