Breast MRI reveals occult cancers in PBI candidates

LOS ANGELES-Breast MRI identifies mammographically occult secondary tumors in about 6% of women with early-stage breast cancer who would otherwise qualify for partial breast irradiation, Rahul D. Tendulkar, MD, a radiation oncologist at the Cleveland Clinic, reported at the 49th annual meeting of ASTRO (abstract 41).

"Breast MRI is a sensitive tool for the evaluation of disease extent in comparison with mammography, and at our institution, the majority of patients with a new diagnosis of invasive breast cancer or DCIS nonselectively undergo pretreatment bilateral breast MRI," Dr. Tendulkar said.

To assess whether this modality can help guide the selection of appropriate patients for PBI, he and his colleagues retrospectively studied the incidence of MRI-detected secondary tumors in 260 consecutive patients who met the trial eligibility criteria (based on mammography, ultrasound, physical examination, and surgical pathology) and underwent a dynamic, contrast-enhanced MRI read by a dedicated breast radiologist using computer-aided detection. Targeted ultrasound with or without biopsy was used to further evaluate any MRI-identified suspicious lesions.

Additional lesions detected

MRI detected suspicious additional lesions in 20% of the patients, and on the basis of the results of targeted ultrasound imaging, 10% underwent a biopsy of these lesions, Dr. Tendulkar reported.

Overall, MRI identified biopsy-proven, mammographically occult secondary tumors in 15 (5.8%) of the patients; the tumors were ipsilateral in 4.2% and contralateral in 1.5% (see the Figure on page 40).

The overall finding of 5.8% mammographically occult breast cancers "may represent a baseline incidence in this relatively unselected population of women with early breast cancer," Dr. Tendulkar commented at the meeting.

The rate of ipsilateral secondary tumors was significantly higher in patients whose primaries were infiltrating lobular carcinoma (ILC), compared with tumors of other histologies (18% vs 3%).

It was also significantly higher in those with pathologic T2 stage, compared with other T stages (13% vs 3%), and stage II disease, compared with other disease stages (12% vs 1%-3%), and marginally higher in patients with pathologic N1 stage.

Patients with pathologic T2 stage were also significantly more likely to have contralateral secondary tumors (6% vs 0%-1%).

The rate of ipsilateral secondary tumors did not differ according to menopausal status, clinical T stage, or age. Dr. Tendulkar pointed out, however, that none of the 18% of patients aged 70 years or older had occult secondary tumors identified by MRI.

Also, all cases of ILC as the primary had the same histology in the secondary tumor, and the secondary tumors tended to be small (0.5 cm to 3.3 cm).

These findings suggest that patients with ILC should probably have an MRI before definitive treatment, Dr. Tendulkar said, while patients aged 70 years or older may be adequately served by mammography alone.

The NSABP/RTOG trial

Accelerated partial breast irradiation is being investigated on a multi-institutional protocol for patients with early breast cancer and DCIS (NSABP B-39/RTOG 0413), Dr. Tendulkar said.

Eligible patients must have unicentric tumors measuring no more than 3 cm in diameter, as well as negative margins and no more than three involved lymph nodes.

"Utilization of breast MRI may possibly influence failure rates or failure patterns of patients entered on the NSABP/RTOG protocol," Dr. Tendulkar commented. "The hypothesis is that patients who had a pretreatment MRI may have lower failure patterns than those who did not, possibly on either or both arms of the study."