Performance Measures of Magnetic Resonance Imaging Plus Mammography in the High Risk Ontario Breast Screening Program

Anna M. Chiarelli; Kristina M. Blackmore; Derek Muradali; Susan J. Done; Vicky Majpruz; Ashini Weerasinghe; Lucia Mirea; Andrea Eisen; Linda Rabeneck; Ellen Warner


J Natl Cancer Inst. 2020;112(2):136-144. 

In This Article


To the best of our knowledge, the High Risk OBSP is the first population-based breast screening program for high-risk women. Although many of the screening centers did not have experience with screening MRI before program initiation, the results compare favorably to those reported in meta-analyses of observational studies[10,21] and case series from academic centers.[16–19]

Not surprisingly, when stratified by risk, the highest cancer detection rates were among mutation carriers (26.8 per 1000) and when stratified by age, in women age 50–69 years (19.3 per 1000), comparable to previous studies.[13,16–19] In our cohort, 23% of screen-detected cancers were DCIS and the majority were detected by MRI only. This is consistent with results from the Toronto MRI study[20] but not those of earlier studies, which reported a lower DCIS rate[11,14,21] likely attributable to inexperience recognizing the characteristic non-mass enhancement on MRI.[54]

In the High Risk OBSP, the benefit of adding mammography to MRI was particularly small among mutation carriers age 30–39 years. Mammography detected only one additional cancer compared with 25 detected by MRI, whereas the combination reduced specificity by 8.2%. A recent meta-analysis based on observational studies also found that adding mammography to MRI did not statistically significantly improve sensitivity in BRCA1 mutation carriers age 40 years or younger.[29] In addition, no interval cancers were diagnosed in mutation carriers age 30–39 years, suggesting that annual MRI may be sufficient in this subgroup and does not justify the higher cost and reduced specificity of more-frequent MRI screening.[18] The use of mammography in addition to MRI when screening high-risk women age 30–39 years requires further consideration, particularly among BRCA1/2 mutation carriers, in whom the risk of radiation-induced cancers may be higher because of a defect in DNA repair.[3]

In our cohort, had women age 50–69 years been screened with mammography alone, more than one-half of the screen-detected cancers would have been undetected. Our findings do not support the National Institute for Health and Care Excellence guidelines recommending screening MRI in high-risk women only until age 50 years, except in women with dense breasts.[27] Contrary to findings in younger women, the addition of mammography to MRI in older women resulted in high sensitivity with little loss of specificity, particularly among mutation carriers. Similar results have been reported for BRCA1/2 mutation carriers older than 50 years.[28,29] These findings are not unexpected given sensitivity and specificity of mammography is higher in older compared with younger women in the general population.[17,55]

The main strength of our study was the inclusion of women screened in an organized screening program, therefore ensuring all radiologists and equipment met minimum quality standards and methods of evaluation and follow-up were similar for all women. Our study had a few limitations. Women excluded were primarily age 50–69 years or postmenopausal; this was, however, unlikely to appreciably alter results because they comprised only 5.6% of the cohort. Information on mammographic density was not collected; it is therefore unknown how this may have affected our results. Our results for women with previous chest radiation therapy are limited by sample size and require further follow-up. Although radiologists are aware of mammogram results before interpreting MRI and may be influenced, referral is independent of mammogram findings. Lastly, owing to the small number of women with mutations in genes other than BRCA1 or BRCA2, we are unable to comment specifically on this subgroup.

Our performance measures demonstrate that a population-based, high-risk breast screening program of annual MRI plus mammography delivers comparable results to those reported in prospective cohort studies and tertiary academic centers. Screening high-risk women age 30–39 years with an annual MRI only may be sufficient and should be evaluated further, particularly for mutation carriers. Among women age 50–69 years, screening is most effective when mammography is included with annual MRI.