A Closer Look
New Guidance for Personalized Breast Cancer Screening
The most recent update of the U.S. Preventive Services Task Force (USPSTF) recommendations on breast cancer screening caused widespread confusion among many women and their physicians when it was released in November 2009.
The most important change made to the recommendations was small but significant—instead of recommending routine mammography for all women beginning at age 40, the task force suggested that the decision to be screened before the age of 50 "should be an individual one and take into account patient context."
"The recommendations were widely interpreted as meaning that no woman under the age of 50 should ever have a screening mammogram, and that was not the intention of the task force, nor would that be very good policy," said Dr. Diana Petitti, professor of Biomedical Informatics at Arizona State University, who was the vice chair of the task force at the time of the 2009 release. "Unfortunately, people couldn't get past the phrase 'recommends against routine screening mammography' for that age group," she continued.
"We were attempting to point physicians and women in the direction of individualized decision making, taking into account a woman's specific constellation of risk factors," explained Dr. Petitti. One persistent obstacle to that approach has been the dearth of guidance available to help individual women predict their risk of breast cancer.
The Influence of Individual Risk Factors
In a study published last month in the Annals of Internal Medicine, a group of researchers led by Dr. John Schousboe of the Park Nicollet Health Services in Minnesota examined the health benefits and cost effectiveness of screening mammography for different groups of women based on a set of known risk factors. They found that breast density was a powerful indicator of who might benefit from earlier screening.
"We tried to go beyond just reinforcing [the USPSTF guidelines] and give some guidance as to how one might go about assessing individual risk based on what we currently know," said Dr. Schousboe.
The researchers collected data on breast cancer incidence and mortality by age from the Surveillance, Epidemiology, and End Results (SEER) database, as well as data on breast cancers and false-positive results found during mammography from the Breast Cancer Surveillance Consortium—a project that NCI's Division of Cancer Control and Population Sciences (DCCPS) has been funding for over 15 years. The researchers also included data on change in quality of life after breast cancer diagnosis from a Swedish database.
Dr. Schousboe and his colleagues used this information to build models examining the health benefits and lifetime costs of mammography every year, every 2 years (biennially), or every 3 to 4 years for women ages 40 to 49, 50 to 59, 60 to 69, and 70 to 79. Within each age group, they refined their estimates based on breast density, history of breast biopsy, and family history of breast cancer—all known risk factors. Breast density was reported by radiologists in clinical practice using the Breast Imaging Reporting and Data System (BI-RADS), which classifies women's breast tissue into four categories, with 1 being the least dense and 4 being the most dense.
—Dr. John Schousboe
For women ages 40 to 49 without dense breast tissue (BI-RADS category 1 or 2) and no other risk factors, biennial mammography was not cost effective. In contrast, biennial mammography was cost effective for women ages 40 to 49 with dense breasts (BI-RADS category 3 or 4), as well as for women with average breast density (BI-RADS category 2) in that age group with a family history of breast cancer and a previous breast biopsy.
Additional findings suggested that density continues to influence risk for older women. Biennial mammography was cost effective for women ages 50 to 59 and 60 to 69 in BI-RADS category 2, 3, or 4, but not for women with the least dense breasts and no other risk factors. This small subset of women could likely go 3 or 4 years between mammograms, Dr. Schousboe and his colleagues proposed.
Cost effectiveness can help women and physicians compare the health benefits of a screening procedure across risk groups, explained Dr. Rachel Ballard-Barbash, associate director of DCCPS's Applied Research Program. "In the area of prevention, we need to screen a large number of people to find the relatively few who are actually at risk," she said.
"So when you see a very high monetary value for a year of life gained, what that says is that the overall benefit is relatively low, and the adverse effects are relatively high. It becomes expensive because you have to screen so many people to benefit one," she summarized.
Can Personal Risk Be Communicated?
The authors of the Annals of Internal Medicine study contend their findings make the case for basing future screening decisions on a baseline mammogram performed at age 40. "Women may choose to have mammography at age 40 years, and those with average or low breast density and no other breast cancer risk factors may choose to repeat screening at age 50 years (including reassessment of breast density) and start periodic screening at that point," they wrote.
"Knowing your risk is really important, and the reason breast density is so helpful is that other risk factors, such as family history, are present in only a relatively small minority of people. But half of the population has high breast density, including even more than half of women between the ages of 40 and 49," said Dr. Schousboe.
"The finding that breast density can be used to determine optimal screening strategies is a provocative result, because it suggests the presence of a biological mechanism that could be exploited to develop biomarkers, monitor risk, and screen for breast cancer on the basis of risk," wrote Dr. Jeanne Mandelblatt of the Lombardi Comprehensive Cancer Center and several colleagues in an accompanying editorial.
Although the biological mechanisms behind breast density are not fully understood, and "while those are important research questions, I think we have enough information now to start using [risk-based personalization]," said Dr. Petitti. "But whether or not doctors can communicate this kind of information is an open question."
"The time constraints primary care physicians are under make this communication difficult," commented Dr. Schousboe. Electronic health records that could provide automatic screening prompts based on risk information in a woman's medical history would help overextended physicians, he added.
What's Next for Personalized Cancer Screening
Several groups of researchers across the country are developing brief questionnaires to capture personal information and family histories that could be translated into a breast cancer risk score, according to Dr. Ballard-Barbash.
In addition, a new NCI program called Population-Based Research Optimizing Screening through Personalized Regimens (PROSPR) will study ways to individualize and improve screening for breast, colon, and cervical cancers based on recognized risk factors.
The idea of refining screening recommendations based on individual risk has long been the standard in other areas of medicine, explained Dr. Ballard-Barbash. For example, in cardiovascular disease, guidelines such as the Framingham Risk Score are used to identify the patients at high risk who would likely benefit from regular heart-disease screening. "We have not done this to any quantitative degree yet in breast cancer," she said.
The study by Dr. Schousboe and his colleagues and the studies that will likely emerge from PROSPR "are exactly the kind of research that we need—based on data taken from very large populations of women. [These studies are] really trying to understand what biological characteristics may influence risk and how we can identify the risk profiles that will determine which women would benefit or not benefit from specific types of screening," Dr. Ballard-Barbash concluded.
Read the related story in this issue: "Higher Breast Density Linked to Increased Cancer Risk and Aggressive Tumors"