Factors associated with receipt of surveillance breast MRI among racially/ethnically diverse women with a personal history of breast cancer.

1584 Background: Surveillance breast imaging among breast cancer (BC) survivors is recommended for early detection of local recurrence or contralateral BC. Breast MRI is increasingly used as a screening modality given its heightened sensitivity compared to mammography, and in 2018, the American College of Radiology released guidelines recommending supplemental breast MRI among women diagnosed with BC under age 50 or with dense breasts. We investigated demographic and clinical factors associated with receipt of surveillance breast MRI among women with a personal history of BC. Methods: We conducted a retrospective cohort study of women with stage 0-III BC between January 2018-June 2023 at Columbia University Irving Medical Center in New York, NY. We excluded women with bilateral mastectomies. The primary outcome was receipt of at least one breast MRI > 1 year after initial diagnosis. Data from the electronic health record included age at diagnosis, race/ethnicity, primary language, health insurance, first degree family history of BC, germline genetic testing results, stage at diagnosis, mammographic density (MD), and BC treatments (surgery, radiation, chemotherapy, and hormonal therapy). We performed multivariable logistic regression to assess factors associated with receipt of surveillance breast MRI. Results: Among 1,990 evaluable patients, mean age was 59.7 years (SD 12.2), and the cohort included 14% non-Hispanic Black, 33% Hispanic, and 6% Asian women. About 22% of women were diagnosed before age 50, 53% had dense breasts (BIRADS C-D), and 18% received at least one surveillance breast MRI > 1 year after diagnosis. On adjusted analysis, younger age at diagnosis, higher MD, first-degree family history of BC, receipt of germline genetic testing, and having a germline pathogenic variant were associated with receipt of breast MRI. Compared to non-Hispanic White women, Hispanic and non-Hispanic Black women had lower odds of receiving breast MRI (odds ratio [OR] = 0.44, 95% confidence interval [CI] = 0.31-0.63 and OR = 0.57, 95% CI = 0.38-0.86, respectively). However, compared to patients with commercial insurance, those with Medicaid were more likely to undergo breast MRI (OR = 1.57, 95% CI = 1.10-2.25). Results were similar when restricting the analysis to those diagnosed before age 50 or with dense breasts. Conclusions: Hispanic and non-Hispanic Black women with BC were less likely to receive surveillance breast MRI than their non-Hispanic White counterparts. Also, as patients with Medicaid were more likely to undergo breast MRI than those with commercial insurance, there may be varying health insurance coverage for breast MRI. These results highlight the need for more standardized guidelines surrounding surveillance breast MRI among BC survivors, which may inform public health initiatives aimed at promoting equitable breast imaging practices in this population.