A simple risk assessment tool could identify women at higher risk of developing invasive breast cancer and so allow clinicians to better target mammography-based breast cancer screening to those who need it most, the results of an Australian study suggest.
The researchers analyzed data on more than 40,000 women aged 50 to 59 years who had taken part in the national breast cancer screening program. They examined whether the Gail risk assessment model (now known as the Breast Cancer Risk Assessment Tool) could identify those most at risk following a negative screening result.
Although the model overestimated the number of cancer cases overall, particularly for patients with the highest risk, it was able to identify a group of women who had a more than twofold increased risk for breast cancer relative to those with the lowest risk.
"This suggests that the existing Gail model is suitable for assigning women into groups at significantly different risk of invasive breast cancer in the 5 years following a negative screen," the authors write.
"This information has the potential to enable more personalised, risk-based screening strategies that aim to improve the balance of the benefits and harms of screening," they add.
The study was published online December 20 in Breast Cancer Research.
Lead author Carolyn Nickson, PhD, Melbourne School of Population and Global Health, University of Melbourne, Carlton, Victoria, Australia, told Medscape Medical News that it would be feasible to collect the data necessary for developing screening protocols based on risk.
"However, changing breast cancer screening protocols for different risk groups would be much more complicated, and we need further evidence to determine which protocols would be optimal," she said.
In practice, this implies "a wide range of options, with differing levels of complexity when it comes to implementation in terms of, eg, costs, workforce requirements, data management, and clinical workflow," she said.
Nickson emphasized that although changing screening intervals may be the simplest approach, "we do not yet understand the costs, benefits, and harms of targeted screening intervals compared to the current program, or how these would compare to, say, offering screening technologies such as digital breast tomosynthesis or adjunctive ultrasound to specific groups of women."
As to whether it would be feasible to persuade women to accept a change in their screening interval, she pointed out that the shift to more personalized medicine means providing high-quality information to help women make informed decisions.
This information, Nickson said, includes the potential harms and benefits of screening, such as estimates of overdiagnosis and information on the reduced accuracy of mammographic screening for women with dense breasts.
"In this context, I think this is plausible that women would accept being assigned to or offered a screening protocol based on their risk profile," she added.
"Here in Australia, screening participation is around 55% in the target age range of 50 to 74 years," she added. "Perhaps a more tailored program of screening would increase the program's appeal to some women who do not participate."
Interest in Risk-Stratified Screening Protocols
Although universal age-based breast cancer screening is conducted in many developed countries around the world, there are increasing concerns about overdiagnosis and the appropriateness of mammography for women with dense breasts.
Furthermore, in recent years, there has been an increase in measurable risk factors for breast cancer, which has fueled an interest in personalized, risk-stratified screening protocols that take into account both potential benefits and possible harms.
The Breast Cancer Risk Assessment Tool is a relatively simple tool for predicting the risk for invasive breast cancer in women who do not have a history of the disease, the authors comment.
To examine the tool's performance in a large population, the researchers gathered data on women in the lifepool cohort. Participants were recruited from Australia's population-based mammography screening program from 2010.
The women were at least 40 years old at enrollment. They completed a detailed baseline questionnaire on sociodemographic, lifestyle, and health-related factors, and their records were linked to screening and cancer registry data.
For the analysis, the team included 40,158 women aged 50 to 69 years.
During a median follow-up of 4.3 years, 564 women (1.4%) were diagnosed with invasive breast cancer. The median time from reference screen to diagnosis was 2.2 years; the maximum was 5.3 years.
Three women were diagnosed with lobular carcinoma in situ.
There were 243 deaths during the study period, of which eight were due to breast cancer.
Women who developed invasive breast cancer tended to be older at enrollment, were more likely to have first-degree female relatives with breast cancer, and were more likely to have undergone a breast biopsy.
The Gail model predicted 612 cancer cases; the observed rate was 564, giving an expected-observed ratio (EO) of 1.09.
The model overestimated risk for women with the top decile of Gail scores, at an EO of 1.65.
With regard to 5-year risk quintiles, the team calculated that, compared with women with a median level of risk (Q3), those in Q1 had a lower risk for invasive breast cancer, at a hazard ratio of 0.59 (P < .001), as did those in Q2, at a hazard ratio of 0.63 (P = .001).
Women in the highest quintile (Q5) had a significantly greater 5-year risk of developing invasive breast cancer than those in Q3, at a hazard ratio of 1.34 (P = .014).
Compared with women in Q1, the risk for invasive breast cancer increased with increasing quintiles, becoming significant at Q3. Women in Q5 had a substantially increased risk vs those in Q1, at a hazard ratio of 2.28 (P < .0001).
Overall, the Gail model had a receiver operating characteristic area under the curve of 0.59 when continuous scores were used.
The results were similar for women aged 50 to 59 years and 60 to 69 years with use of a reduced Gail model that excluded information on hyperplasia and ethnicity.
Similar to Study From Last Year
The current findings from Australia are similar to those from a British study published in mid-2018, which explored hypothetically the idea of providing breast screening only for women who were genetically at higher risk for the disease.
The study showed that limiting breast cancer screening only to women at higher levels of risk for the disease could improve cost-effectiveness and reduce overdiagnosis while maintaining the benefits of screening.
However, at that time, Daniel B. Kopans, MD, founder, Breast Imaging Division, Massachusetts General Hospital, Boston, commented that being able to target screening to women with a high risk for breast cancer means "you first need to prove that you can safely identify women at no risk."
He said, however, that no one "has ever been able to define women who have no risk of developing breast cancer."
Kopans pointed out that women with a genetic abnormality that places them at high risk account for less than 10% of breast cancers diagnosed annually. Family history and other factors add another 15%.
"This means that 75% of women diagnosed with breast cancer each year have no demonstrable factors to elevate their risk," he said, adding: "As far as we can tell in 2018, everyone is at risk."
The study was supported by Cancer Australia, the National Breast Cancer Foundation Collaborative, and the National Breast Cancer Foundation. The authors have disclosed no relevant financial relationships.
Breast Cancer Res. Published online December 20, 2018. Full text
For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc
Medscape Medical News © 2019
Cite this: Could Gail Model Allow Personalized Breast Screening? - Medscape - Jan 08, 2019.