In his book, How We Do Harm: A Doctor Breaks Ranks About Being Sick in America, Otis Brawley1 writes, “I believe that a man should know what we know, what we don’t know, and what we believe about prostate cancer. I have been concerned that many patients and physicians have confused what is believed with what is known.” I agree.
Common sense is what we believe. Does common sense trump science? Did the US Preventive Services Task Force (USPSTF) get it wrong? I don’t think so.
The USPSTF bases its recommendations on an explicit assessment of the science that informs us of the benefits and harms of a preventive service, and a judgment about the magnitude of net benefit.
So what do we know about the benefits of prostate cancer screening? When attempting to answer the question of whether an intervention is beneficial, there is a hierarchy of evidence, from most likely to be wrong to most likely to be right. Relying on our personal stories is the former; relying on well-conducted randomized trials is the latter.
In the multicenter Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial2 conducted in the United States, there was a nonsignificant increase in prostate-cancer mortality in the screening group, while the European Randomized Study of Screening for Prostate Cancer (ERSPC)3 trial showed a statistically significant absolute reduction of 0.10 prostate-cancer deaths per 1000 person-years after a median follow-up of 11 years. In the ERSPC trial, all-cause mortality was 19.1% in the screened group and 19.3% in the control group, a difference that was not significant. What we know is that after 10 years, even with aggressive treatment of 80% to 90% of screen-detected cancers, very few, if any, men will have lived longer because they were screened.
What don’t we know about the benefits? We don’t know whether following screened and nonscreened men for 15 or 20 years or longer will demonstrate a larger difference in mortality. Competing causes of mortality make it progressively less likely that men who are screened will actually live longer. The average age of death from prostate cancer is 80 years, and 70% of all deaths occur after age 75.4 Contrast those statistics to breast cancer, for which the average age of death is 68 years and 63% of all deaths occur before age 75.5
What do we believe about the benefits? Some certainly believe the trials must be wrong; common sense tells us that early detection and treatment must provide more benefit than what the evidence has shown. Common sense tells us that the decline in prostate cancer mortality over the past 2 decades must be due to screening, although the ERSPC results clearly show that neither the magnitude nor the timing of the decline can be attributed to screening.
What do we know—and not know—about the harms?
We know that much of the suffering from prostate cancer is a consequence of the diagnosis and management of the disease, rather than the disease itself. Complications of both diagnosis and treatment of prostate cancer are frequent and serious.6 We also know that many screen-detected cancers would never become apparent in a man’s lifetime without screening.
We don’t know the precise magnitude of overdiagnosis, although all estimates suggest it is substantial. In the ERSPC trial, 9.6% of the screened group received a prostate cancer diagnosis, vs 6.0% of the control group—a 60% increase in the rate of diagnosis. The recently published long-term results from the Prostate Cancer Prevention Trial7 are enlightening. Finasteride reduced the incidence of screen-detected cancers by 30%, with no impact on all-cause mortality at 18 years. If those screen-detected cancers had been a significant threat to health, then after 18 years we would have expected some mortality benefit from finasteride.
What do we believe about the harms of screening?
We believe that by being more conservative about who gets treated, we shift the balance of benefits and harms of screening. There is no question that reducing the burden of overdiagnosis and overtreatment would provide a welcome reduction in the harms. But can we do it?8
In the United States 90% of men found to have prostate cancer are treated (including about 75% of men with low-risk cancers).6 And although we hope to be able to reduce harms without changing benefits, we do not know what impact more conservative management of screen-detected cancers would have on the already small effect of screening on prostate cancer mortality.