Cost-effectiveness of Breast MRI Screening by Cancer Risk

Institution: Stanford University
Investigator(s): Allison Kurian, M.D. -
Award Cycle: 2005 (Cycle 11) Grant #: 11FB-0051 Award: $49,647
Award Type: Postdoctoral Fellowship
Research Priorities
Community Impact of Breast Cancer>Health Policy and Health Services: better serving women's needs



Initial Award Abstract (2005)
Introduction: Breast magnetic resonance imaging (MRI) is increasingly used as a screening tool for the early diagnosis of breast cancer. Although breast MRI has been shown in some settings to detect cancers when they are smaller, and potentially more curable, than mammography alone, MRI is more costly than mammography and can lead to a high rate of unnecessary breast biopsies, causing patients anxiety and discomfort. Screening breast MRI is likely to be of greatest benefit (measured as a ratio of its cost to its effectiveness) in women at increased risk of breast cancer for whom mammography performs poorly: women with dense breast tissue on mammography and women with high inherited breast cancer risk, particularly women in both of these groups who have other risk-increasing characteristics.

Hypotheses: The goal of this project is to identify cost-effective protocols involving the addition of screening MRI to mammography for early breast cancer detection. The most cost-effective screening protocol will be the one which saves the greatest number of years of life (accounting, using women's preferences, for quality of life) at the lowest cost (in side-effects and in dollars) to screened women and to society. We will focus on women with dense breast tissue and women with high inherited breast cancer risk, and evaluate the impact of factors which increase (such as prior abnormal breast biopsy or hormone replacement therapy) and reduce (such as anti-hormonal medications or surgical removal of the ovaries) breast cancer risk on the cost-effectiveness of breast MRI screening. Depending on a woman's risk level and age, we will determine her most cost-effective breast MRI screening schedule (for example, yearly or every 2 years).

Methodology: We will modify a computer simulation model of mammographic screening, which generates breast cancer outcomes for individual women, and presents their aggregated results on a population level. This model, which was developed and validated by our research group, will be adapted to include the cancer detection ability of MRI, and the characteristics of women with dense breast tissue and women with high inherited breast cancer risk, including the impact of risk-modifying factors. The model will be modified using data from previous literature and from cancer registries. The preferences and experiences of high-risk women about breast MRI screening and its alternatives will be collected using survey instruments, and incorporated into the model. It will be used to calculate the estimated mortality reduction of screening MRI, its cost-effectiveness, and its optimal schedule according to a woman's age and breast cancer risk level.

Innovative elements: The question of how best to use screening breast MRI would be answered with a randomized clinical trial, which could determine whether screening MRI leads to reduced mortality through early breast cancer detection, but such a trial is ethically questionable given evidence of MRI's ability to find cancer early in some clinical settings. This project is innovative in its use of a computer simulation model to address the questions of mortality benefit, cost-effectiveness, and the optimal schedule of screening breast MRI, which are not likely to be answered otherwise in a timely fashion. Moreover, it is innovative in the inclusion of detailed information about breast cancer risk levels and women's measured preferences about MRI screening and its alternatives; this will result in a highly practical guide as to which women will most benefit from breast MRI, in terms of their survival and their quality of life.

Advocacy involvement: This research aims to address the key human issue of how best to use a costly technology that has been shown to improve breast cancer detection, and promises to reduce mortality, among the many women who are not well served by standard mammographic methods of breast cancer screening. It aims to ensure that women at increased risk of breast cancer are significantly included in this project by using a survey instrument to measure their preferences regarding breast MRI screening and its alternatives, and by presenting results which incorporate those measured preferences. The design of this research project has benefited from the advice and involvement of Ms. Sue Friedman, founder and executive director of the breast cancer advocacy group FORCE (Facing Our Risk of Breast Cancer Empowered), whose letter of support is attached; her participation and that of other breast cancer advocates will be encouraged throughout the completion of this work.


Final Report (2007)
Introduction: Breast magnetic resonance imaging (MRI) is increasingly used as a screening tool for breast cancer. Although breast MRI has been shown to detect cancers when they are smaller, MRI is more costly than mammography and can lead to a high rate of unnecessary breast biopsies, causing patients anxiety and discomfort. Screening breast MRI is likely to be of greatest benefit in women at increased risk of breast cancer for whom mammography performs poorly.

Topic Addressed: The topics addressed in this research project are the effectiveness (measured in terms of breast cancer mortality reduction) and cost-effectiveness (measured as a ratio of cost versus effectiveness) of screening women with breast MRI, incorporating women’s preferences for management of their breast cancer risk, and adjusting to a woman’s individual level of breast cancer risk.

Progress Toward Specific Aims: We have made significant progress toward completion of Specific Aim 1 (to evaluate the cost-effectiveness of screening breast MRI as an adjunct to mammography in women for whom screening mammography has low sensitivity. We have adapted our pre-existing computer simulation model of mammography screening in the general population to simulate screening with mammography and breast MRI in women at high inherited breast cancer risk. After extensive review of published literature, we have incorporated the best data currently available on the incidence and prognosis of breast cancers in women with high inherited risk, the performance of mammography and breast MRI in high-risk women, the efficacy of breast cancer therapy, the costs of breast cancer screening and treatment, and the quality of life impact of breast cancer screening and treatment into this computer model. The results of our model-based analysis are that annual breast MRI screening protocols which are comparably cost-effective to other widely accepted interventions can be identified in BRCA1 mutation carriers, and in BRCA2 mutation carriers with dense breast tissue. Adding breast MRI to screening mammography is estimated to reduce breast cancer mortality in both BRCA1 and BRCA2 mutation carriers. We recently published these results in the Journal of the American Medical Association. We have also made significant progress toward completion of Specific Aim 2 (to measure preferences about screening breast MRI in women at high inherited risk) through questionnaire-based studies of high-risk women ascertained in cancer genetics clinics and an internet-based support group for women with inherited cancer predisposition; we presented preliminary results of this work at the 2005 San Antonio Breast Cancer Symposium. More recently, we have initiated work on a decision analysis evaluating the impact of breast cancer risk-modifying characteristics and interventions, such as pre-menopausal oophorectomy, on the life expectancies of women selecting strategies of prophylactic mastectomy versus screening with breast MRI.

Future Direction: We will continue to evaluate the impact of breast cancer risk-modifying characteristics and interventions on the effectiveness and optimal interval of screening with breast MRI. We will also consider the impact of high-risk women’s preferences for strategies to manage their predisposition to develop breast cancer.

Impact: Our research has resulted in estimates of the mortality reduction and cost-effectiveness associated with breast MRI screening added to mammography in BRCA1 and BRCA2 mutation carriers, which have direct relevance to clinical recommendations and health policy. We have also begun evaluation of high-risk women’s cancer risk management preferences. We anticipate that the results of this research may inform both individualized patient recommendations and population-level guidelines, by identifying optimal breast screening strategies according to women’s level of breast cancer risk and risk management choices.

Note: The postdoctoral PI resigned this award mid-way through year 2 as a result of receiving a faculty appointment.


Symposium Abstract (2007)
GOAL: To evaluate the effectiveness (measured in terms of breast cancer mortality reduction) and cost-effectiveness (measured as a ratio of cost versus effectiveness) of screening women with breast magnetic resonance imaging (MRI), incorporating women’s preferences and adjusting to a woman’s individual level of risk.

BACKGROUND: Breast MRI is increasingly used as a screening tool for breast cancer, but no study has yet shown that it decreases mortality. Although breast MRI has been shown to detect tumors when they are smaller, MRI is more costly than mammography and can lead to a higher rate of breast biopsies, causing patients anxiety and discomfort. Screening breast MRI is likely to be of greatest benefit to women at increased risk of breast cancer for whom standard screening mammography performs poorly.

METHODS: We adapted a previously developed computer simulation model of breast cancer screening with mammography to reproduce the natural history of breast cancer in women at high risk due to BRCA1/2 mutations. We incorporated the use of screening breast MRI into this simulation model, and estimated the impact on breast cancer mortality and cost-effectiveness. Simultaneously, we performed a questionnaire-based study of high-risk women’s preferences for breast MRI screening versus the alternative of prophylactic mastectomy.

RESULTS: Adding annual screening breast MRI to mammography in BRCA1/2 mutation carriers is estimated to yield a 23% relative reduction in breast cancer mortality, and a cost per quality-adjusted life year gained of $55,420 in BRCA1 and $130,695 in BRCA2 mutation carriers ages 35-54. These cost-effectiveness ratios are similar to those of widely accepted interventions in breast cancer management. Further model analyses suggest that MRI will add less benefit to screening women at lower breast cancer risk. High-risk women were more likely to choose breast screening over prophylactic mastectomy if they were older (mean age 47 versus 40) and had no prior breast cancer.

POTENTIAL IMPACT AND FUTURE RESEARCH: Our research produced estimates of the breast cancer mortality reduction and cost-effectiveness of MRI screening in BRCA1/2 mutation carriers, which were published in the Journal of the American Medical Association and cited in the recent American Cancer Society guidelines on screening breast MRI. In ongoing work, we are evaluating the impact of breast cancer-risk modifying characteristics and interventions, such as prophylactic oophorectomy, on the life expectancies of high-risk women who must choose between MRI-based breast screening or prophylactic mastectomy.