MRI for breast cancer: Current indications

Abstract Mammography is the only imaging study that has been proven in multiple large randomized trials to decrease breast cancer mortality. Mammography, however, has its limitations and, as such, other modalities that can complement it are being studied. One of these is dynamic contrast-enhanced breast MRI, which has emerged as an important adjunctive modality and is at present the most sensitive modality that we have to evaluate the breast. The American College of Radiology, in its 2004 practice guidelines, has outlined the 12 current indications for breast MRI. This manuscript reviews and provides examples of each of these.


Introduction
Mammography is the only imaging study that has been shown in multiple large clinical trials to decrease mortality in breast cancer. However, mammography has well-known limitations; for example, it has limited use when there is increased breast tissue density, in the diagnosis of lobular carcinoma, in the postoperative breast, and in patients with BRCA and other gene mutations. As such, there has been an effort to develop other modalities to complement mammography. Dynamic contrast-enhanced breast MRI (DCE-MRI) has emerged as an invaluable adjunctive tool. The most recent American College of Radiology (ACR) practice guidelines for the performance of breast MRI [1] outline 12 indications [ Table 1] for DCE-MRI. This article presents an example of each of these indications and reviews the literature in support of the recommendations.

Warnings and protocols
There is no standard recommended protocol for performing DCE-MRI. Protocols vary with the equipment being used and the clinician's preference. For example, while some clinicians favor evaluating images in the axial planes, others choose to interpret in the sagittal plane. There are, however, minimum standards for the performance of breast MRI and these are outlined in Table 2. A dedicated breast coil, at least 1.5-Tesla magnet strength, and dynamic contrast administration are absolute requirements for the performance of breast MRI. A power injector is highly recommended to standardize contrast administration from study to study.  Breast MRI should not be used instead of mammography; it is a complementary study to other breast imaging modalities. It is recommended that a current mammogram be available for comparison when interpreting the breast MRI. Finally, DCE-MRI should not be used in lieu of a biopsy of a suspicious lesion found by USG, mammogram, or physical examination, or for the evaluation of characteristically benign lesions. Women undergoing breast MRI should be advised that although the study is highly sensitive, its low specificity could lead to recommendations for additional imaging follow-up studies or biopsy.
A final but important warning is the need for the establishment of MRI-guided needle localization or biopsy capability in any breast MRI practice. Since there will be lesions found at the time of DCE-MRI that are clinically, mammographically, and sonographically occult, there needs to be a program in place to address such lesions.

Current ACR practice guidelines Lesion characterization
Lesion characterization is probably the weakest and least investigated indication for DCE-MRI. [2] However, when mammography and USG fail to fully evaluate a finding, we have found breast MRI to be a useful complementary study to conventional breast imaging modalities [ Figure  1]. Although DCE-MRI is highly sensitive, the specificity and negative predictive value (reported, for example, by Bluemke et al, [3] to be 67.4% and 85.4%, respectively) are not sufficiently high to preclude biopsy when there are suspicious imaging findings.

Response to neoadjuvant chemotherapy
Neoadjuvant chemotherapy is routinely used in advanced breast cancers to reduce the size of the tumor so that conservation surgical therapy can be performed. MRI has been shown to be better than physical examination, mammography, and USG for assessing residual disease after neoadjuvant chemotherapy. [4,5] There are, however, limitations to DCE-MRI evaluation of residual disease after neoadjuvant chemotherapy. MRI tends to overestimate the size of residual disease and, because of the antiangiogenic effects of certain chemotherapeutic agents on tumor, the ability of DCE-MRI to evaluate lesion enhancement can be significantly decreased. Figure 2 demonstrates the pre-and post-chemotherapy MRI appearance of breast cancer in a patient who had complete response to therapy.

Extent of infiltrating lobular carcinoma
Infiltrating lobular carcinoma is known to be a diagnostic challenge in mammography. It is often seen in only one projection and is well known to be underestimated by both mammography and USG. [6] Breast MRI has been shown in several studies to better depict the extent of lobular carcinoma. [7,8] In Figure 3 it can be seen that the DCE-MRI

Extent of infiltrating ductal carcinoma
DCE-MRI has been shown in multiple studies to be capable of accurately evaluating the extent of infiltrating ductal carcinoma and finding additional, mammographically occult, areas of disease. [9][10][11] Evaluating the extent of disease in patients with newly diagnosed breast cancer is the most common indication for DCE-MRI at our institution. Figure  4 demonstrates the dramatic extent of disease revealed by DCE-MRI in a patient with recently diagnosed breast cancer. In this patient who clinically had inflammatory changes, the study also served to establish a baseline prior to neoadjuvant chemotherapy.

Axillary node metastases with an unknown primary
Occult breast cancer is an uncommon presentation of breast carcinoma. When axillary lymph node metastasis is identified and the mammogram is negative, breast MRI is able to locate the primary site in 75-86% of women. [12,13] Although identifying the primary breast tumor will not affect the prognosis in such cases, it will allow the patient to consider breast conservation surgery as a treatment option. Figure 5 shows a patient with biopsy-proven lymph node metastasis of a breast primary and a negative mammogram who had a positive breast MRI. Second-look USG was then performed and the lesion was identified and targeted for biopsy.

Postoperative tissue reconstruction
Many patients who have undergone mastectomy choose to have breast reconstruction with autologous tissue such as a transverse rectus abdominis myocutaneous flap (TRAM), latissimus dorsi flap, or gluteal flap. Discussion of the reconstructive technique is beyond the scope of this manuscript; briefly, the chest wall is covered with fatty tissue and in some cases an implant is added to give bulk to the reconstructed mound. Follow-up of these patients, both clinically and with mammography, gives only limited information. Mammography has not conclusively been shown to help in detecting recurrence, [14] but it can be used as part of routine surveillance in patients with a history of breast cancer. DCE-MRI, which can clearly depict the chest wall, is helpful in identifying local recurrent disease. [15] In addition, DCE-MRI can identify benign changes that can present clinical dilemmas in patients with autologous tissue reconstruction. Figure 6 demonstrates the case of a patient who presented with a new palpable abnormality at the site of a TRAM flap site and was noted to have suspicious enhancement. In this case, however, there was no histologic confirmation since the patient refused further evaluation and therapy.

Silicone and non-silicone breast augmentation
Before the availability of gadolinium contrast enhancement, evaluation of implant integrity in patients with silicone breast augmentation was the first indication for MRI imaging of the breast. The many types of implants and the appearance of rupture have been nicely reviewed by Middleton and McNamara [16] In patients with implants, the mammogram may be difficult to interpret, and evaluation for the presence of cancer in such cases is another indication for DCE-MRI. An additional advantage of imaging with MRI is that it can visualize lesions behind the implant and it is advocated as the study of choice for evaluation of breast cancer in patients with implants. [2] Figure 7 demonstrates the usefulness of DCE-MRI in not only defining the extent of disease but also in finding additional foci of carcinoma in a patient with a breast implant. At our institution, for patients with silicone implants being evaluated by DCE-MRI, we perform implant-specific sequences. This procedure adds just minutes to the examination and provides useful information which, in the case of implant rupture, can be of clinical significance.

Invasion deep to fascia
Both mammography and USG have limitations in the evaluation of the chest wall. MRI is able to visualize the entirety of the chest wall. Enhancement of the pectoralis and intercostal muscles is indicative of chest wall invasion in patients with a posterior breast tumor. [17] Obliteration of the overlying fat plane, in contrast, is not sufficient to suggest chest wall invasion. Knowledge of chest wall invasion is invaluable for preoperative planning. In Figure 8, the mammogram [ Figure 8A] does not completely image the posterior chest wall. USG [ Figure 8C], because of posterior shadowing caused by the tumor, gives no information on the status of the chest wall. The DCE-MRI [ Figure 8C], in contrast, demonstrates no abnormal enhancement of the muscle, and at surgery there was no facial invasion.

Contralateral breast screening
Women with a history of breast cancer are at increased risk for additional breast cancers. As many as 7% of women will be diagnosed with metachronous disease and up to 3% will have contralateral synchronous disease. [18] In recent studies, DCE-MRI has been able to identify occult contralateral cancer in 3-5% of the cases. [18,19] MRI, thus, is of value as a study to screen the contralateral breast in patients with a new diagnosis of breast cancer. At present, there is no data on the impact that DCE-MRI might have on survival when a synchronous contralateral tumor is identified. With the increased use of partial breast irradiation it will become even more critical to identify additional areas of disease. Figure 9 demonstrates the findings in a patient with a contralateral synchronous tumor. With high-resolution axial and parallel imaging it is now feasible to image both breasts simultaneously.

Residual disease post-lumpectomy
Lumpectomy followed by radiation is an acceptable choice in the treatment of stage I and II breast cancer and has been shown to provide the same survival as radical and modified radical mastectomies. [20] Positive margins are known to increase local recurrence rates. The rate of positive margins varies between surgeons but, in general, it is accepted that 40% of lumpectomies will have positive margins. The advantage of obtaining MRI prior to returning to the operating room for re-excision is that MRI helps in identifying multifocal or multicentric disease, which would change the management from lumpectomy to mastectomy. [21] MRI evaluation can also inform the surgeon as to the extent of residual disease and its location. Because of postoperative inflammatory changes, it is accepted that the specificity of DCE-MRI is limited in the postoperative period and that imaging earlier than 28 days post surgery will adversely affect accuracy. [22] Figure 10 shows how a patient with positive post-lumpectomy margins benefitted from DCE-MRI prior to re-excision.

Surveillance of high-risk patients
The generally accepted risk factors for breast cancer are outlined in Table 3. For patients with these risk factors there is sufficient evidence to recommend annual DCE-MRI in addition to annual mammography for screening for breast cancer. [23] In patients with genetic mutations, cancer is diagnosed at an earlier age, breast tissue is denser and, typically, the lesion is relatively larger in size at the time of diagnosis. There is insufficient evidence to recommend DCE-MRI screening in patients who have a personal history of breast cancer, prior atypical ductal hyperplasia, or other high-risk lesions at breast biopsy, and also in patients with heterogeneously dense or very dense breast glandularity. Figure 11 shows a case of mammographically occult breast cancer identified by DCE-MRI in a known breast cancer (BRCA1) gene mutation carrier.

Recurrence of breast cancer
Evaluation of breast cancer in patients with autologous tissue or implant breast reconstruction has been described earlier in this manuscript. Postoperative changes are also known to hinder the evaluation of breast cancer recurrence at the lumpectomy site. Although there is little published data, MRI may be useful in evaluating for recurrent disease in patients in whom conventional imaging is confusing due to considerable postoperative scarring. While scar may enhance on MRI for 1-2 years following surgery, a negative MRI may be helpful in excluding recurrent disease. This may be more difficult when the postoperative scar is still enhancing. In general, a scar tends to present as a thin rim or cloud of enhancement around the cavity, whereas recurrent tumor tends to be more clumpy or mass-like.

Conclusion
In this article we review the indications for DCE-MRI examination as per the current ACR guidelines and present examples for each of these indications. Breast MRI has emerged as the most sensitive modality for evaluation of the breast; however, it is limited by low specificity. Breast MRI does not replace mammography for screening of breast cancer in the general population. MRI-guided localization or a biopsy system and the requisite expertise are needed for any breast MRI program as there will be lesions that will not be seen by other imaging modalities.