Background parenchymal signal enhancement ratio on preoperative MRI may predict local recurrence in ductal carcinoma in situ patients after breast conserving surgery

Abstract

Purpose: To retrospectively investigate whether the signal enhancement ratio (SER) of the background parenchyma around the tumor on dynamic contrast enhanced MR imaging was associated with ipsilateral breast tumor recurrence (IBTR) in breast ductal carcinoma in situ (DCIS) patients after breast conserving surgery (BCS).

Materials and Methods: Between 2004 and 2009, 215 consecutive women (median age, 47years

range, 24-74years) with pure DCIS (mean size, 2.90cm ± 1.99

range, 0.2cm – 9.8cm) who underwent preoperative MRI, curative BCS and had at least 2-year follow-up data were identified. Their clinicopathologic features (age, menopausal status, surgery type, adjuvant therapy, ER, PR, HER2 status, nuclear grade, margin width) and MRI features [lesion size, morphology, fibroglandular density, background parenchymal enhancement grade, parenchymal SER defined as (Se-Sp)/(Sd-Sp), where Sp, Se, and Sd represent the signal intensity on the precontrast, early postcontrast, and delayed postcontrast images] were analyzed. Receiver operating characteristic curves were used to determine the best cut-off value of variables for the prediction of IBTR. The reproducibility of the SER measurements was evaluated by using the intraclass correlation coefficient (ICC). RFS was estimated using the Kaplan-Meier method. A multivariate Cox proportional hazards model was used to determine associations between survival outcome and MRI variables, adjusting for clinicopathologic variables.

Results: There were 15 (7.0%, 15 of 215) ipsilateral breast tumor recurrences (9 DCIS, 6 invasive recurrences). The median follow-up period for the no recurrence group (n=200) was 48 months (range 27-100 months). The ICC between the two radiologists was 0.852 (95% confidence interval [CI]: 0.811, 0.885

P < .001) for measurements of the SER, which indicates excellent agreement. On multivariate analysis, high mean background parenchymal SER around tumor was an independent factor associated with early IBTR : The hazard ratio (HR) for high SER were 17.837 (95% CI: 4.958, 64.472

P< .001), and 10.136 (95% CI: 3.392, 30.288

P< .001) for reader 1 and reader 2, respectively. Omission of the adjuvant endocrine therapy and larger size of tumor measured at surgical specimen were also found to be independent poor prognostic factors for IBTR on multivariate analysis.

Conclusion: High SER in the background parenchyma around the tumor, omission of adjuvant endocrine therapy and larger tumor size at specimens were independent factor associated with IBTR in breast DCIS patients treated with BCS.