The pros and cons of screening for three cancers
By Phyllis Maguire
PHILADELPHIA—When The Lancet published an article last fall questioning the efficacy of mammography, it breathed new life into a long-running debate about that screening procedure. At an Annual Session presentation on cancer screening, however, a group of panelists said that the debate over mammography is only one part of a much larger controversy surrounding cancer screening.
With experts challenging the value of screening for some cancers, physicians find themselves in a quandary, said panel moderator Barnett S. Kramer, FACP, associate director for disease prevention at the National Institutes of Health (NIH). "The issue has led to a clash between our intuition—which says that detecting cancers sooner must save lives—and medical science, which is calling into question the efficacy of some screening," he said.
Dr. Kramer pointed out that this tension also has the potential to damage physician-patient relationships. "When we convince the public that a procedure is beneficial before we have the facts," he said, "it's hard to come back and say, 'We've now found out we were incorrect.' It calls into question the credibility of organizations releasing guidelines and leads patients to question other recommendations we make."
During the presentation, panelists talked about screening measures for breast cancer, lung cancer and colorectal cancer. Here is an overview of the screening benefits and limitations they presented for each disease.
Back in the 1990s, experts argued about the age when women should start being screened. Now, "some investigators are asking whether we should screen any women at all," said Joann G. Elmore, MD, "and they're concluding the answer is 'no.' "
Dr. Ellmore says women have been oversold on breast cancer screening and its benefits, without being informed about its potential risks and limitations.
The data. Dr. Elmore, head of general medicine at Harborview Medical Center in Seattle and associate professor of medicine at the University of Washington School of Medicine, pointed out that most physicians still agree that mammograms help save lives. "We think that the data show that women over age 50 should be screened every one to two years with mammography and clinical exam," she said. "We think this can reduce breast cancer mortality by 20% to 30%."
(For evidence about other screening methods, see "Breast cancer screening: self exams and clinical exams."
Recently, however, some researchers have attacked that assumption. New meta-analyses, including the controversial article published in The Lancet, have challenged the efficacy of mammography. Researchers claim that many of the studies traditionally used to support mammography were tainted by fatal research flaws—from faulty randomization methods to unacceptable variations in screening techniques.
Confusing matters is uncertain evidence about mammography's benefits for women under 50. "We have to screen 4,000 women in their 30s to detect one cancer," Dr. Elmore said. For women in their 40s, she added, one cancer is typically found among every 800 women screened. (Two separate National Institutes of Health consensus panels, the latest in 1997, concluded that data didn't bear out universally recommending screening for women in their 40s.)
The unclear evidence for mammography benefits in women under 50 has led to conflicting guidelines. The U.S. Preventive Services Task Force (USPSTF), for example, now recommends mammography for women 40 and older, although it notes that benefits are uncertain in younger women. Canadian experts, on the other hand, do not recommend mammography for women under the age of 50.
Risks. While the debate over data rages on, Dr. Elmore said that one thing is clear: "Women have been oversold on breast cancer screening and its benefits, without being informed about its potential risks and limitations."
Mammography exposes patients to radiation as well as discomfort, she explained. In addition, the kind of cancers that mammograms frequently detect—ductal carcinomas in situ—may never develop into aggressive malignancies, even though many women opt for aggressive treatment, including mastectomy and chemotherapy. As a result, Dr. Elmore said, many experts now question whether the cancers that mammography detects are being overtreated.
Another major limitation is the subjectivity of reading mammograms. Dr. Elmore said that when she showed 150 mammograms to 10 different radiologists, they disagreed about whether to order biopsies for one-quarter of them. Even more startling, she said, was the fact that 9% of the physicians who claimed a biopsy was needed disagreed about which breast to examine.
Given those kinds of discrepancies, she said, it's no wonder that mammograms lead to false negatives, a major limitation she said physicians need to guard against. "Do not be falsely reassured," Dr. Elmore said. "If you feel a suspicious lump and the mammogram is normal, go after it anyway. Schedule follow-up, and get tissue."
If false negatives are one downside of mammography, false positives are even more common. Women run a 10% risk of having a false positive with one screening; after 10 screenings, that risk jumps to 50%.
Economics also come into play: For every $1 million spent on screening mammograms, an additional $260,000 to $330,000 is spent working up false positives—which can come, Dr. Elmore pointed out, with steep psychological costs.
"If we have a hard time with all this," she said, "imagine what our patients are feeling." Patients already have a host of misperceptions about breast cancer, Dr. Elmore added. Women in their 40s are convinced they run the same risk of being diagnosed with breast cancer as older women. Women in their 50s and 60s, meanwhile—who might benefit the most from screening—are often so anxious about discovering they have the disease that they avoid getting screened.
"We have a lot of patient education to do," Dr. Elmore concluded, adding that physicians need to raise the issue of breast cancer screening routinely with patients and discuss both the benefits and risks.
While the debate continues over mammography, most experts believe that there still isn't enough evidence to recommend mass screening for lung cancer, the country's biggest cancer killer.
Pamela M. Marcus, PhD, an epidemiologist with the cancer prevention division at the National Cancer Institute in Bethesda, Md., explained that the search for effective screening has traditionally focused on chest X-rays. One of the largest studies—which ended in 1983—was able to detect more cancers in the intervention group, but more participants in the intervention arm died of the disease than did control group participants.
Follow-up research in the 1990s confirmed those puzzling results, but it also found that members of the intervention arm survived longer after their lung cancer diagnosis. Researchers concluded that the most likely explanation was what Dr. Marcus called "overdiagnosis," a phenomenon that occurs when screening detects asymptomatic tumors that would never have been diagnosed during a person's lifetime.
Overdiagnosis will occur if that screening detects indolent lesions that don't pose a fatal threat. Although not all researchers agree that such a class of indolent lung cancer lesions exists, Dr. Marcus said support for their existence comes from other organ sites.
Overdiagnosis is a serious issue because it challenges the value of screening. By identifying lesions with little clinical relevance, Dr. Marcus explained, screening can cause patients to undergo unnecessary workups, anxiety, depression and cancer therapy that wouldn't have otherwise occurred.
Computed tomography (CT) has shown some promising results in lung cancer screening, and the U.S. government is now funding a major randomized controlled trial to compare CT scan screening results to chest X-rays. Meanwhile, the jury is still out on whether lung cancer screening saves lives.
"The bottom line," Dr. Marcus said, "is that screening benefits must outweigh the harms."
While the efficacy of a screening test is critical to its success, the fight to detect colorectal cancer is only half the battle.
Gastroenterologist David F. Ransohoff, FACP, professor of medicine and epidemiology and director of the NIH-supported K30/Clinical Research Curriculum at the University of North Carolina, Chapel Hill, said that while evidence shows colorectal screening works in average risk patients, convincing patients to get screened is now the major challenge.
Dr. Ransohoff said that 56,000 Americans die of colorectal cancer each year, making it the second most deadly cancer in the country. Randomized controlled trials held in the early 1990s showed that screening with annual fecal occult blood tests (FOBTs) for patients 50 and older reduced colorectal cancer mortality by about 30%. Those results led the USPSTF in 1996 to endorse annual FOBT screening.
The task force also recommended flexible sigmoidoscopy screening every five years based on findings from case-control studies. (Randomized trials on flexible sigmoidoscopy are currently underway.)
Although the American Cancer Society has in the past recommended a combination of FOBT and flexible sigmoidoscopy, many organizations now recommend both those screening routes as well as colonoscopy, a screening technique that has recently received media attention. Medicare began reimbursing for screening colonoscopy last year.
While prominent medical organizations have endorsed a range of screening techniques, far fewer patients are screened for colorectal cancer than for breast cancer. Dr. Ransohoff pointed out that less than 30% of patients 50 and older receive FOBT screening and less than 15% undergo flexible sigmoidoscopy.
What's behind those low numbers? "Physicians don't recommend colorectal cancer screening," he said. "It's something we have to stop being shy about."
Also, ordering screening colonoscopy is more complicated and time-consuming than ordering a mammogram, and reimbursements for several screening tests—including flexible sigmoidoscopy, which general internists have performed for years—have fallen.
While screening for colorectal cancer may be efficacious, many of the dilemmas seen in other types of cancer screening apply. How, for example, can physicians distinguish potentially deadly cancers or pre-cancerous lesions from more innocuous growths to avoid overdiagnosis and over-treatment?
"We need to figure out which of those neoplasms we need to remove," Dr. Ransohoff explained. "We also need to figure out which adenomas after removal indicate a high future risk of colon cancer."
Adenomatous polyps, which are the precursors to cancer, are common, Dr. Ransohoff said: About 30% to 50% of patients above age 50 have one or more. But only 10% of those adenomas are 1 cm or bigger, and only 1% of polyps over 1 cm will become cancerous in any given year.
Experts agree that average risk patients over 50 should be screened, but they still argue about how to manage those who have had adenomas removed, as well as other high-risk patients. Some experts now believe that a patient with a single small adenoma doesn't need surveillance colonoscopy, while other groups are considering lengthening the intervals for recommended surveillance.
"Post-screening follow-up and surveillance will be huge cost drivers," Dr. Ransohoff said. "We need to sort this out over the next few years."
While most of the controversy over breast cancer screening centers on mammography, physicians commonly recommend or perform other screening techniques. Here is a look at the evidence for each:
Self exams. This procedure is simple, inexpensive and has the added benefit of helping women get more comfortable with their bodies, explained Joann G. Elmore, MD, head of general medicine at Harborview Medical Center in Seattle and associate professor of medicine at the University of Washington School of Medicine.
The problem, she said, is that the technique has real limitations, such as low accuracy and many false positives. And no data have shown that self exams reduce breast cancer mortality.
The result is a gamut of recommendations. The American Cancer Society recommends that women perform self exams, while the U.S. Preventive Services Task Force (USPSTF) claims there is insufficient evidence to recommend either for or against the test. A Canadian group last year gave the technique a thumbs-down, saying self exams provide no clinical benefit while inducing "undue stress" in women who think they feel a lump.
Clinical breast exams. Physicians' clinical breast exams are also simple and inexpensive. While the accuracy of the clinical exam is not as good as mammography, Dr. Elmore said that studies show that clinical exams can catch from 3% to 45% of the cancers that mammograms miss.
The test's big limitation, however, is time. "Studies recommend that we spend five minutes per breast, or a total of 10 minutes, and we just don't have that kind of time," said Dr. Elmore. She noted that she often asks patients to make a separate appointment for a clinical breast exam and a discussion about breast cancer screening.
One Canadian study weighed clinical exam alone against clinical exam plus mammogram and found no difference in subsequent breast cancer mortality, Dr. Elmore said. Researchers did find, however, that women randomized to mammography had more breast biopsies.
Expert recommendations only add to the confusion. The American Cancer Society recommends clinical exam with mammogram for women age 40 and older, while the USPSTF says there is insufficient evidence for clinical exams, either pro or con.
Internist Archives Quick Links
Have questions about the new ABIM MOC Program?
One Click to Confidence - Free to members
ACP Smart Medicine is a new, online clinical decision support tool specifically for internal medicine. Get rapid point-of-care access to evidence-based clinical recommendations and guidelines. Plus, users can easily earn CME credit. Learn more