Int J Radiat Oncol Biol Phys 2013 Mar 24;85(3):615-22. Epub 2012 Jul 24.
Department of Radiation and Cellular Oncology, University of Chicago Medical Center, Chicago, Illinois, USA.
Purpose: Retrospective data have demonstrated that breast magnetic resonance imaging (MRI) may change a patient's eligibility for partial breast irradiation (PBI) by identifying multicentric, multifocal, or contralateral disease. The objective of the current study was to prospectively determine the frequency with which MRI identifies occult disease and to establish clinical factors associated with a higher likelihood of MRI prompting changes in PBI eligibility.
Methods And Materials: At The University of Chicago, women with breast cancer uniformly undergo MRI in addition to mammography and ultrasonography. From June 2009 through May 2011, all patients were screened prospectively in a multidisciplinary conference for PBI eligibility based on standard imaging, and the impact of MRI on PBI eligibility according to National Surgical Adjuvant Breast and Bowel Project protocol B-39/Radiation Therapy Oncology Group protocol 0413 entry criteria was recorded. Univariable analysis was performed using clinical characteristics in both the prospective cohort and in a separate cohort of retrospectively identified patients. Pooled analysis was used to derive a scoring index predictive of the risk that MRI would identify additional disease.
Results: A total of 521 patients were screened for PBI eligibility, and 124 (23.8%) patients were deemed eligible for PBI based on standard imaging. MRI findings changed PBI eligibility in 12.9% of patients. In the pooled univariable analysis, tumor size ≥ 2 cm on mammography or ultrasonography (P=.02), age <50 years (P=.01), invasive lobular histology (P=.01), and HER-2/neu amplification (P=.01) were associated with a higher likelihood of MRI changing PBI eligibility. A predictive score was generated by summing the number of significant risk factors. Patients with a score of 0, 1, 2, and 3 had changes to eligibility based on MRI findings in 2.8%, 13.2%, 38.1%, and 100%, respectively (P<.0001).
Conclusions: MRI identified additional disease in a significant number of patients eligible for PBI, based on standard imaging. Clinical characteristics may be useful in directing implementation of MRI in the staging of PBI candidates.