Breast Magnetic Resonance Imaging in the Management of Early Breast Cancer: Implications for Partial Breast Irradiation

Reviewer: Christine Hill, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 30, 2007

Presenter: R.D. Tendulkar
Presenter's Affiliation: Cleveland Clinic, Cleveland, OH
Type of Session: Scientific


  • For patients with early stage breast cancer, breast conserving treatment (BCT), consisting of lumpectomy followed by breast irradiation is currently a standard of care. Current radiotherapy techniques involve delivery of 45 – 50 Gray (Gy) to the whole breast in fractions of 1.8 – 2 Gy fractions, followed by a radiation boost delivered to the tumor bed.
  • Recently, accelerated partial breast irradiation (APBI) techniques have been developed and are being utilized in some centers. APBI generally consists of delivery of radiation to only the tumor bed with a small margin of surrounding tissue. This radiation is frequently hypofractionated and may be delivered using either external beam photon radiation or brachytherapy.
  • Currently, no long term data exists regarding clinical outcomes with APBI, and a multi-institutional protocol, NSABP B-39/ RTOG 0413 is currently underway.
  • Eligibility requirements for NSABP B-39/ RTOG 0413 include unicentric disease, tumor size ≤ 3 cm, tumor classification Tis – T2, nodal classification N0 – N1, and negative margins.
  • Magentic resonance imaging (MRI) is a sensitive tool for evaluation of breast disease, and has been demonstrated to increase detection of mammographically occult disease in both the ipsilateral and contralateral breasts at the time of diagnosis of breast cancer [Lehman CD, NEJM, 2007].
  • In determining patient eligibility for APBI, detection of occult disease within the ipsilateral breast may be especially important. This study was carried out to evaluate the rate of MRI detection of incidental second breast lesions in patients who would otherwise have been candidates for APBI.

Materials and Methods

  • The records of patients diagnosed with early-stage breast cancer between 6/2005 and 12/2006 at the Cleveland Clinic were retrospectively reviewed.
  • All patients analyzed in this study had undergone bilateral breast MRI as part of a standard pre-operative work-up using a 1.0T Siemens Harmony MRI unit. Images were interpreted by radiologists specializing in breast imaging, with computer-based assistance.
  • Concerning or suspicious lesions detected on MRI were further evaluated with ultrasonography. Biopsy was performed if clinically indicated.
  • Eligibility criteria for this study included the eligibility requirements for NSABP B-39/ RTOG 0413 (unicentric disease, tumor size ≤ 3 cm, tumor classification Tis – T2, nodal classification N0 – N1, and negative margins). Additionally, patients were excluded if mammograms or pathology were not available for review.
  • Rates of detection of occult disease by MRI were used to calculate sensitivity, specificity, positive predictive value, and negative predictive value of the test in this clinical setting.
  • Results obtained from MRI were used to determine patients who would have been ineligible for NSABP B-39/ RTOG 0413 based only on MRI findings.


  • 260 patients who met eligibility criteria for NSABP B-39/ RTOG 0413 based on mammogram/ ultrasound, physical exam, and surgical pathology were identified.
  • Median age was 57 years (range 28 – 86 years).
  • 69 % of patients evaluated were post-menopausal, and 30% had breast tumors that were palpable at the time of diagnosis.
  • The majority of patients (75%) had invasive breast cancer, and the remainder had ductal carcinoma in situ without an invasive component. Of patients with invasive disease, 80% had invasive ductal carcinoma and 9% had invasive lobular carcinoma (11% had mixed or other pathology).
  • Of the patients with invasive disease, 165 had pathologic T1 tumors, 31 had pathologic T2 tumors, and 39 had pathologic N1 nodal disease. With regards to the entire group, 64 patients had stage 0 disease, 139 had stage I disease, 44 had stage IIA disease, and 13 had stage IIB disease.
  • MRI identified concerning second lesions in the unilateral breast in 36 patients, and in the contralateral breast in 16. After ultrasound evaluation, 27 patients underwent biopsy of the lesions identified.
  • Mammographically occult malignancies were identified in 16 patients. Biopsy proven multifocal, ipsilateral disease was identified in 3 patients, multicentric, ipsilateral disease in 9 patients, and contralateral breast cancer in 4 patients. These lesions represented an overall rate of detection of mammographically-occult malignant lesions of 5.8%.
  • The sensitivity of MRI for detection of multifocal/ multicentric ipsilateral disease was calculated to be 85%. The specificity was 87%. Positive predictive value was 26%, and negative predictive value was 99%. Overall accuracy of MRI was 87% (95% CI 83 – 91%).
  • Increased risk of second tumor site was observed for patients with invasive lobular carcinoma, for whom rate of detection was 18% versus 3% for patients with invasive ductal carcinoma (p < 0.01). The majority of these patients had occult ipsilateral disease. In all cases of invasive lobular carcinoma, second tumor pathology was also invasive lobular carcinoma. 
  • A trend towards increased risk of second lesion detection was observed for patients with T2 tumor classification and/ or stage II disease (p = 0.06). Of this group, stage II disease (T2 N1) appeared to increase risk of occult ipsilateral disease. Tumor classification T2 in the absence of nodal disease appeared to increase risk of occult contralateral breast cancer.
  • Age and menopausal status did not significantly affect rates of detection; however, no occult cancer was identified in patients over 70 years of age.

Author's Conclusions

  • The authors conclude that breast MRI can identify mammographically occult multifocal, multicentric, and contralateral breast cancer in early breast cancer patients who might otherwise be candidates for APBI.
  • The authors note that this may be particularly important for patients with invasive lobular pathology.
  • Finally, the authors note that no occult disease was identified in women over 70 years in this study. As a result, they recommend that mammography with or without ultrasound may be sufficient studies for this subset of women.

Clinical/Scientific Implications

  • APBI is currently being investigated in the NSABP B-39/ RTOG 0413 trial, and is being utilized to treat early-stage breast cancer in several centers.
  • Long-term clinical outcomes with these techniques are not currently available; however, one concern raised by breast cancer specialists is the lack of radiation treatment to areas with occult disease within the breast that may be discontiguous from the primary tumor with APBI. With standard radiotherapy, these areas may be sterilized during the whole-breast phase of treatment.
  • This study demonstrates a rate of 5.8% of occult second cancers in patients with early-stage breast cancer that were identified on MRI. The presence of second sites of malignancy would certainly be expected to increase local recurrence rates and potentially death from breast cancer if left untreated. Based on these results, MRI appears to be a tool that should be utilized to further identify patients who are not optimal candidates for APBI.
  • Pathologic invasive lobular carcinoma has been demonstrated previously to increase risk of second breast cancers, and this is confirmed by this study. These patients in particular may benefit from use of breast MRI for detection of second lesions, and may be suboptimal candidates for APBI.
  • Although MRI has been demonstrated to be superior to mammography in sensitivity for detection of occult breast malignancy, no data exists regarding the presence of occult disease that may not be detected by MRI. The absence of a lesion identified on MRI certainly does not eliminate the possibility of malignant foci within the breast that are not radiographically detectable. The concern for such areas of disease does raise questions regarding use of APBI even with MRI. The results of the current NSABP B-39/ RTOG 0413 trial will certainly speak to this question as local recurrence rates and other clinical outcomes are evaluated with APBI versus conventional breast radiotherapy.

Partially funded by an unrestricted educational grant from Bristol-Myers Squibb.

Frequently Asked Questions