Routine use of magnetic resonance imaging does not improve outcomes in women undergoing lumpectomy for ductal carcinoma in situ, according to the results of a retrospective cohort study conducted at the Memorial Sloan-Kettering Cancer Center in New York.
A team led by Dr. Melissa L. Pilewskie studied 2,321 patients who underwent lumpectomy for ductal carcinoma in situ (DCIS) between 1997 and 2010. A quarter had breast MRI before or immediately after their surgery, in addition to conventional imaging with mammography and/or ultrasound.
Dr. Melissa Pilewskie
Study results, being reported in full later this week at the breast cancer symposium sponsored by the American Society of Clinical Oncology, showed that the 5-year rate of locoregional recurrence was about 8%, with no significant difference between the groups who did and did not have MRI.
The findings were the same after adjustment for potential confounders and also when analyses were restricted to the subset of patients who did not receive radiation therapy, according to Dr. Pilewskie, a breast surgeon at Sloan-Kettering.
In addition, perioperative MRI did not reduce the rate of contralateral breast cancer, which stood at about 4% in each group.
"In the absence of evidence that MRI is improving our surgical management or – as we showed here – long-term outcomes, the routine use of this test for DCIS should be questioned," Dr. Pilewskie commented in a related press briefing.
She outlined circumstances that may justify this additional testing. "I think that MRI can be a useful adjunct if there are discrepancies or still clinical questions when someone comes in, between the imaging that they have on their mammogram or ultrasound and their physical exam or their presentation," she said. "The majority of women who present with DCIS have a normal exam and calcifications or changes on a mammogram. So when there are differences, and someone has a palpable mass or nipple discharge or a different presentation that wasn’t answered by their imaging, I think MRI can help give additional information. But that again is not the routine woman who presents with DCIS."
With a median 59-month follow-up, the groups who did and did not receive MRI were statistically indistinguishable in terms of the actuarial rate of locoregional recurrence at 5 years (8.5% vs. 7.2%) and at 8 years (14.6% vs. 10.2%).
"When comparing the MRI and no-MRI groups, there were some differences between them in that the women who had an MRI had more high-risk features," Dr. Pilewskie noted, and those features might have influenced the decision to obtain this additional imaging and outcomes.
However, MRI was not a significant predictor of locoregional recurrence in a multivariate analysis that adjusted for these and other potential confounders: age, family history, mode of presentation, tamoxifen or other hormonal therapy, margin status, number of excisions, and year of surgery.
On the other hand, receipt of radiation therapy, receipt of endocrine therapy, and negative margins were all significantly associated with a lower risk of locoregional recurrence.
In a subset analysis, MRI also failed to reduce the risk of locoregional recurrence in the roughly one-third of patients who did not receive radiation therapy.
Similarly, patients who did and did not receive MRI were statistically indistinguishable in terms of the actuarial rate of contralateral breast cancer at 5 years (3.5% vs. 3.5%) and at 8 years (3.5% vs. 5.1%). Again, results were essentially the same in the subset who did not receive radiation therapy.
Dr. Steven O’Day, director of clinical research at the Beverly Hills Cancer Institute in California and moderator of the press briefing, commented, "There has been a tremendous increase in the use of MRI perioperatively and postoperatively in invasive as well as noninvasive breast cancer. And this [study] I think just grounds us to continue to – as new technologies, new imaging is used – be sure that we are actually" improving outcomes.
He concurred that the added sensitivity of MRI may be helpful in challenging cases. "But its routine use certainly in DCIS from this large retrospective study has not been shown as an independent predictor to improve locoregional or contralateral breast outcomes," he said. "So I think this is an important study" and leads us to want "to study MRI further prospectively both in DCIS and invasive cancer."
Giving some background to the research, Dr. Pilewskie noted that current guidelines do not address when MRI should be used in the work-up of patients with DCIS.
"However, about 30% of physicians currently obtain a perioperative MRI to look for areas of additional disease in patients with DCIS, and theoretically, treating this additional disease found by MRI could result in a lower risk of local recurrence or contralateral breast cancer down the road," she said. "And, again theoretically, this effect could be most pronounced in women treated with excision alone, meaning just having lumpectomy and no radiation therapy."