Magnetic resonance imaging of breast cancer: does the time interval between biopsy and MRI influence MRI-pathology discordance in lesion sizing?

Background Breast magnetic resonance imaging (MRI) is more accurate than ultrasound and mammography in estimating local extension of both invasive breast cancer and ductal carcinoma in situ (DCIS) and it is part of a breast cancer patient's preoperative management. Purpose To verify if time interval between breast biopsy and preoperative MRI, lesion margins, and biopsy technique can influence tumor sizing on MRI. Material and Methods By a database search, we retrospectively identified all women with a newly diagnosed, biopsy-proven, primary breast cancer who underwent MRI before surgery. The time interval between biopsy and MRI, the type of biopsy procedure, and various pathological features of tumors were collected. We defined the concordance between MRI and pathology measurements as a difference of <5 mm in lesion sizing. Results One hundred and sixty-six women (mean age, 51.4 ± 10.4 years) were included. The time interval between biopsy and MRI showed only a weak correlation with the absolute MRI-pathology difference (r = 0.236). Stratifying the whole cohort of patients using a cutoff value of 30 days, we found that the MRI-pathology discordance was significantly higher in patients with a biopsy-MRI time interval >30 days ( P < 0.05). By means of multivariate analysis, we found that DCIS subtype and the presence of poorly defined margins on MRI are the only two factors independently and strongly associated with MRI-pathology discordance in lesion sizing. Conclusion Size, histology, and margins of tumors may affect the accuracy of MRI measurements. The type of biopsy procedure and the time interval between biopsy and preoperative MRI are not independently associated to MRI-pathology discordance.

Breast findings incidentally detected on body MRI.

OBJECTIVES: To evaluate breast findings incidentally detected on body MRI. METHODS: A retrospective review of the institutional database identified 1752 body MRI performed between January 2015 and September 2015. MRI of women with breast tissue visible in the field-of-view were reviewed for breast findings. Breast findings were classified with the breast imaging reporting and data system (BI-RADS) lexicon. The standard statistic, costs of additional work-up, and the clinical relevance were used to describe breast findings, and we calculated 95 % exact confidence intervals (CIs). RESULTS: 440 body MRI of 440 women (mean age: 57 ± 20 years) included breast tissue in the field-of-view. A total of 41 breast findings were identified in 41 patients. Breast findings were classified BI-RADS 2 N = 25, BI-RADS 3 N = 13, BI-RADS 4 N = 3. A total of 3.6 % [95 % CI 1.6 %, 5.6 %] women with breast tissue visible on MRI had a recommendation for further imaging work-up for a breast finding. The 18.7 % (3 of 16) of these patients had a clinically important finding (breast cancer). Further imaging evaluation increased costs of €108.3 per patient with a breast finding. CONCLUSIONS: Clinically important breast findings could be detected on body MRI in up to 0.7 % (3 of 440) of women.