Thomas Farrington’s father and both of his grandfathers died of prostate cancer, but for many years, Farrington wasn’t concerned about his own risk. Then, in 2000, he was diagnosed with moderately aggressive stage I prostate cancer at age 57. An elevated score on a routine prostate-specific antigen (PSA) test led to the diagnosis for Farrington, who lives in Quincy, Massachusetts, and worked for many years in the information technology industry.
During his diagnosis and treatment, Farrington was surprised to learn that, as an African-American man, he was part of a group at increased risk of prostate cancer. As of 2016, African-American men are 1.7 times more likely than white men to get prostate cancer and 2.4 times more likely to die of it.
In 2003, Farrington founded the Prostate Health Education Network (PHEN), a nonprofit organization dedicated to education and to reducing disparities in prostate cancer incidence and mortality affecting African-American men. Likely due in part to increased awareness of prostate cancer risk in African-American men, PSA testing rates have risen in this group since the 1990s. In 2012, African-American men 45 to 60 years old in the U.S. reported higher rates of screening than white men, according to one study, although older African-American men were still screened at slightly lower rates than white men.
Also in 2012, the U.S. Preventive Services Task Force (USPSTF), a panel of volunteer experts that rates screening and preventive health services, issued new prostate cancer screening guidelines. The USPSTF gave PSA testing its lowest grade: D. Physicians are advised to discourage patients from using D-rated screening tests. Prior to the new guidelines, the USPSTF had said there was insufficient evidence to determine if risks of PSA testing outweighed harms for men under age 75. The D score alarmed patient advocates.
“We worked hard to get the level of PSA testing for black men at the same level it was for white men,” says Farrington. “We worked years to make that happen, and we achieved that goal.” After all the hard work, the USPSTF was saying the test should not be offered routinely to any men.
A cancer screening test is successful when it finds an early-stage cancer that can be treated effectively, when a later-stage diagnosis would have been fatal. At one time, the notion that early detection of cancer must save lives was often embraced uncritically. But recently, researchers have taken a more nuanced look by considering screening’s possible harms, such as false-positive results and early-stage diagnoses that can sometimes lead to unneeded treatments.
The “catch-it-early” screening strategy works better for some cancers than for others. For example, colorectal cancer usually grows slowly and is highly treatable early on. Colonoscopy, one method of colorectal cancer screening, even allows for the removal of precancerous polyps. Many experts agree that the risks of colorectal screening, such as injury during a colonoscopy, are worth taking for the relatively large potential benefit. “Colon cancer is the perfect cancer for screening,” says gastroenterologist Philip Schoenfeld of the John D. Dingell VA Medical Center in Detroit. Cervical cancer screening also can catch lesions before they become cancerous, making its benefits potentially quite large. While recommendations on the frequency and method of cervical cancer screening have shifted over the years, U.S. guidelines are now largely in harmony.
The same cannot be said for screening for the two most common cancers in U.S. women and men, respectively: breast cancer and prostate cancer.
In 2009, the USPSTF updated its guidelines on breast cancer screening to say that every-other-year mammograms should be given routinely to women starting at age 50. From 2002 to 2009, the task force had recommended that women begin mammography routinely at age 40. The changing guideline prompted major backlash from some members of the breast cancer community. After all, breast cancer deaths in the U.S. had declined by approximately one-third since their peak around 1990. Why was the USPSTF weakening its recommendations for mammograms, which grew in popularity in the 1980s and 1990s, in the face of such seemingly positive statistics?
In making the change to mammograms starting at age 50, the USPSTF cited a relatively low number of lives lengthened in women aged 40 to 49, amid high rates of false-positives. Women in their 50s benefit more from mammograms than those in their 40s, and women in their 60s benefit the most, according to the USPSTF, which reaffirmed its 2009 mammography recommendations in 2016. The task force gave mammography for women in their 40s a C grade, meaning doctors and patients should decide whether to start screening on a case-by-case basis.
Besides guiding primary care doctors’ recommendations, USPSTF guidelines help determine preventive care that insurance companies must make available at no cost to patients under the 2010 Affordable Care Act, with services rated A and B making the cut. This link has made the task force’s C recommendations particularly controversial—so much so that Congress has twice mandated that mammograms be provided by private insurers at no cost to women in their 40s, in keeping with the 2002 USPSTF guidelines rather than the current ones.
A common downside of a mammogram is a false-positive result. Each time a woman in her 40s gets a mammogram, she has about a 12 percent chance of having a false alarm, according to research published in February 2016 in
Annals of Internal Medicine. She has a 1.6 percent chance of getting a breast biopsy. The rate of false-positives falls as women enter their 50s and 60s. Some screening experts have expressed concern that unnecessary call-backs and biopsies can cause distress, while others have argued that gaining information is more valuable than avoiding possible false alarms.
“As a woman, I feel it’s even paternalistic to say, ‘We don’t want you to have this anxiety coming back for additional imaging and a possible biopsy,’ ” says radiologist Elizabeth Morris, chief of the Breast Imaging Service at Memorial Sloan Kettering Cancer Center in New York City and president of the Society of Breast Imaging. Memorial Sloan Kettering recommends yearly mammograms starting at 40 for average-risk women. The Society of Breast Imaging, an organization for breast imaging professionals based in Reston, Virginia, is currently running a campaign called “End the Confusion” to advocate for annual mammograms beginning at age 40.
Another problem with mammograms is overdiagnosis, in which “you detect a lesion that wouldn’t have caused a problem for the woman in her lifetime,” says epidemiologist Kirsten Bibbins-Domingo of the University of California, San Francisco. Bibbins-Domingo also is the current chairperson of the USPSTF. A positive result on a mammogram can sometimes lead to unnecessary surgery, chemotherapy and radiation, with life-altering and often long-lasting or permanent side effects. Researchers are working to get better at differentiating between tumors that will never cause problems and ones that will progress. For now, most women diagnosed with any type of breast cancer are treated.
Download a PDF chart to review screening guidelines from the U.S. Preventive Services Task Force and the American Cancer Society.
Experts generally agree that overdiagnosis occurs, but disagree on how common it is. A paper published in the
New England Journal of Medicine in 2012 found that a “best-guess” model suggested 31 percent of breast cancers in women age 40 and over are overdiagnosed. The authors argue that the decline in breast cancer deaths in the past several decades resulted more from improved treatments than increased screening.
Morris calls this estimate of overdiagnosis “wildly overinflated,” pegging the rate at about 10 percent. She and others argue that screening itself does not lead to unnecessary treatment. Rather, the steps following screening do. But Vinay Prasad, a hematologist-oncologist at Oregon Health & Science University in Portland who has argued that studies have never fully proved that screening saves lives, compares screening and its consequences to a freight train hurtling down the tracks. “Almost nobody can find out, ‘You may have cancer,’ ” and then take no action, Prasad says. “Most people can’t do that, including most doctors.”
Parsing PSA’s Benefits and Harms
The protein in the blood measured by the PSA test often rises in the presence of prostate cancer. PSA also rises due to benign conditions, such as an enlarged prostate or prostate inflammation. According to a large U.S. trial, among men who received annual PSA tests for four years, around one in eight had at least one false-positive result and one in 18 had at least one biopsy due to a false-positive result. Just a quarter of men who get biopsies after PSA screening are diagnosed with prostate cancer. Meanwhile, the side effects of a prostate biopsy can be serious and include bleeding, pain, urinary tract problems and infections.
Prostate cancer screening also results in high rates of overdiagnosis, due to the high number of prostate cancers that do not cause problems, particularly in men who are nearing the end of their lives. According to a review article published in the October 2015 issue of the International Journal of Cancer, an estimated 59 percent of all men 80 years or older with no known prostate cancer history who were autopsied upon their deaths were found to have some form of prostate cancer. (Just 5 percent of men autopsied under age 30 had prostate cancer.) Despite these findings, many in the prostate cancer community argue in favor of PSA testing.
Some patient advocates say the USPSTF should have come up with a separate prostate cancer screening recommendation for African-American men because of their increased risk. The two main studies the USPSTF cited in support of the D designation for the PSA test, one from Europe and one from the U.S., included few men of African descent, PSA advocates argue.
“If [the USPSTF doesn’t] have the scientific evidence based upon their own criteria for issuing a recommendation, they have to issue [a letter] I,” for insufficient evidence, says Farrington, the prostate cancer survivor who founded PHEN. “They violated their own criteria because they had no evidence for African-American men.”
In September 2016, Farrington invited prostate cancer researchers, clinicians and patient advocates, as well as Bibbins-Domingo, to participate in PHEN’s 12th annual African-American Prostate Cancer Disparity Summit in Washington, D.C. The USPSTF plans to release a revised prostate cancer screening guideline in 2017 and has solicited comments from the prostate cancer community to help it develop a research plan.
At the summit, disappointment with the USPSTF’s recommendations remained fresh. “In the process of going from [I to D], it struck a big blow for those men at highest risk,” said J. Jacques Carter, an internist at Beth Israel Deaconess Medical Center and a professor of medicine at Harvard Medical School in Boston, in remarks at the summit.
“It’s had a terrifying impact,” said Charlie Hill, a prostate cancer survivor and co-founder of the Hampton Roads Prostate Health Forum in Hampton, Virginia, of the downgrade. Hill described his years spent encouraging local African-American men to get prostate cancer screening and said that interest has declined since the new recommendations.
A coalition of urologist associations and patient groups, including PHEN, sent comments to the USPSTF advocating for a focus on high-risk groups in the upcoming recommendations. St. Louis-area doctors, survivors and community health groups formed the Prostate Cancer Coalition to advocate for doctors to talk about screening pros and cons with African-American men.
Using Genetics as a Guide
While advocates for cancer testing call for customized guidelines for different groups of people, some researchers are working to establish even finer distinctions with the help of genetics. People with gene mutations that greatly increase their risk of getting cancer already are sometimes advised to get extra screening and to have preventive surgeries. These include people with a BRCA gene mutation that puts them at greater risk for breast and other cancers, or those who have Lynch syndrome, which is caused by variations in certain genes involved in DNA repair and increases the risk of colorectal and other cancers. But researchers are also becoming aware of more common genetic variants that can lead to smaller changes in cancer risk that could nevertheless be significant.
A coalition of scientists and clinicians at a group of University of California medical centers and Sanford Health, a medical system in North and South Dakota, are testing personalized, risk-based screening. Their study, called WISDOM, will screen some women for breast cancer with mammograms yearly beginning at 40 and will personalize screening for other people based on factors that can affect their breast cancer risk, such as breast density, family history, age, race, ethnicity, history of benign breast conditions and genetics. WISDOM began enrolling people in California at the end of August 2016.
“It’s clear that there are various recommendations for mammographic screening for women by different organizations, and none of them actually are recommending the same thing,” says Laura van ’t Veer, principal investigator for the WISDOM trial at the University of California, San Francisco site. She and her colleagues hypothesize that all these recommended screening regimens might be right—just each for a different group of women.
Beyond taking into account traditional risk factors for breast cancer, van ’t Veer and her colleagues will check for mutations in nine breast cancer-related genes, including genes that dramatically elevate risk and more common mutations that increase risk more modestly. They will also look at a larger number of spots in the genome where common variations in a single DNA base-pair affect cancer risk very slightly. Alone, any one of these variants, called single nucleotide polymorphisms (SNPs), would be an inconsequential blip in a woman’s cancer risk. Taken together, a number of SNPs can add up to a moderate increase in breast cancer risk that may be clinically significant.
The initial list of SNPs is based on data from largely white participants, says van ’t Veer, but she is making an effort to include data from women of more diverse ethnic and racial backgrounds. For instance, she is working to incorporate SNPs related to elevated risk in African-American and Asian women into the model, as well as a SNP that seems to protect some Latinas from breast cancer.
Prostate cancer screening could also benefit from a precision approach that takes genetics into account, says genetic epidemiologist Jianfeng Xu of NorthShore University Health System in Illinois. In the summer of 2016, he began a trial of personalized screening at NorthShore, estimating inherited risk of prostate, breast and colorectal cancer using genetics and providing modified screening schedules for patients based on their results. He has developed different lists of risk-related SNPs from reviews of previously published literature for patients who are white, African-American and Asian.
Some urge caution in pursuing precision screening, however. Morris, of the Society of Breast Imaging, worries that strategies that decrease mammography in supposedly low-risk women will end up preventing women from getting screened who could have benefited. “Seventy percent of women who get breast cancer don’t have identifiable risk factors other than the fact that they’re female and they’re getting older,” she says.
Bibbins-Domingo of the USPSTF reports that the task force is closely watching efforts to personalize screening. But as researchers formulate algorithms and harness sequencing technology in the service of screening, the bar for medical intervention in individuals without symptoms should still remain high, says Bibbins-Domingo.
“We’ve always tried to do a one-size-fits-all screening for everybody. … That probably is the wrong way to do it,” says Prasad of the Oregon Health & Science University. “There may be a group who benefits from screening. It’s almost surely unlikely to be [true that] everybody benefits from a single screening test.”
December 30, 2016