Most Preoperative Breast MRIs Requested by Nononcologists
San Francisco, CA—Preoperative breast magnetic resonance imaging (MRI) failed to influence clinical management in any subgroup of 257 patients with breast cancer, according to the results of a retrospective review that led one cancer center to discontinue the practice.
Preoperative MRI slightly increased the rate of total mastectomy, but it led to additional imaging studies and biopsies in more than 60% of cases. Although the MRI-prompted biopsies identified additional cancers, most of the lesions were low risk, reported Sarah P. Cate, MD, a breast surgeon at Mount Sinai Beth Israel Hospital, New York, at the 2014 Breast Cancer Symposium.
“Given these results, we are no longer using preoperative MRI routinely at our institution,” Dr Cate and colleagues concluded in a poster presentation. “Additional studies will be needed to determine strict indications for the use of MRI in the preoperative setting.”
Another report at the meeting showed that physicians other than oncologists requested most of the preoperative MRI studies of patients with breast cancer.
Despite their frequency, the use of preoperative breast MRIs remains controversial. In an effort to help clarify the imaging modality’s role in breast cancer, investigators retrospectively reviewed the records of 257 patients who underwent preoperative breast MRI during 2010-2012.
The records showed that 155 of the patients subsequently had breast-conserving surgery, and that MRI led to upstaging and subsequent total mastectomy in 8 (3%) patients. Breast density was scattered fibroglandular dense in 117 cases, heterogeneously dense in 95 cases, extremely dense in 19 cases, and predominantly fatty in 8 cases.
The preoperative MRI was followed by additional imaging in 158 cases, and 78 patients had additional biopsies as a result of their MRI findings. The biopsies revealed cancer in 32 cases. However, given the low rate of upgrading to total mastectomy, the clinical significance of the cancers remained open to question, said Dr Cate.
“We did not find any statistically significant differences in patients who were upgraded to TM [total mastectomy] by clinical factors,…though there was a trend towards upgrade to TM in those who required additional imaging,” the investigators noted in their presentation. “We found no clinical factors associated with the need for additional biopsies.”
A study from the Dubin Breast Center at Mount Sinai Hospital in New York showed that 75% of preoperative breast MRI requests came from physicians other than the operating surgeon.
The analysis included 964 patients with newly diagnosed breast cancer during 2011-2013. Of 423 breast MRI studies requested for the patients, 241 (57%) occurred before the patients were seen by their breast cancer surgeons. Medical records showed that 181 (75%) of the preoperative MRIs were requested by physicians who were not specialists in breast disease or surgery.
Socioeconomic factors significantly influenced requests for preoperative MRI. Anya M. Romanoff, MD, a research general surgery resident at the Icahn School of Medicine at Mount Sinai, New York, and colleagues, found that 65% of white patients and 66% of Hispanic patients had preoperative MRI compared with 42% of black patients (P = .001) and 26% of Asian patients (P <.001). Preoperative MRI was requested for 62% of patients who had private insurance coverage, 64% of patients with Medicare coverage, and 55% of patients who paid for their care versus 37% of patients with Medicaid coverage (P = .002).
An analysis of factors associated with orders for preoperative MRI showed that only race (P = .001) and insurance (P = .018) were significant predictors. The factors that were not associated with preoperative breast MRI included age, history of breast disease, family history of breast cancer, genetic mutation, breast density, mode of breast cancer diagnosis, and biopsy pathology.
“Further research is needed to develop guidelines for breast MRI use in newly diagnosed cancer patients,” Dr Romanoff and colleagues concluded.