Racial Disparities in Preoperative Breast MRI Use and Surgical Margin Outcomes among Patients with Recently Diagnosed Breast Cancer

Abstract

Purpose

To evaluate racial disparities in preoperative breast MRI use and surgical margin outcomes among patients with recently diagnosed breast cancer.

Materials and Methods

This retrospective study included patients with breast cancer who presented to a single cancer center between 2008 and 2020, underwent breast surgery, and self-identified as White or Black. Patients were divided into MRI or no-MRI cohorts based on preoperative MRI use. MRI use and positive surgical margin rates were determined for all patients and racial subgroups. Data were collected from the electronic medical record and analyzed using the χ² test for significance.

Results

The study included 28 384 female patients (mean age, 56 years ± 13 [SD]) with recently diagnosed breast cancer who self-identified as White (90.36%, n = 25 647) or Black (9.64%, n = 2737). Of the entire cohort, 32.78% (9305 of 28 384) underwent preoperative MRI. Black patients had a lower rate of preoperative MRI use than White patients (29.85% [817 of 2737] vs 33.10% [8488 of 25 647], respectively; P < .001). The MRI cohort had a lower positive margin rate compared with the no-MRI cohort (2.90% [133 of 4592] vs 4.78% [539 of 11 271], respectively; P = .03). In the no-MRI cohort, Black patients had a higher positive margin rate than White patients (6.17% [68 of 1103] vs 4.63% [471 of 10 168], respectively; P = .03). In the MRI cohort, there was no evidence of a difference in positive margin rates between Black and White patients (3.17% [12 of 379] vs 2.87% [121 of 4213], respectively; P = .90).

Conclusion

Compared with White patients, Black patients with breast cancer were less likely to undergo preoperative MRI and had a higher positive surgical margin rate.

Keywords: MR Imaging, Breast Cancer, Breast MRI

© RSNA, 2024

See also the commentary by Tran and Miles in this issue.

Summary

Black patients with recently diagnosed breast cancer were less likely to undergo preoperative breast MRI and more likely to have positive surgical margins when preoperative MRI was not performed, in comparison with White patients.

Key Points

• ■ Among patients recently diagnosed with breast cancer, Black patients had a lower preoperative MRI use rate compared with White patients (29.85% [817 of 2737] vs 33.10% [8488 of 25 647], respectively; P < .001).

• ■ Patients who underwent preoperative breast MRI had a lower positive surgical margin rate compared with those who did not (2.90% [133 of 4592] vs 4.78% [539 of 11 271], respectively; P = .03).

• ■ Among patients who did not undergo preoperative MRI, Black patients had a higher positive surgical margin rate compared with White patients (6.17% [68 of 1103] vs 4.63% [471 of 10 168], respectively; P = .03).

Introduction

Racial disparities in breast cancer outcomes have been well documented. Despite a lower incidence rate of breast cancer in Black compared with White female individuals, Black female individuals have a 40% higher breast cancer mortality rate (1,2). Variable MRI use contributes to existing racial disparities within breast cancer diagnosis and treatment. While prior literature has established that Black patients are less likely to undergo preoperative breast MRI (3–5) and more likely to have positive surgical margins compared with White patients (6), no study to date has examined these trends within the same patient cohort.

There are substantial gaps in knowledge regarding which patients newly diagnosed with breast cancer will benefit from preoperative breast MRI during cancer workup. Although there are relative indications for preoperative breast MRI, such as dense breasts and aggressive disease (ie, triple-negative breast cancer, high-grade ductal carcinoma in situ, and invasive lobular carcinoma) (7), clinical practice patterns are variable and based on surgeon preference.

In this study, we evaluated racial disparities in preoperative breast MRI use and surgical margin outcomes among patients with recently diagnosed breast cancer.

Materials and Methods

Patient Cohort

This is a Health Insurance Portability and Accountability Act– and institutional review board–compliant single-institution retrospective study. The need for informed consent was waived. There were no author conflicts of interest with regard to control of study data.

Data Collection

The patient cohort consisted of all female patients with recently diagnosed breast cancer (within 6 months of initial diagnosis) who presented to our institution between January 2008 and December 2020, underwent breast surgery as part of their cancer treatment, and racially self-identified as White or Black. Patients who identified as mixed race or Hispanic were excluded. Two subcohorts were created: patients who underwent breast MRI during initial diagnostic workup prior to breast surgery (MRI cohort) and patients who did not (no-MRI cohort). For each patient, birth date, age at diagnosis, race, ethnicity, preoperative MRI use, surgery type, and positive margin status (if applicable) were collected from the electronic medical record. Preoperative MRI use was defined as a breast MRI examination performed, at our institution or at an outside facility, within 6 months of the cancer diagnosis and prior to surgical intervention. Indications for preoperative MRI examinations at our institution include evaluating the extent of disease in patients with known cancer, typically determined by surgeon preference. Patients who underwent screening MRI examinations were not included. Surgery type included total or partial mastectomy. For patients who underwent partial mastectomy, surgical margin status was collected from the hospital’s central tumor registry database. Margin positivity was defined as invasive tumor cells present at the inked tissue edge of the resection specimen at histologic examination. Threshold for re-excision at our institution is a minimum clear margin of at least 1 mm for ductal carcinoma in situ and tumor at ink for invasive carcinoma. Patients who underwent total mastectomy or lacked follow-up data were excluded from the surgical margin analysis.

MRI Protocol

All patients underwent a contrast-enhanced breast MRI examination with a 1.5- or 3.0-T system (Discovery 750; GE Medical Systems) with a dedicated eight- or 16-channel breast coil. The gadolinium-based contrast agent was administered at a concentration of 0.1 mmol gadobutrol per kilogram of body weight (Gadavist; Bayer Healthcare Pharmaceuticals), at a rate of 2 mL/sec. The acquisition parameters for conventional steady-state dynamic contrast-enhanced MRI were repetition time/echo time of 7.9 msec/4.3 msec, flip angle of 12°, in-plane spatial resolution of 1.0 × 1.0 mm, thickness of 1.0 mm, temporal resolution of approximately 90 seconds, and axial orientation.

Statistical Analysis

Preoperative MRI use rates for the entire cohort as well as for racial subgroups were determined. Positive margin rates were calculated for all patients, the MRI and no-MRI cohorts, and the Black and White racial subgroups within each of the two cohorts. Differences between rates across cohorts and racial subgroups were compared and tested for significance using the χ² test of independence with a type 1 error of 0.05. All statistical analysis was conducted using R software version 4.2 (R Project for Statistical Computing).

Results

Patient Cohort Characteristics

The final patient cohort consisted of 28 384 female patients (mean age, 56 years ± 13 [SD]), of which 90.36% (n = 25 647) self-identified as White and 9.64% (n = 2737) self-identified as Black (Table 1). MRI use status was available for all patients, and surgical margin data following partial mastectomy were available for 55.89% (n = 15 863) of patients (Fig 1).

Table 1:

Demographics of the Study Cohort

Figure 1:

Flowchart of patients included in the study.

Preoperative MRI Use Rate among Racial Subgroups

Of the 28 384 patients in the final patient cohort, 32.78% (n = 9305) underwent preoperative breast MRI while 67.22% (n = 19 079) did not. Black patients had a significantly lower MRI use rate compared with White patients (29.85% [817 of 2737] vs 33.10% [8488 of 25 647], respectively; P < .001) (Table 2). Figure 2 shows examples of preoperative MR images in White and Black patients with positive or negative margins at surgery.

Table 2:

Rates of Preoperative MRI Use Stratified by Racial Subgroup and Rates of Positive Surgical Margins Stratified by Racial Subgroup and Preoperative MRI Use

Figure 2:

Examples of MRI in the patient cohort. (A) Preoperative contrast-enhanced axial MR image in a White patient (70-year-old female) who had negative margins at surgery, showing an enhancing mass in the right breast, upper outer quadrant, posterior depth. (B) Preoperative contrast-enhanced axial MR image in a Black patient (60-year-old female) who had negative margins at surgery, showing focal nonmass enhancement in the left breast, upper inner quadrant, posterior depth. (C) Preoperative contrast-enhanced axial MR image in a White patient (73-year-old female) who had positive margins at surgery, showing two enhancing masses in the right outer breast, anterior and posterior depths. (D) Preoperative contrast-enhanced axial MR subtraction image in a Black patient (44-year-old female) who had positive margins at surgery, showing an irregular enhancing mass with an in situ biopsy marker denoting site of carcinoma.

Positive Surgical Margin Rate among MRI Use and Racial Subgroups

The overall positive margin rate was 4.24% (672 of 15 863). Patients who underwent preoperative MRI had a significantly lower positive surgical margin rate compared with those who did not (2.90% [133 of 4592] vs 4.78% [539 of 11 271], respectively; P = .03) (Table 2). Within the MRI cohort, there was no evidence of a difference in positive surgical margin rates between Black and White racial subgroups (3.17% [12 of 379] vs 2.87% [121 of 4213], respectively; P = .90). Within the no-MRI cohort, Black patients had a significantly higher positive surgical margin rate compared with White patients (6.17% [68 of 1103] vs 4.63% [471 of 10 168], respectively; P = .03) (Table 2).

Discussion

In this study, we evaluated racial disparities in preoperative breast MRI use and surgical margin outcomes among patients with recently diagnosed breast cancer. Our results showed that Black patients were significantly less likely than White patients to undergo preoperative breast MRI (30% vs 33%, respectively; P < .001). Overall, patients who underwent preoperative MRI had a lower rate of positive surgical margins (2.90% vs 4.78%, P = .03), with no evidence of a difference in positive surgical margin rates when results were stratified by racial subgroups (3.17% vs 2.87%, P = .90). However, among patients who did not undergo preoperative MRI, Black patients were disproportionately more likely to have positive surgical margins compared with White patients (6.17% vs 4.63%, respectively; P = .03).

This study builds upon the growing literature demonstrating racial disparities in breast cancer care and treatment. Black female individuals are more likely to die of breast cancer at any age, receive a breast cancer diagnosis at a younger age, and have more aggressive disease; yet, they are less likely to obtain adequate treatment (1). Our study is consistent with previous studies reporting that Black patients are less likely than White patients to undergo preoperative MRI for breast cancer (3–5) and more likely to have positive surgical margins following breast-conserving surgery when controlling for clinicopathologic features that correlate with positive surgical margins (6). To our knowledge, there are no studies to date examining these trends in the same patient cohort or investigating positive surgical margin rates among racial subgroups stratified by preoperative MRI use. This study demonstrates evidence of racial disparities in preoperative breast MRI use, as well as increased positive surgical margin rates in Black patients compared with White patients who do not undergo preoperative MRI.

While it is an important observation that a lower percentage of Black patients with breast cancer underwent breast MRI preoperatively compared with White patients, our study did not explore the underlying reasons for this disparity. The causes of these disparities are likely multifactorial, given the many factors that contribute to ordering and undergoing a breast MRI examination. The ordering of a breast MRI examination is highly dependent on breast surgeon preference. Our cancer center has a large surgical department, and some of our cohort initially underwent breast MRI examinations performed at outside facilities prior to presenting to our center. Other potential causes for this disparity include barriers to undergoing a breast MRI examination, such as geographic access to a breast MRI facility, health insurance, financial means to support out-of-pocket costs, and breast MRI offered at hours that accommodate work schedules (5,8,9). Patient preference and education level may play a role in advocating for breast MRI preoperatively. The effect of unconscious biases, which unintentionally alter interactions and decision-making, on health care outcomes may also contribute to these trends (10). Given that White and Black patients in our study were not matched for clinical factors that may influence whether breast MRI is performed, such as individual surgeon preference, tumor type, or breast density, further investigation is needed to better understand this disparity.

We note that only 32.78% of all patients with breast cancer in our study underwent preoperative breast MRI, and overall MRI use was low for both White and Black patients (33.10% vs 29.85%, respectively) in comparison with use rates at other institutions, which have ranged from 30% to 75% in recent reports (11–15). This broad range may reflect the ongoing challenges in identifying which patients will benefit from preoperative MRI, leading to MRI use rates that are often surgeon- or institution-dependent. It is also recognized that throughout the span of our study, guidelines and practice patterns for preoperative MRI use have evolved. Currently, the American Society of Breast Surgeons recommends preoperative breast MRI when determining the extent of cancer in patients with proven cancer who have associated clinical or indeterminate imaging findings (16). These guidelines reference a meta-analysis that found no proven criteria for a specific patient population that would most benefit from routine preoperative breast MRI (17), highlighting the ambiguity regarding the indications for preoperative MRI in patients with breast cancer.

One decade-long retrospective study at a large regional cancer center investigated the value of conventional patient and tumor features most often used by clinicians for decision-making regarding preoperative breast MRI use (eg, age, breast density, tumor phenotype) (18). Of note, none of these factors were associated with the finding of additional cancer, while breast density and younger age were associated with false-positive outcomes (no additional cancer found). These findings suggest that the predictive value of such features may not be as accurate as previously thought and that other factors should be investigated (18). Some studies have demonstrated that preoperative breast MRI is particularly useful in patients with mammographically occult breast cancer, invasive lobular carcinoma, triple-negative disease, dense breast tissue, luminal B and HER2 molecular subtype disease, and intermediate- or high-grade ductal carcinoma in situ (19–25). However, none of these studies evaluated primary outcomes by race or ethnicity. Thus, further investigation is needed to study these potential outcome differences and to determine factors unique to racial and ethnic minority individuals that may be used to more appropriately determine whether preoperative MRI is indicated. Some studies have also shown that preoperative MRI is associated with reduced risk of positive margins, reduced reoperation rates, and lower cancer recurrence rates (26–28). As such, inconsistencies in preoperative MRI use between racial subgroups may be associated with downstream clinical consequences that exacerbate racial health disparities. This study thus provides further justification to investigate and mitigate disparities in breast MRI use and improve poor surgical outcomes among racial and ethnic minority patients, potentially through addressing barriers patients may face in obtaining a breast MRI examination or exploring subjective practice patterns in MRI triaging.

Of note, there has been conflicting evidence on whether preoperative MRI in patients with breast cancer is associated with margin status. A recent meta-analysis showed an association between preoperative MRI and a lower risk of positive resection margins, reoperations, and recurrence in patients with breast cancer (27). However, other studies of patients with breast cancer showed that preoperative MRI was not an independent significant factor in predicting margin status (29,30).

This study had several limitations. The patient cohort was obtained from a single tertiary cancer care center with a larger White patient population than Black, allowing for potential selection bias and limiting the generalizability of our findings to the wider U.S. population. For instance, while this is a large patient cohort overall (28 384 patients with recently diagnosed breast cancer), there were only 12 patients who were Black, underwent preoperative MRI, and had positive surgical margins, which may have limited the power of our results. Additionally, race and ethnicity data were self-reported, introducing the potential for subjective data. Potential confounding variables that were not collected in this study but could be included in future work include time to treatment, family history of cancer, comorbid medical conditions, education, income, insurance status, and zip code.

The impact and clinical applicability of this work has far-reaching ramifications, given that nearly 30 000 Black female individuals are diagnosed with breast cancer annually in the United States (31). Our study adds to the growing literature showing racial disparities in preoperative breast MRI use and demonstrates increased positive margin rates among Black patients when compared with White patients with breast cancer who do not undergo MRI. These findings demonstrate the need to further investigate the causes of and solutions to racial disparities in breast MRI use.