For Five Cancers, Screening Skews Risk Factors
January 02, 2018
Screening and diagnostic testing can affect the incidence of certain cancers and result in "self-fulfilling" risk factors associated with those cancers, according to a new essay by two major figures in the US cancer scene.
The article was published online January 1 in the Annals of Internal Medicine.
These various forms of "scrutiny" can also result in "misleading feedback loops" in cancer epidemiology, say H. Gilbert Welch, MD, MPH, professor of medicine, Dartmouth Institute for Health Policy & Clinical Practice in Hanover, New Hampshire, and Otis W. Brawley, MD, chief medical officer, American Cancer Society, Atlanta, Georgia.
For example, female sex is considered a prominent risk factor for thyroid cancer because women are three times more likely to be diagnosed with it than men in the United States. But women also are more likely to access health services and undergo related imaging tests. Importantly, thyroid cancer mortality is nearly identical between men and women, which suggests that sex as a risk factor is "more apparent than real," say the essayists.
Dr Welch and Dr Brawley say that cancer risk-factor epidemiology needs to change in order to make risk factors more "meaningful" and appropriate.
"When clinicians hear the word 'risk factor,' I'd like them to reflexively ask 'Risk factor for what?' In the case of these cancers at least, risk factors for cancer death may be very different that risk factors for cancer diagnosis," Dr Welch told Medscape Medical News in an email. "I worry this concern may be equally relevant for genetic risk factors."
Prostate cancer is the "poster child" of scrutiny-dependent cancer, say the essayists. The incidence of prostate cancer skyrocketed in the mid-1980s and early 1990s and then declined steadily in the following decades. "No known tumor biology or carcinogenic process can explain its rapid risk and quick decline," write Dr Welch and Dr Brawley.
Instead, the process of looking for prostate cancer, especially with the blood test for prostate-specific antigen (PSA), was driving the rise in incidence. When physicians agreed that PSA testing in men had significant limitations, use of the test plummeted (first in older men and then in all men).
Any type of cancer with a "substantial disease reservoir of indolent, subclinical forms" (such as prostate cancer) is a candidate to be a scrutiny-dependent cancer, say the essayists.
But the significance of this phenomenon does not stop here. Scrutiny can also affect risk factor assessment, they explain.
The essayists discuss family history as a risk factor for prostate cancer to show how scrutiny can warp its true impact.
They review a recent analysis (J Clin Oncol. 2016;34:4338-4344) of two clinical trials showing that family history is a "relatively weak risk factor" when all men undergo biopsy, with a relative risk of less than 1.5. But, among men exposed to the combination of contemporary screening and biopsy, the relative risk of family history is greater than 2.0.
The essayists reprint what prominent urologist Peter Albertsen, MD, from the University of Connecticut in Farmington, wrote about the likely cause of these differing risks in the editorial related to that analysis (J Clin Oncol. 2016;34:4310-4311).
"Men with a family history of prostate cancer are frequently encouraged to undergo PSA testing. This, in turn, leads to a higher biopsy rate, which, because of the large pool of subclinical disease, leads to higher rates of prostate cancer diagnosis in this group of men, thereby 'confirming' the original hypothesis [that family history in a risk factor]," wrote Dr Albertsen.
In short, family history is indeed a risk factor for prostate cancer. But that risk is "spuriously strengthened because men with a family history are exposed to greater scrutiny," write Dr Welch and Dr Brawley.
There may be other scrutiny-dependent cancers, said Dr Welch.
"It all depends on how hard we look for them and how big the reservoir of undetected disease is," he said. "I would have never included lung cancer on the list, for example, before we started looking for it using spiral CT [computed tomography]."
If it becomes commonplace, genetic testing may eventually give rise to other scrutiny-dependent cancers, the essayists warn. That's because of the potential for "aggressive" screening for people with positive test results and thus more cancer detection.
Risk factor epidemiology must shift away from diagnosis to "harder" outcomes that are more "directly related to the disease process," and death is the best candidate because it is least ambiguous, say Dr Welch and Dr Brawley.
"As cancer diagnosis becomes increasingly sensitive to observer scrutiny, we hope that those investigating the risk for this disease focus on risk factors for death from cancer and not cancer diagnosis," they conclude.
The authors have disclosed no relevant financial relationships.
SOURCE: Medscape, January 02, 2018. Ann Intern Med. Published online January 1, 2018.