Eline E. Deurloo MD, PhD, Presenter: Nothing to Disclose

Jincey D. Sriram MD, Abstract Co-Author: Nothing to Disclose

Hendrik Jelle Teertstra MD, Abstract Co-Author: Nothing to Disclose

Claudette Elisabeth Loo MD, Abstract Co-Author: Nothing to Disclose

Jelle Wesseling, Abstract Co-Author: Nothing to Disclose

Emiel Rutgers, Abstract Co-Author: Nothing to Disclose

Kenneth G.A. Gilhuijs PhD, Abstract Co-Author: Nothing to Disclose

00030490-DMT et al, Abstract Co-Author: Nothing to Disclose

Pure ductal carcinoma in situ (DCIS) is a pre-invasive breast lesion without the ability to metastasize. DCIS can, however, be associated with presence of invasive cancer. Core biopsy has been reported to underestimate presence of invasion in up to 20% of patients with preoperatively diagnosed DCIS. Because axillary sampling is mandatory in patients with invasive breast cancer, uncertainty exists when to perform sentinel lymph node biopsy in patients with DCIS. The aim of the current study was to evaluate the efficacy of preoperative MRI to identify patients with DCIS who are at low risk of invasive breast cancer.

Patients preoperatively diagnosed with pure DCIS on core biopsy (absence of invasion) were prospectively included. All patients underwent contrast-enhanced MRI of both breasts prior to surgery. MRI was interpreted with respect to morphology, early and late kinetics of contrast uptake. In addition to MRI, clinical, mammography, and histology features from core biopsies were assessed. All patients underwent breast surgery (wide-local excision or ablation). Univariate and multivariate analyses were performed to identify features associated with presence of invasion in the resection specimens. Chi square statistics and ROC analyses were employed.

One-hundred-and-thirty-seven DCIS lesions in 134 patients were included. Mean age was 52.6 years (range 27–84 years). Fifty-seven lesions (41.6%) showed no suspicious enhancement at MRI. Of the enhancing lesions, 12 (8.8%) showed a type-1 curve (continuous increase), 22 (16.0%) a type-2 curve (plateau) and 46 (33.6%) a type-3 curve (washout). Twenty-three lesions showed invasive cancer on final histology. The most predictive features to exclude presence of invasive disease at multivariate analysis were absence of enhancement or a type-1 curve at MRI (risk of invasion 4%; area-under-the ROC-curve: 0.80, p=0.00002). Lesions with a type-2 or type-3 curve were associated with larger risk of invasion.

Complementing clinical and conventional imaging parameters, contrast-uptake kinetics at MRI provide high negative-predictive value to exclude presence of invasion. The technique shows potential to facilitate selection of patients with DCIS in whom sentinel node procedures should not be considered.

MRI shows potential to facilitate selection of patients with DCIS in whom sentinel node procedures should not be considered.

Deurloo, E, Sriram, J, Teertstra, H, Loo, C, Wesseling, J, Rutgers, E, Gilhuijs, K, et al, 0, MRI of the Breast in Patients with DCIS to Improve Selection for the Sentinel Lymph Node Procedure.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8016604.html