In addition to today’s resource, please see and review Final Recomendation Statement
Prostate Cancer: Screening
May 08, 2018 from U.S. Preventive Services Task Force.
In this post, I link to and excerpt from Reconsidering Prostate Cancer Mortality
— The Future of PSA Screening [PubMed Abstract] [Full-Text HTML]. H. Gilbert Welch, M.D., M.P.H., and Peter C. Albertsen, M.D. N Engl J Med. 2020 Apr 16;382(16):1557-1563.
All that follows is from the above resource.
From its peak in the early 1990s, U.S. mortality due to prostate cancer has decreased from 39 per 100,000 men to 19 per 100,000 men — essentially by half. Although everyone agrees
that this reduction is good news, there is considerable disagreement about why it happened. The controversy has profound implications for the future of prostate-specific antigen (PSA) screening.Why Not to Screen
So why not advocate for PSA screening? Unfortunately, the decrease in prostate cancer mortality has been achieved at enormous human cost: incidence soared to over 200 per 100,000, andmore than a million men were diagnosed with
a clinically insignificant “cancer” and received treatment for pathologic findings not destined to cause symptoms or death.18 PSA screening represents a textbook case of overdiagnosis and
overtreatment in medical care. On that basis alone, we believe that the U.S. Preventive Services Task Force was right not to recommend it.Furthermore, decreasing prostate cancer mortality may be a misleading metric in evaluating PSA screening. A reduction in cancer-specific mortality does not reliably translate into increased longevity, which requires a reduction in all-cause mortality. This problem was highlighted in the 30-year follow-up of arguably the most influential randomized trial of colorectal cancer screening: the Minnesota Colon Cancer
Control Study.19 The curves illustrating cumulative colorectal cancer mortality showed a clear advantage for annual fecal occult blood screening, which resulted in a 33% relative reduction (or a 1% absolute reduction) in cancer-specific
mortality. Nevertheless, there was no change in all-cause mortality — the curves illustrating cumulative all-cause mortality were perfectly superimposed over the entire 30-year period (see figure at NEJM.org).In other words, screening may more easily change the distribution of causes of death (trading off one cause for another) than extend life (as implied by promises to “save lives”). This issue is particularly relevant to PSA screening,
since the median age at death due to prostate cancer is so high — 80 years (as compared with 72 for lung cancer and 68 for breast cancer).20 For the elderly, the combination of a high burden of competing risks for death and high rates of intervention-related complications conspires to limit any reduction in all-cause mortality offered by screening.21On balance, we would continue to argue against contemporary PSA screening, particularly in light of our volume-driven health care system. But we acknowledge that our position reflects a value judgment. A few people receive a substantial benefit (avoiding death from prostate cancer), while many more are exposed to needless biopsies, operations, and another
source of financial stress. We have no common metric for comparing these benefits and harms— they are like apples and oranges. Thus, there is no calculus or decision model that can produce a single “right” answer.If You’re Going to Screen Anyway
We believe that providers who arrive at the opposite value judgment and support PSA screening must offer patients a better deal by protecting them from overdiagnosis and overtreatment. Doing so requires an incidence target — a “not
to exceed” incidence benchmark. Currently, prostate cancer incidence is a bout where it was in1975: roughly 100 per 100,000. That should bethe incidence target.Meeting this target while screening will require a higher test threshold for biopsy.22 The conventional PSA threshold of 4 ng per milliliter was chosen to maximize cancer detection. We
believe that this threshold is too low, identifying far too many men with low-grade, clinically insignificant disease. A higher biopsy threshold would not only reduce overdiagnosis, it would
also reduce the number of biopsies and their associated harms.
