https://orcid.org/0000-0003-4835-1478
Abstract
Breast MRI is the most sensitive imaging modality for the assessment of newly diagnosed breast cancer extent and can detect additional mammographically and clinically occult breast cancers in the ipsilateral and contralateral breasts. Nonetheless, appropriate use of breast MRI in the setting of newly diagnosed breast cancer remains debated. Though highly sensitive, MRI is less specific and may result in false positives and overestimation of disease when MRI findings are not biopsied prior to surgical excision. Furthermore, improved anatomic depiction of breast cancer on MRI has not consistently translated to improved clinical outcomes, such as lower rates of re-excision or breast cancer recurrence, though there is a paucity of well-designed studies examining these issues. In addition, current treatment paradigms have been developed in the absence of this more accurate depiction of disease span, which likely has limited the value of MRI. These issues have led to inconsistent and variable utilization of preoperative MRI across practice settings and providers. In this review, we discuss the history of breast MRI and its current use and recommendations with a focus on the preoperative setting. We review the evidence surrounding the use of preoperative MRI in the evaluation of breast malignancies and discuss the data on breast MRI in the setting of specific patient factors often used to determine breast MRI eligibility, such as age, index tumor phenotype, and breast density. Finally, we review the impact of breast MRI on surgical outcomes (re-excision and mastectomy rates) and long-term breast recurrence and survival outcomes.
Preoperative breast MRI performed for extent of disease is more commonly obtained in women of younger age, of higher breast cancer risk, and with invasive lobular histology.
In certain clinical scenarios, breast MRI may contribute to decreased surgical re-excision rates, though data on long-term local recurrence and survival outcomes between women who underwent MRI and those who did not remain unclear.
Introduction
Early studies in the 2000s demonstrated that preoperative breast MRI (pMRI) identifies a significant number of mammographically and clinically occult additional breast cancers in the ipsilateral and contralateral breast, which led to an increase in its use (1–4). However, many retrospective studies, and a few randomized controlled studies, have shed doubt on the belief that the detection of additional disease burden provides a meaningful clinical benefit, such as improved surgical outcomes or recurrence rates in the setting of current treatment paradigms (5–7). Some have also suggested that pMRI may increase mastectomy rates unnecessarily (7–10). A common thread connecting most of these data has been the inability to control for important factors, such as age, socioeconomic status, surgeon approach and ability to incorporate pMRI findings, MRI interpretation standards, and tumor subtypes. These challenges are due to flaws in prospective trial designs and limitations when evaluating retrospective data. Nonetheless, given the costs associated with MRI examinations and the uncertain risk-benefit ratio, it remains important to continually evaluate the available evidence and identify gaps that can be addressed with future research so that judicious use of pMRI can be achieved.
Current Recommendations and Utilization of pMRI
Current evidence-based indications for breast MRI include supplemental screening of women at high risk for breast cancer, further evaluation of newly diagnosed breast cancer (ie, pMRI), evaluation of silicone breast implant integrity, evaluation of treatment response to neoadjuvant chemotherapy, and selected problem-solving scenarios (eg, evaluation of suspicious nipple discharge) (11,12). However, the precise population of patients for whom pMRI is appropriate remains controversial, with recommended use scenarios ranging widely; Table 1 lists the current subgroup guidelines for when pMRI may be recommended. It is important to recognize that these guidelines are based on expert consensus opinion rather than data indicating clear benefit in the recommended patient populations.
Preoperative Breast MRI Subgroup Recommendations in the Setting of New Breast Cancer
| Subgroup | Preoperative Breast MRI Recommendation |
|---|---|
| National Comprehensive Cancer Network (2022) (12) | DCIS: controversial. Select circumstances in which additional information is warranted during the initial workup Invasive cancer: optional |
| American College of Radiology (Practice Parameter) (2018) (13) | Useful for extent of disease and occult ipsilateral and contralateral tumors Evaluation of invasion deep to pectoralis fascia |
| National Institute for Health and Care Excellence (2018) (14) | Discrepancy in EOD between clinical exam, mammogram, or US Breast density precludes accurate mammogram assessment Invasive lobular carcinoma with breast conservation surgery considered |
| American Society of Breast Surgeons (2017) (15) | Significant discrepancies in EOD between clinical exam, mammogram, or US Extremely dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Imaging (2008) (16) | Screening of contralateral breast (all patients) Dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Cancer Specialists (2010) (17) | High breast cancer risk <60 years of age with >1-cm size difference between mammogram and US size Invasive lobular carcinoma Accelerated partial breast irradiation candidates based on clinical exam and conventional imaging |
| German Gynecological Oncology Group (2018) (18) | Optional in following circumstances: High breast cancer risk Reduced lesion detection on mammography (dense breasts) and US Nipple involvement Invasive lobular carcinoma Multicentric disease suspected on conventional imaging |
| Subgroup | Preoperative Breast MRI Recommendation |
|---|---|
| National Comprehensive Cancer Network (2022) (12) | DCIS: controversial. Select circumstances in which additional information is warranted during the initial workup Invasive cancer: optional |
| American College of Radiology (Practice Parameter) (2018) (13) | Useful for extent of disease and occult ipsilateral and contralateral tumors Evaluation of invasion deep to pectoralis fascia |
| National Institute for Health and Care Excellence (2018) (14) | Discrepancy in EOD between clinical exam, mammogram, or US Breast density precludes accurate mammogram assessment Invasive lobular carcinoma with breast conservation surgery considered |
| American Society of Breast Surgeons (2017) (15) | Significant discrepancies in EOD between clinical exam, mammogram, or US Extremely dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Imaging (2008) (16) | Screening of contralateral breast (all patients) Dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Cancer Specialists (2010) (17) | High breast cancer risk <60 years of age with >1-cm size difference between mammogram and US size Invasive lobular carcinoma Accelerated partial breast irradiation candidates based on clinical exam and conventional imaging |
| German Gynecological Oncology Group (2018) (18) | Optional in following circumstances: High breast cancer risk Reduced lesion detection on mammography (dense breasts) and US Nipple involvement Invasive lobular carcinoma Multicentric disease suspected on conventional imaging |
Abbreviations: DCIS, ductal carcinoma in situ; EOD, extent of disease.
Preoperative Breast MRI Subgroup Recommendations in the Setting of New Breast Cancer
| Subgroup | Preoperative Breast MRI Recommendation |
|---|---|
| National Comprehensive Cancer Network (2022) (12) | DCIS: controversial. Select circumstances in which additional information is warranted during the initial workup Invasive cancer: optional |
| American College of Radiology (Practice Parameter) (2018) (13) | Useful for extent of disease and occult ipsilateral and contralateral tumors Evaluation of invasion deep to pectoralis fascia |
| National Institute for Health and Care Excellence (2018) (14) | Discrepancy in EOD between clinical exam, mammogram, or US Breast density precludes accurate mammogram assessment Invasive lobular carcinoma with breast conservation surgery considered |
| American Society of Breast Surgeons (2017) (15) | Significant discrepancies in EOD between clinical exam, mammogram, or US Extremely dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Imaging (2008) (16) | Screening of contralateral breast (all patients) Dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Cancer Specialists (2010) (17) | High breast cancer risk <60 years of age with >1-cm size difference between mammogram and US size Invasive lobular carcinoma Accelerated partial breast irradiation candidates based on clinical exam and conventional imaging |
| German Gynecological Oncology Group (2018) (18) | Optional in following circumstances: High breast cancer risk Reduced lesion detection on mammography (dense breasts) and US Nipple involvement Invasive lobular carcinoma Multicentric disease suspected on conventional imaging |
| Subgroup | Preoperative Breast MRI Recommendation |
|---|---|
| National Comprehensive Cancer Network (2022) (12) | DCIS: controversial. Select circumstances in which additional information is warranted during the initial workup Invasive cancer: optional |
| American College of Radiology (Practice Parameter) (2018) (13) | Useful for extent of disease and occult ipsilateral and contralateral tumors Evaluation of invasion deep to pectoralis fascia |
| National Institute for Health and Care Excellence (2018) (14) | Discrepancy in EOD between clinical exam, mammogram, or US Breast density precludes accurate mammogram assessment Invasive lobular carcinoma with breast conservation surgery considered |
| American Society of Breast Surgeons (2017) (15) | Significant discrepancies in EOD between clinical exam, mammogram, or US Extremely dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Imaging (2008) (16) | Screening of contralateral breast (all patients) Dense breast tissue Invasive lobular carcinoma |
| European Society of Breast Cancer Specialists (2010) (17) | High breast cancer risk <60 years of age with >1-cm size difference between mammogram and US size Invasive lobular carcinoma Accelerated partial breast irradiation candidates based on clinical exam and conventional imaging |
| German Gynecological Oncology Group (2018) (18) | Optional in following circumstances: High breast cancer risk Reduced lesion detection on mammography (dense breasts) and US Nipple involvement Invasive lobular carcinoma Multicentric disease suspected on conventional imaging |
Abbreviations: DCIS, ductal carcinoma in situ; EOD, extent of disease.
Among this landscape of conflicting evidence and lack of consensus, pMRI is inconsistently ordered in surgical practice. Though the American Society of Breast Surgeons does not recommend routinely obtaining MRI in patients with new breast cancer, a survey of breast surgeons indicated that 41% used pMRI in daily practice (19). A 2017 study of general and breast surgeons demonstrated that pMRI is most often ordered for patients with higher breast density, family history of breast cancer, mammographically occult cancer, and invasive lobular carcinoma (ILC) (20). Additional factors to consider included radiologist recommendation for pMRI and young patient age (<40 years).
The Multicenter International Prospective Analysis (MIPA) study was a large European observational study providing important insights on the practice of obtaining pMRI for breast cancer in clinical practice at 27 centers (21). In this nonrandomized study of women aged 18 to 80 years, it was found that pMRI was obtained more often for younger patients who were pre- or perimenopausal, women with dense breasts, women planned for mastectomy, women with ILC, and women with tumor size >2 cm. Re-excision rates were lower among the pMRI group, though mastectomy rates were higher. Preoperative MRI was used to help confirm need for mastectomy, with an absolute 11% increase in mastectomy rate (9% when accounting for patient preference), mainly due to additional malignant findings.
Overall Evidence for pMRI
Accuracy to Assess Pathological Extent of Index Malignancy
In the setting of a new breast cancer diagnosis, studies have demonstrated that pMRI estimates final pathological extent of disease (EOD) more accurately than mammography, US, and clinical examination (Figure 1) (1,22). Preoperative MRI is more accurate than conventional imaging for estimation of ILC tumor size when compared with pathological extent (23,24). Rominger et al (25) and Mann et al (26) have also found that there is both under- and overestimation of tumor size with MRI, and correlation of MRI and pathology extent is limited by both imaging and pathological measurement techniques, pathological reporting, and unclear thresholds as to what constitutes a meaningful discrepancy in size.
Forty-year-old woman with screen-detected right invasive ductal carcinoma and ductal carcinoma in situ (DCIS). Diagnostic spot magnification image (A) shows a right breast asymmetry with calcifications in the superior breast posteriorly measuring 18 mm (circle). A targeted right breast US image (B) shows an irregular mass with indistinct margins, measuring 12 mm (arrow), corresponding with the mammographic asymmetry. Maximum intensity projection image (C) from preoperative MRI shows segmental non-mass enhancement (NME) in the right breast (arrows), encompassing the entire upper inner quadrant and extending into the upper outer quadrant, measuring 52 mm in the anteroposterior dimension and within 7 mm of the nipple. Biopsy of the anterior aspect of the NME confirmed DCIS, which was mammographically occult.
Detection of Additional Occult Malignancies
It has been firmly established that pMRI identifies additional cancers that are mammographically and clinically occult in both the ipsilateral and contralateral breasts (27–29). A meta-analysis of 50 studies (N = 10 811 women) established that breast MRI detects additional ipsilateral and contralateral disease in 20% and 5.5% of women, respectively (Figure 2) (30). Still, moderate specificity remains a challenge with MRI, with a 33% false positive rate for detection of ipsilateral breast cancers and 63% for contralateral disease (Figure 3) (30,31). This false positive rate has decreased, however, with higher magnetic field strength (≥1.5T) resulting in potential reductions in false positive findings (16% for the ipsilateral breast, 21% for the contralateral breast) (30). Regardless, pathological verification of MRI findings should be performed prior to modifying surgical recommendations to limit the potential for performing unnecessary surgeries because of false positive MRI findings.
Mammographically occult disease seen on preoperative MRI in a 69-year-old woman with newly diagnosed left breast invasive ductal carcinoma (IDC) presenting as a 28-mm mass at the 11-o’clock position, upper inner quadrant. Maximum intensity projection image from a preoperative MRI performed for extent of disease shows the primary mass (arrowhead), as well as a 4-mm mass in the upper outer quadrant at the 2-o’clock position (circle) and a 7-mm mass in the lower outer quadrant at the 4-o’clock position (arrow). Biopsies of these masses showed additional IDC at both sites.
False positive contralateral breast finding on preoperative MRI in a 46-year-old woman with screen-detected right breast invasive ductal carcinoma and ductal carcinoma in situ. Preoperative MRI showed segmental non-mass enhancement (NME) in the right breast posteriorly (circle) at the site of known malignancy. In the left breast, there is focal NME at the 3-o’clock position, measuring 14 mm (arrow). MRI-guided biopsy of this area showed pseudoangiomatous stromal hyperplasia, which was benign and concordant.
Impact on Breast Surgical Outcomes—Re-excision and Mastectomy Rates
To date, and to our knowledge, there have been five multicenter randomized controlled trials evaluating the value of pMRI in decreasing re-excision rates in breast cancer patients, four of which have reported results (Table 2). The Comparative Effectiveness of MRI in Breast Cancer (COMICE) trial randomized women with a diagnosis of breast cancer who were planning to undergo breast conservation surgery (BCS) based on conventional imaging and clinical exam to receive pMRI versus no pMRI (5). The investigators found no difference in re-excision rates with pMRI compared to conventional imaging alone; however, there were several important limitations. Specifically, this study did not require radiologists to have breast MRI interpretation experience, nor did it require that all MRI-detected findings be sampled prior to surgery. Furthermore, patients who were converted to mastectomy based on false positive MRI findings were included as “re-operations” without accounting for the lack of MRI-guided biopsy availability and the appropriateness of the operation. In some cases, patients underwent surgery prior to pMRI interpretation.
Multicenter Randomized Clinical Trials of Preoperative MRI on Surgical Outcomes
| Trial and Location | Patient Recruitment Rates and Number of Sites Included | Population | Final Number of Patients Included and Randomization | Results | Limitations |
|---|---|---|---|---|---|
| Comparative Effectiveness of MRI in Breast Cancer (COMICE) trial (5), United Kingdom | 2002–2007; published 2010; 45 centers | Women aged 18 years and older with biopsy-proven breast cancer planned for wide local excision after clinical exam and imaging assessment (mammogram ± US, pathology) | Total, 1623: 816 pMRI, 807 no pMRI Randomized by age (<50 or ≥50 years) or breast density (fatty or other) | pMRI did not decrease re-excision rates compared to conventional imaging 7% of patients with pMRI had mastectomy, versus 1% without pMRI | Lack of radiology interpretation experience Sampling of all pMRI-detected findings prior to surgery was not required Patients underwent surgery prior to pMRI interpretation |
| MR Mammography of Nonpalpable Breast Tumors (MONET) trial (6), the Netherlands | 2006–2008;published 2011; three centers; two large community teaching hospitals, one academic | Women aged 18–75 years with a nonpalpable BI-RADS category 3–5 lesion detected on mammography who were referred for biopsy | Total, 418: 207 MRI, 211 no MRI Randomized by independent trial center, stratified by site once inclusion criteria met | No significant difference in number and type of initial surgery (BCS versus mastectomy) between MRI and no MRI groups MRI was associated with higher re-excision rate—18/53 (34%) versus 6/50 (12%) without MRI (P = 0.008) Re-excision rate to mastectomy was lower with MRI though not significant, 6/53 (11%) versus 8/50 (14%) without MRI (P = 0.489) | Not a true pMRI trial as MRI was performed before biopsy of findings Surgical volume of tissue excised was less in MRI versus no MRI group, in particular in patients with no suspicious MRI findings Relatively high percentage of patients (60%) included with suspicious calcifications as the only finding Technical limitations with inhomogeneous fat suppression on MRI images, limiting interpretation at the beginning of the study |
| Preoperative MRI of the Breast (POMB) trial (32), Sweden | 2007–2011; published 2014; three centers | Women aged <56 years with a new diagnosis of breast cancer | Total, 440: 220 pMRI, 220 no pMRI Most randomized by computer algorithm after diagnosis of cancer but before tumor board discussion; some were after tumor board | pMRI contributed to altered treatment plan in 18% of women, with 15% of patients converting initial surgery to mastectomy pMRI decreased re-excision rate to 5%, versus 15% without pMRI (P < 0.001) pMRI did not increase overall mastectomy rates | Patients randomized to pMRI had higher planned initial BCS (70%) compared with no pMRI (60%) Not all pMRI-detected lesions were required to be biopsied and MRI biopsy capability was not available |
| Preoperative Breast MRI in Women With Local DCIS to Optimize Surgical Outcomes (IRCIS) trial (33), France | 2010–2014; published 2019; 10 centers | Women aged 18–80 years with biopsy-proven pure DCIS corresponding to unifocal group of calcifications or mass <30 mm in size on mammogram and/or US | Total, 352: 178 pMRI, 174 no pMRI Randomized by block randomization, stratified by center | Lower re-excision rate with pMRI, 20%, versus 27% without pMRI, though without statistical significance (P = 0.13) Initial mastectomy rate higher (9%) in pMRI, versus 4% in no pMRI (P = 0.06) due to larger or residual lesions on pMRI and confirmed histology. Additional 15 mastectomies performed as second surgery (9%) in pMRI group versus 23 (13%) in no pMRI group. Overall, no difference in mastectomy rates, pMRI 18% versus without pMRI 17% | Majority (82%) of DCIS appeared as non-mass enhancement on pMRI, with some findings too subtle or nonspecific for surgeons to integrate into surgical plan |
| Trial and Location | Patient Recruitment Rates and Number of Sites Included | Population | Final Number of Patients Included and Randomization | Results | Limitations |
|---|---|---|---|---|---|
| Comparative Effectiveness of MRI in Breast Cancer (COMICE) trial (5), United Kingdom | 2002–2007; published 2010; 45 centers | Women aged 18 years and older with biopsy-proven breast cancer planned for wide local excision after clinical exam and imaging assessment (mammogram ± US, pathology) | Total, 1623: 816 pMRI, 807 no pMRI Randomized by age (<50 or ≥50 years) or breast density (fatty or other) | pMRI did not decrease re-excision rates compared to conventional imaging 7% of patients with pMRI had mastectomy, versus 1% without pMRI | Lack of radiology interpretation experience Sampling of all pMRI-detected findings prior to surgery was not required Patients underwent surgery prior to pMRI interpretation |
| MR Mammography of Nonpalpable Breast Tumors (MONET) trial (6), the Netherlands | 2006–2008;published 2011; three centers; two large community teaching hospitals, one academic | Women aged 18–75 years with a nonpalpable BI-RADS category 3–5 lesion detected on mammography who were referred for biopsy | Total, 418: 207 MRI, 211 no MRI Randomized by independent trial center, stratified by site once inclusion criteria met | No significant difference in number and type of initial surgery (BCS versus mastectomy) between MRI and no MRI groups MRI was associated with higher re-excision rate—18/53 (34%) versus 6/50 (12%) without MRI (P = 0.008) Re-excision rate to mastectomy was lower with MRI though not significant, 6/53 (11%) versus 8/50 (14%) without MRI (P = 0.489) | Not a true pMRI trial as MRI was performed before biopsy of findings Surgical volume of tissue excised was less in MRI versus no MRI group, in particular in patients with no suspicious MRI findings Relatively high percentage of patients (60%) included with suspicious calcifications as the only finding Technical limitations with inhomogeneous fat suppression on MRI images, limiting interpretation at the beginning of the study |
| Preoperative MRI of the Breast (POMB) trial (32), Sweden | 2007–2011; published 2014; three centers | Women aged <56 years with a new diagnosis of breast cancer | Total, 440: 220 pMRI, 220 no pMRI Most randomized by computer algorithm after diagnosis of cancer but before tumor board discussion; some were after tumor board | pMRI contributed to altered treatment plan in 18% of women, with 15% of patients converting initial surgery to mastectomy pMRI decreased re-excision rate to 5%, versus 15% without pMRI (P < 0.001) pMRI did not increase overall mastectomy rates | Patients randomized to pMRI had higher planned initial BCS (70%) compared with no pMRI (60%) Not all pMRI-detected lesions were required to be biopsied and MRI biopsy capability was not available |
| Preoperative Breast MRI in Women With Local DCIS to Optimize Surgical Outcomes (IRCIS) trial (33), France | 2010–2014; published 2019; 10 centers | Women aged 18–80 years with biopsy-proven pure DCIS corresponding to unifocal group of calcifications or mass <30 mm in size on mammogram and/or US | Total, 352: 178 pMRI, 174 no pMRI Randomized by block randomization, stratified by center | Lower re-excision rate with pMRI, 20%, versus 27% without pMRI, though without statistical significance (P = 0.13) Initial mastectomy rate higher (9%) in pMRI, versus 4% in no pMRI (P = 0.06) due to larger or residual lesions on pMRI and confirmed histology. Additional 15 mastectomies performed as second surgery (9%) in pMRI group versus 23 (13%) in no pMRI group. Overall, no difference in mastectomy rates, pMRI 18% versus without pMRI 17% | Majority (82%) of DCIS appeared as non-mass enhancement on pMRI, with some findings too subtle or nonspecific for surgeons to integrate into surgical plan |
Abbreviations: BCS, breast-conserving surgery; DCIS, ductal carcinoma in situ; pMRI, preoperative breast MRI.
Multicenter Randomized Clinical Trials of Preoperative MRI on Surgical Outcomes
| Trial and Location | Patient Recruitment Rates and Number of Sites Included | Population | Final Number of Patients Included and Randomization | Results | Limitations |
|---|---|---|---|---|---|
| Comparative Effectiveness of MRI in Breast Cancer (COMICE) trial (5), United Kingdom | 2002–2007; published 2010; 45 centers | Women aged 18 years and older with biopsy-proven breast cancer planned for wide local excision after clinical exam and imaging assessment (mammogram ± US, pathology) | Total, 1623: 816 pMRI, 807 no pMRI Randomized by age (<50 or ≥50 years) or breast density (fatty or other) | pMRI did not decrease re-excision rates compared to conventional imaging 7% of patients with pMRI had mastectomy, versus 1% without pMRI | Lack of radiology interpretation experience Sampling of all pMRI-detected findings prior to surgery was not required Patients underwent surgery prior to pMRI interpretation |
| MR Mammography of Nonpalpable Breast Tumors (MONET) trial (6), the Netherlands | 2006–2008;published 2011; three centers; two large community teaching hospitals, one academic | Women aged 18–75 years with a nonpalpable BI-RADS category 3–5 lesion detected on mammography who were referred for biopsy | Total, 418: 207 MRI, 211 no MRI Randomized by independent trial center, stratified by site once inclusion criteria met | No significant difference in number and type of initial surgery (BCS versus mastectomy) between MRI and no MRI groups MRI was associated with higher re-excision rate—18/53 (34%) versus 6/50 (12%) without MRI (P = 0.008) Re-excision rate to mastectomy was lower with MRI though not significant, 6/53 (11%) versus 8/50 (14%) without MRI (P = 0.489) | Not a true pMRI trial as MRI was performed before biopsy of findings Surgical volume of tissue excised was less in MRI versus no MRI group, in particular in patients with no suspicious MRI findings Relatively high percentage of patients (60%) included with suspicious calcifications as the only finding Technical limitations with inhomogeneous fat suppression on MRI images, limiting interpretation at the beginning of the study |
| Preoperative MRI of the Breast (POMB) trial (32), Sweden | 2007–2011; published 2014; three centers | Women aged <56 years with a new diagnosis of breast cancer | Total, 440: 220 pMRI, 220 no pMRI Most randomized by computer algorithm after diagnosis of cancer but before tumor board discussion; some were after tumor board | pMRI contributed to altered treatment plan in 18% of women, with 15% of patients converting initial surgery to mastectomy pMRI decreased re-excision rate to 5%, versus 15% without pMRI (P < 0.001) pMRI did not increase overall mastectomy rates | Patients randomized to pMRI had higher planned initial BCS (70%) compared with no pMRI (60%) Not all pMRI-detected lesions were required to be biopsied and MRI biopsy capability was not available |
| Preoperative Breast MRI in Women With Local DCIS to Optimize Surgical Outcomes (IRCIS) trial (33), France | 2010–2014; published 2019; 10 centers | Women aged 18–80 years with biopsy-proven pure DCIS corresponding to unifocal group of calcifications or mass <30 mm in size on mammogram and/or US | Total, 352: 178 pMRI, 174 no pMRI Randomized by block randomization, stratified by center | Lower re-excision rate with pMRI, 20%, versus 27% without pMRI, though without statistical significance (P = 0.13) Initial mastectomy rate higher (9%) in pMRI, versus 4% in no pMRI (P = 0.06) due to larger or residual lesions on pMRI and confirmed histology. Additional 15 mastectomies performed as second surgery (9%) in pMRI group versus 23 (13%) in no pMRI group. Overall, no difference in mastectomy rates, pMRI 18% versus without pMRI 17% | Majority (82%) of DCIS appeared as non-mass enhancement on pMRI, with some findings too subtle or nonspecific for surgeons to integrate into surgical plan |
| Trial and Location | Patient Recruitment Rates and Number of Sites Included | Population | Final Number of Patients Included and Randomization | Results | Limitations |
|---|---|---|---|---|---|
| Comparative Effectiveness of MRI in Breast Cancer (COMICE) trial (5), United Kingdom | 2002–2007; published 2010; 45 centers | Women aged 18 years and older with biopsy-proven breast cancer planned for wide local excision after clinical exam and imaging assessment (mammogram ± US, pathology) | Total, 1623: 816 pMRI, 807 no pMRI Randomized by age (<50 or ≥50 years) or breast density (fatty or other) | pMRI did not decrease re-excision rates compared to conventional imaging 7% of patients with pMRI had mastectomy, versus 1% without pMRI | Lack of radiology interpretation experience Sampling of all pMRI-detected findings prior to surgery was not required Patients underwent surgery prior to pMRI interpretation |
| MR Mammography of Nonpalpable Breast Tumors (MONET) trial (6), the Netherlands | 2006–2008;published 2011; three centers; two large community teaching hospitals, one academic | Women aged 18–75 years with a nonpalpable BI-RADS category 3–5 lesion detected on mammography who were referred for biopsy | Total, 418: 207 MRI, 211 no MRI Randomized by independent trial center, stratified by site once inclusion criteria met | No significant difference in number and type of initial surgery (BCS versus mastectomy) between MRI and no MRI groups MRI was associated with higher re-excision rate—18/53 (34%) versus 6/50 (12%) without MRI (P = 0.008) Re-excision rate to mastectomy was lower with MRI though not significant, 6/53 (11%) versus 8/50 (14%) without MRI (P = 0.489) | Not a true pMRI trial as MRI was performed before biopsy of findings Surgical volume of tissue excised was less in MRI versus no MRI group, in particular in patients with no suspicious MRI findings Relatively high percentage of patients (60%) included with suspicious calcifications as the only finding Technical limitations with inhomogeneous fat suppression on MRI images, limiting interpretation at the beginning of the study |
| Preoperative MRI of the Breast (POMB) trial (32), Sweden | 2007–2011; published 2014; three centers | Women aged <56 years with a new diagnosis of breast cancer | Total, 440: 220 pMRI, 220 no pMRI Most randomized by computer algorithm after diagnosis of cancer but before tumor board discussion; some were after tumor board | pMRI contributed to altered treatment plan in 18% of women, with 15% of patients converting initial surgery to mastectomy pMRI decreased re-excision rate to 5%, versus 15% without pMRI (P < 0.001) pMRI did not increase overall mastectomy rates | Patients randomized to pMRI had higher planned initial BCS (70%) compared with no pMRI (60%) Not all pMRI-detected lesions were required to be biopsied and MRI biopsy capability was not available |
| Preoperative Breast MRI in Women With Local DCIS to Optimize Surgical Outcomes (IRCIS) trial (33), France | 2010–2014; published 2019; 10 centers | Women aged 18–80 years with biopsy-proven pure DCIS corresponding to unifocal group of calcifications or mass <30 mm in size on mammogram and/or US | Total, 352: 178 pMRI, 174 no pMRI Randomized by block randomization, stratified by center | Lower re-excision rate with pMRI, 20%, versus 27% without pMRI, though without statistical significance (P = 0.13) Initial mastectomy rate higher (9%) in pMRI, versus 4% in no pMRI (P = 0.06) due to larger or residual lesions on pMRI and confirmed histology. Additional 15 mastectomies performed as second surgery (9%) in pMRI group versus 23 (13%) in no pMRI group. Overall, no difference in mastectomy rates, pMRI 18% versus without pMRI 17% | Majority (82%) of DCIS appeared as non-mass enhancement on pMRI, with some findings too subtle or nonspecific for surgeons to integrate into surgical plan |
Abbreviations: BCS, breast-conserving surgery; DCIS, ductal carcinoma in situ; pMRI, preoperative breast MRI.
The MR Mammography of Nonpalpable Breast Tumors (MONET) trial sought to determine the impact of MRI on the management of nonpalpable Breast Imaging Reporting and Data System (BI-RADS) category 3 to 5 lesions in 418 women (6,34). This study was not a true pMRI trial because MRI examinations were performed prior to knowing whether a lesion was malignant. Patients were assigned to either a control group, who had core-needle biopsy after conventional imaging (mammography and US), or an intervention group, who underwent core-needle biopsy after mammography, US, and breast MRI. When evaluating the subset of women who had a breast cancer, the trial found no impact of MRI on initial BCS rate and counterintuitively demonstrated MRI to be associated with a higher rate of tumor-positive surgical margins. A viable explanation for this paradoxical result is that the surgeons participating in the trial incorporated MRI information to potentially perform smaller excisions for improved cosmetic outcomes, a notion circumstantially supported by the fact that excision volumes were smaller in the MRI versus no-MRI cohort. In this study, there was no overall increase in conversion to mastectomy or overall mastectomy rate between women with MRI versus no MRI.
The Preoperative MRI of the Breast (POMB) trial included women at three sites who were <56 years old and had a new diagnosis of breast cancer (total enrollment = 440) (32). Half of the participants were randomized to the pMRI arm and the other half to the no pMRI arm. The results demonstrated that pMRI changed the treatment approach in 18% of women and significantly decreased the rate of re-excision without increasing overall final mastectomy rates. The overall final mastectomy rate was the same between both groups. Of note, 10 of the women who were randomized to the pMRI group did not undergo pMRI but were analyzed as such based on the intention-to-treat principle. Strengths of the study were clear definitions of what constituted altered tumor extent and inclusion of only younger patients; however, not all additional lesions identified on pMRI were required to undergo biopsy. In addition, no data on factors leading to conversion to mastectomy in the MRI group were provided. Furthermore, patients may have changed their mind about undergoing mastectomy without first having biopsy-proven additional foci of disease, and the impact of pMRI on patients’ decisions about treatment was not assessed.
The Evaluation of Diagnostic Performance of MRI ± Biopsy to Optimize Resection of Ductal Carcinoma In Situ Breast Cancer (IRCIS) trial randomized 352 patients with pure ductal carcinoma in situ (DCIS) to pMRI or no pMRI prior to surgery and found a lower re-excision rate in the pMRI arm, but this difference was not significant in the intention-to-treat analysis (33). Similar to the MONET trial, there was no overall increase in conversion to mastectomy or overall mastectomy rate between the two groups.
Results are awaited from the U.S.-based Alliance for Clinical Trials in Oncology 11104/Eastern Cooperative Oncology-American College of Radiology Imaging Network (ECOG-ACRIN) 6694 trial that randomized women with stage I or II human epidermal growth factor receptor-2 (HER2)-positive or triple-negative invasive cancers to pMRI or no MRI. A challenge of that study is that, over its course, the treatment paradigm of primary surgery shifted to a more common recommendation for neoadjuvant chemotherapy (35).
Several retrospective studies have provided support for the use of pMRI to improve surgical outcomes in some patient populations. Sung et al (36) performed a retrospective matched cohort study involving 174 women with early stage breast cancer who had pMRI and were treated with BCS. The control group with no pMRI was matched by age, tumor histology, stage, and surgeon. The two groups had similar rates of final negative margins, axillary lymph node involvement, and adjuvant therapy received; however, re-excision rates were lower in the pMRI group (29% vs. 45%; P = 0.02).
Kuhl et al (37) performed a prospective two-center study evaluating patients with DCIS components of invasive breast cancers in those who had a pMRI after a diagnosis of invasive breast cancer. The investigators specifically studied the diagnostic accuracy of pMRI versus mammography and breast US on pathological EOD and surgical outcomes (positive margins and mastectomy rates). The sensitivity of identifying DCIS on pMRI was significantly higher than for conventional imaging (89.4% vs. 36.7%; P < 0.0001), with more than half (51%) of DCIS cases only seen on pMRI. With regard to surgical outcomes, positive margin and mastectomy rates were overall very low (3.7% and 9.5%, respectively) compared with previously published data of 20%–40% and 25%–45%, respectively. Despite the promise of several of these individual studies, a meta-analysis of 19 studies of all breast cancer types showed higher odds of both ipsilateral and contralateral mastectomy associated with pMRI but no significant differences in positive margins or rate of re-excision (7).
Although rates of BCS with adjuvant whole breast radiation have, overall, increased since the 1990s, the past two decades have seen a rise in the frequency of mastectomy for the surgical treatment of breast cancer, including among those women eligible for BCS (38). Because this trend has paralleled the increasing utilization of breast MRI, several studies have posited that these rising mastectomy rates are due in part to pMRI use (9). A retrospective study of 13 097 women with stage 0 to III breast cancer demonstrated 32% higher odds of mastectomy compared with BCS among the 2217 women with pMRI. In addition, women with pMRI were more likely to undergo contralateral prophylactic mastectomy and breast reconstruction (9). Important limitations of this study were that initial eligibility for BCS, margin status, and whether pMRI directly led to a subsequent change in surgical management could not be determined. The study also did not account for patient and provider preferences.
Data suggesting that pMRI is associated with a rise in mastectomies are confounded by a general trend toward increasing mastectomy rates, particularly among younger women with higher breast density and a family history of breast cancer, or those with DCIS pathology (38–40). Given that the rise in mastectomy rates has paralleled the rise in pMRI use, it is plausible that the observed association of pMRI with increased mastectomies is at least in part due to its association with other known patient factors that correlate with mastectomy use (41,42).
Impact on Breast Cancer Recurrence and Survival Outcomes
To date, the literature has not clearly demonstrated improved longer-term outcomes in patients who receive pMRI versus those who do not. A retrospective study by Solin et al (43) in women with early stage invasive breast cancer or DCIS undergoing BCS with radiation therapy found no difference in eight-year overall survival, contralateral breast cancer, or freedom from distant metastases with pMRI. A similar retrospective study by Hwang et al (44) in patients with invasive breast cancer undergoing BCS with radiation therapy also found no association between pMRI and eight-year ipsilateral tumor recurrence. However, these findings may be due to unmatched patient and treatment factors between the two groups, with younger women, who are more likely to experience recurrence, overrepresented in the pMRI group. Furthermore, a meta-analysis by Houssami et al (45) showed that the addition of pMRI to conventional imaging did not reduce the risk of eight-year local or distant recurrence.
The retrospective matched cohort study by Sung et al (36) evaluating pMRI outcomes in early breast cancer patients undergoing BCS with radiation therapy found no significant difference in the locoregional recurrence rate (5% vs. 9%; P = 0.33) or disease-free survival rate (median follow-up, eight years) between the pMRI and no pMRI groups. A separate retrospective matched cohort study of 371 patient pairs with a unilateral pMRI protocol by Yi et al (46) showed a significant decrease in ipsilateral recurrence disease in patients compared with no pMRI. Using a bilateral pMRI protocol in 97 patients, there was also a reduced risk of contralateral recurrence (46). More recently, 10-year survival data from the aforementioned POMB trial were analyzed. In this study of 440 women <56 years of age, those who had undergone pMRI had a small increase in disease freedom and overall survival, though results were not statistically significant (47).
Freitas et al (48) recently published a retrospective 2:1 matched sample study in 705 women on the effect of pMRI on long-term outcomes (median follow-up, 10 years). They found more contralateral disease in patients with pMRI compared with those without pMRI (5.7% vs. 2.1%; P = 0.047) and higher overall survival in patients with pMRI (P = 0.01). In particular, patients who had larger tumor sizes (>2 cm; P = 0.03) and higher-grade (III) tumors received the greatest benefit from pMRI (P = 0.02).
The potential impact of pMRI on decisions to recommend adjuvant chemotherapy or radiation therapy should also be considered. A retrospective study of 24,379 women aged 67 to 94 years with early stage breast cancer treated with BCS using the Surveillance, Epidemiology, and end Results (SEER)-Medicare database showed no significant difference in mortality rate or subsequent mastectomy rate between the pMRI and no pMRI groups (49). These results included all patients with and without radiation therapy. However, in a subgroup analysis of women who underwent BCS alone, without radiation therapy, pMRI was associated with lower risks of subsequent mastectomy for recurrence and breast cancer mortality. This finding suggests that pMRI may be helpful in stratifying older women who could safely undergo BCS without radiation. There was no significant difference between the pMRI and no pMRI groups in terms of tumor stage or size; however, women who received pMRI had a higher proportion of moderately to well-differentiated, estrogen receptor–positive cancers and were more likely to have undergone axillary surgical staging.
In a retrospective analysis of a prospectively collected database, Gervais et al (50) investigated the association between pMRI versus no pMRI and ipsilateral local recurrence of invasive breast cancers treated with BCS and radiation therapy. This study included 461 patients, 27% of whom underwent pMRI, with a median follow-up of 97 months. Patients who underwent mastectomy after initial lumpectomy and those with positive lumpectomy margins were excluded. Patients in the pMRI group more often had a palpable breast mass, were younger (aged 50 years vs. 58 years; P < 0.0001), were more likely to have undergone adjuvant chemotherapy, and presented with larger tumors (≥2 cm) with triple-negative and HER2-positive subtypes. The investigators found no significant difference in 10-year ipsilateral local recurrence rates between the pMRI (1.6%) and no pMRI (4.2%) groups after adjusting for age, tumor size, and adjuvant chemotherapy (P = 0.37). However, in patients who did not have a pMRI, there was a significantly higher recurrence rate among those with triple-negative and HER2-positive cancers (11.8%) versus those who did not have this tumor histology (1.8%; P = 0.002), but this difference did not persist in those who did receive a pMRI (3.3% vs. 1.1%; P = 0.3), suggesting possible value of pMRI for these tumor phenotypes.
Similarly, a retrospective review of 398 women with stage I or II triple-negative breast cancer found a higher risk of recurrence among women without pMRI (51). Mammographically dense breast tissue and family history of breast cancer were also independently associated with recurrence. Another single-institution retrospective study in women undergoing BCS found a lower recurrence rate with pMRI (3.7% vs. 5.3% at five years; P = 0.04; and 4.0% vs. 8.0% at eight years; P = 0.04) on univariate analysis of all patients with invasive breast cancer or DCIS, but not multivariate analysis (52). There was no difference in ipsilateral recurrence rates for invasive cancers; however, for patients with DCIS, recurrence was significantly lower in the pMRI group compared with the no pMRI group (1.8% vs. 7.0% at five years; P = 0.06; and 3.6% vs. 10.9% at eight years; P = 0.06).
Patient Factors and pMRI
Because of the uncertain benefit of breast MRI on clinical outcomes and the potential for adverse outcomes such as false positives and delays in primary surgery, many have posited that pMRI may have better performance in certain subpopulations (eg, higher breast density, younger age, ILC) (53). The performance of pMRI across these patient characteristics is described below.
Mammographic Breast Density
Breast density is a known independent risk factor for the development of breast cancer. The sensitivity and specificity of mammography are decreased in women with dense breast tissue, which suggests that more effective screening and diagnostic imaging modalities are needed in this patient population (54,55). To date, the value of pMRI based on density alone appears to be mixed. A retrospective study that stratified breast cancer patients with pMRI into low and high breast density found that, although there were more findings in patients with high breast density, there were no statistically significant differences in the number of additional biopsy-proven malignant findings (ipsilateral, 32% vs. 23%; P > 0.15; contralateral, 6.2% vs. 3.2%; P > 0.15) or recurrence rates between the two groups (54). Similarly, Onega et al (8) retrospectively analyzed data from nearly 19 500 women with newly diagnosed breast cancer. Biopsy was performed four times more often among the 28% of women who underwent pMRI, with three times the additional cancer yield. Women with dense breasts had twice the rate of additional biopsies than women with non-dense breasts; however, there was no significant additional cancer detection among women with dense breasts when compared to women with non-dense breasts.
In contrast, a retrospective 2016 study by Seely et al (56) reviewed outcomes of pMRI in women with breasts of varying densities. Patients with dense breasts were over five times more likely to have additional findings on pMRI. Furthermore, additional lesions detected by pMRI were most likely to affect management in women with dense breasts. A 2021 study showed similar results, with MRI detecting significantly more additional biopsy-proven malignancies in women with dense as opposed to non-dense breasts (57).
Age Considerations
Age has been an important consideration in the utility of pMRI for breast cancer diagnosis. Younger women often have more advanced breast cancer on diagnosis than older women, and the malignancy more often demonstrates aggressive biological features with worse outcomes (58,59). Similar to density, data to support the use of pMRI in younger women are mixed. Onega et al (60) conducted a large population-based observational cohort study that examined the association of pMRI and all-cause mortality. This study included a cohort of women aged 66 years and older in five registries of the Breast Cancer Surveillance Consortium with non-metastatic (stage 0-III) breast cancer. There was no breast cancer–specific or all-cause mortality benefit associated with the use of pMRI in this patient population.
Adkisson et al (61) performed a retrospective review of 710 women with breast cancer stratified by age and found a nonsignificant trend toward additional cancer yield in women aged <50 years with pMRI. Park et al (62) examined the association between pMRI and surgical outcomes in women aged ≤35 years by using propensity score matching. The outcomes assessed were initial mastectomy rate, re-excision rate, and overall mastectomy rate. The study found that 27% of patients had additional suspicious lesions on pMRI, of which 49% were malignant. The pMRI group showed lower odds of repeat surgery and higher odds of initial mastectomy but no change in overall mastectomy rates. These results suggest that pMRI is valuable in younger patients and may aid in improving surgical outcomes by accurately identifying additional areas of involvement in the breast and reducing re-excision rates.
DCIS and Invasive Ductal Carcinoma
The use of pMRI to evaluate DCIS prior to surgical excision is based on two major clinical goals—identification of noncalcified components of disease and identification of occult invasive disease. Of nine studies published to date comparing the accuracy of MRI with that of mammography in determining the extent of DCIS using final pathology as the reference standard, eight concluded that MRI span was more accurate than mammographic span (63). In one study, MRI was able to accurately estimate the pathological extent of DCIS within 5 mm in 60% of cases, whereas the accuracy of mammography was only 38% (64). The overall sensitivity of MRI to accurately determine disease extent is reported to reach almost 89%, compared with 55% for mammography alone (65,66).
However, the practical surgical benefit of improved depiction of DCIS is less clear, with a meta-analysis (67) and recent systematic review (68) showing no reduction in re-excision rate with pMRI, though it is clear that the quality of data available to assess the impact of pMRI on surgery is poor (69). The ECOG-ACRIN E4112 trial demonstrated that when MRI approach and management are standardized, successful wide local excision (WLE) rate is very high (96%), with 79% of such women undergoing a single WLE (70), comparing favorably with historical data from both SEER (88% successful WLE and 57% single successful WLE, respectively) (71) and Danish registries (89% and 63%, respectively) (72). Data from the E4112 trial also indicate greater patient satisfaction when the first surgery to treat DCIS is the final surgery. Two recent single-institution retrospective studies by Lam et al (73) and Yoon et al (74) demonstrated that pMRI was not associated with more mastectomies and that its use may result in fewer surgeries, improved WLE success rate, and lower rates of positive margins. These studies confirm the potential for pMRI to reduce re-excision rates in patients with DCIS without increasing mastectomy rates. It also suggests that larger randomized controlled trials are needed, with appropriate eliminations of confounders, to understand the true value of pMRI (69).
Invasive ductal carcinoma (IDC) exists on a spectrum with DCIS and commonly contains DCIS components (75). A retrospective study of 1113 patients with DCIS or IDC showed that pMRI changed the surgical plan in 13% of patients with DCIS and 9.9% of those with IDC (76). In women with invasive disease with DCIS, the sensitivity of pMRI for DCIS is almost 100%, particularly for higher grade DCIS (32,77). Additional studies indicate that pMRI may be useful in preoperatively predicting upstaging of patients with DCIS to IDC (75,78).
The current literature advises against routinely obtaining pMRI for all cases of IDC. However, as many as 10% of breast cancers with predominantly invasive ductal histology have lobular features on biopsy. A 2021 study of 120 patients with IDC and lobular features showed that MRI findings changed management in 22.5% of cases (79). Preoperative MRI may be helpful in certain scenarios, including cases of IDC with DCIS and IDC with lobular features (76,79).
Invasive Lobular Carcinoma
Invasive lobular carcinoma is challenging to detect on conventional imaging modalities and more commonly presents as multifocal, multicentric, and/or bilateral disease, compared with IDC (80). A survey by Morrow et al (19) estimated that 72% of surgeons obtain pMRI for ILC. Several prior studies have demonstrated the utility of pMRI in providing a more accurate span of ILC disease compared with conventional imaging (23,24,81). However, despite the clear data showing better size concordance of pMRI with pathology, there is currently no evidence that suggests better outcomes or increased survival for patients with ILC who undergo pMRI.
In an observational cohort study of women with invasive breast cancer undergoing BCS, pMRI was associated with a reduced risk of positive lumpectomy margins for all types of invasive breast cancers (82). A higher risk of positive margins was associated with multiple variables, including high breast density and lobular histology. Several additional studies have shown that pMRI decreases re-excision rates with lobular cancer. Ha et al (83) showed that MRI detected new lesions in nearly 40% of patients, 65.5% of which were malignant findings, resulting in a change in surgical plan in 25.5% of patients. Furthermore, this study and others have shown decreased re-excision rates with no increase in mastectomy rates with pMRI in cases of ILC (24,84,85).
Additional Considerations
Despite evidence supporting the use of pMRI in certain patient populations and clinical scenarios, multiple concerns regarding pMRI remain. The high sensitivity of pMRI may result in additional procedures, leading to increased costs and anxiety, and higher rates of mastectomy. Several studies have shown that pMRI may lead to delays in surgery, with increased wait times predominantly attributed to further investigation of additional findings (86–88). Furthermore, studies have cast doubt on the clinical significance of additional disease detected by pMRI. Given the uncertain impact of pMRI on outcomes, including survival, these factors are important to consider for patients and providers in the preoperative setting.
With our current multidisciplinary treatment paradigms, it may be difficult to improve local-regional treatment outcomes with pMRI. However, with a shift toward de-escalation of treatment, such as accelerated partial breast irradiation (APBI) in lieu of whole breast radiation (89) or oncoplastic or multisite lumpectomies in women with multifocal or multicentric disease traditionally treated with mastectomy (90,91), there is potential for pMRI to aid in the tailoring of treatment based upon EOD. Our institution has recently created a multistep approach and criteria for when EOD pMRI would be recommended (Table 3), which considers current evidence and practice patterns. While these criteria do not include every patient factor, further evaluation will be performed after implementation of this pathway to ensure that we are actively identifying patient subgroups for whom pMRI would be of benefit.
Institutional Pathway for Recommending Preoperative Breast MRI for Extent of Disease
| Initial Evaluation and Criteria for MRI Performed for Extent of Disease |
| - Performed within breast imaging clinic at the time of radiology-pathology correlation - Criteria for recommending a breast MRI if any of the following is met: ◦ Premenopausal patient (age ≤55 years) regardless of cancer type (including DCIS) ◦ Post-menopausal patient (aged ≥56 years) with the following cancer types: ▪ Triple negative breast cancer ▪ Human epidermal growth factor receptor 2 (HER2)-positive phenotype ▪ ILC (including IDC with lobular features and mixed IDC/ILC) ◦ Breast cancer axillary nodal metastasis with unknown primary location (negative mammogram ± US) ◦ Women at high risk prior to diagnosis (known BRCA or other gene mutation, >20% lifetime risk by family history, history of mantle radiation) |
| Subsequent Evaluations and Criteria for MRI Performed for Extent of Disease |
| - Performed at time of multidisciplinary discussion, meeting the patient, additional information such as genetic testing acquired, etc. - The following criteria are considered: ◦ Newly discovered genetic mutation/high risk ◦ APBI consideration ◦ Discrepancy >1 cm in size between clinical exam, mammogram, and US ◦ Multifocal/multicentric disease ◦ Neoadjuvant chemotherapy eligibility |
| Initial Evaluation and Criteria for MRI Performed for Extent of Disease |
| - Performed within breast imaging clinic at the time of radiology-pathology correlation - Criteria for recommending a breast MRI if any of the following is met: ◦ Premenopausal patient (age ≤55 years) regardless of cancer type (including DCIS) ◦ Post-menopausal patient (aged ≥56 years) with the following cancer types: ▪ Triple negative breast cancer ▪ Human epidermal growth factor receptor 2 (HER2)-positive phenotype ▪ ILC (including IDC with lobular features and mixed IDC/ILC) ◦ Breast cancer axillary nodal metastasis with unknown primary location (negative mammogram ± US) ◦ Women at high risk prior to diagnosis (known BRCA or other gene mutation, >20% lifetime risk by family history, history of mantle radiation) |
| Subsequent Evaluations and Criteria for MRI Performed for Extent of Disease |
| - Performed at time of multidisciplinary discussion, meeting the patient, additional information such as genetic testing acquired, etc. - The following criteria are considered: ◦ Newly discovered genetic mutation/high risk ◦ APBI consideration ◦ Discrepancy >1 cm in size between clinical exam, mammogram, and US ◦ Multifocal/multicentric disease ◦ Neoadjuvant chemotherapy eligibility |
Abbreviations: ABPI, accelerated partial breast irradiation; DCIS, ductal carcinoma in situ; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma.
Institutional Pathway for Recommending Preoperative Breast MRI for Extent of Disease
| Initial Evaluation and Criteria for MRI Performed for Extent of Disease |
| - Performed within breast imaging clinic at the time of radiology-pathology correlation - Criteria for recommending a breast MRI if any of the following is met: ◦ Premenopausal patient (age ≤55 years) regardless of cancer type (including DCIS) ◦ Post-menopausal patient (aged ≥56 years) with the following cancer types: ▪ Triple negative breast cancer ▪ Human epidermal growth factor receptor 2 (HER2)-positive phenotype ▪ ILC (including IDC with lobular features and mixed IDC/ILC) ◦ Breast cancer axillary nodal metastasis with unknown primary location (negative mammogram ± US) ◦ Women at high risk prior to diagnosis (known BRCA or other gene mutation, >20% lifetime risk by family history, history of mantle radiation) |
| Subsequent Evaluations and Criteria for MRI Performed for Extent of Disease |
| - Performed at time of multidisciplinary discussion, meeting the patient, additional information such as genetic testing acquired, etc. - The following criteria are considered: ◦ Newly discovered genetic mutation/high risk ◦ APBI consideration ◦ Discrepancy >1 cm in size between clinical exam, mammogram, and US ◦ Multifocal/multicentric disease ◦ Neoadjuvant chemotherapy eligibility |
| Initial Evaluation and Criteria for MRI Performed for Extent of Disease |
| - Performed within breast imaging clinic at the time of radiology-pathology correlation - Criteria for recommending a breast MRI if any of the following is met: ◦ Premenopausal patient (age ≤55 years) regardless of cancer type (including DCIS) ◦ Post-menopausal patient (aged ≥56 years) with the following cancer types: ▪ Triple negative breast cancer ▪ Human epidermal growth factor receptor 2 (HER2)-positive phenotype ▪ ILC (including IDC with lobular features and mixed IDC/ILC) ◦ Breast cancer axillary nodal metastasis with unknown primary location (negative mammogram ± US) ◦ Women at high risk prior to diagnosis (known BRCA or other gene mutation, >20% lifetime risk by family history, history of mantle radiation) |
| Subsequent Evaluations and Criteria for MRI Performed for Extent of Disease |
| - Performed at time of multidisciplinary discussion, meeting the patient, additional information such as genetic testing acquired, etc. - The following criteria are considered: ◦ Newly discovered genetic mutation/high risk ◦ APBI consideration ◦ Discrepancy >1 cm in size between clinical exam, mammogram, and US ◦ Multifocal/multicentric disease ◦ Neoadjuvant chemotherapy eligibility |
Abbreviations: ABPI, accelerated partial breast irradiation; DCIS, ductal carcinoma in situ; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma.
Advances in artificial intelligence applied to pMRI to develop and validate imaging biomarkers hold promise to further assist with breast cancer management. The field of radiomics, which refers to the conversion of standard of care images through a quantitative processing pipeline into high-dimensional data for subsequent clinical decision making support (92), has grown substantially over the past decade and has shown clear potential for breast MRI–based radiomic features to predict clinical outcomes, such as correlation with molecular assays (93), 10-year ipsilateral breast recurrence (94), and pathological complete response after neoadjuvant chemotherapy (95). As the breast cancer community continues to refine treatment to match tumor aggressiveness, future radiomics research may uncover additional advantages of pMRI for some breast cancers.
Conclusion
The evidence and consensus guidelines for pMRI in the setting of new breast cancer are not clearly established. In this review, we have discussed multiple factors of consideration for pMRI, including patient age, breast density, histology, and impact on surgical management and patient outcomes. Though the current impact of pMRI on outcomes remains unclear, future directions, including advanced MRI techniques, may lead to increased specificity for malignant findings, thereby decreasing false positive exams and patient anxiety.
Although controversial, and despite the fact that definitive long-term outcomes have yet to be established, pMRI may be beneficial in women who are already at high risk for developing breast cancer, are of a younger age (premenopausal), have mammographically dense breasts, and have invasive lobular histology. Additional treatment factors that should be considered in obtaining pMRI include neoadjuvant chemotherapy and APBI in order to prevent undertreatment.
Funding
None declared.
Conflict of Interest Statement
H.R. and D.L.L. have grant funding from GE Healthcare. H.R. does ad hoc consulting with Guerbet, LLC. D.M.C., M.N.S., and S.H.J. have no disclosures.
