Application of magnetic resonance computer-aided diagnosis for preoperatively determining invasive disease in ultrasonography-guided core needle biopsy-proven ductal carcinoma in situ

The aim of this study was to analyze kinetic and morphologic features using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) with computer-aided diagnosis (CAD) to predict occult invasive components in cases of biopsy-proven ductal carcinoma in situ (DCIS). We enrolled 138 patients with 141 breasts who underwent preoperative breast MRI and were diagnosed with DCIS via ultrasonography (US)-guided core needle biopsy performed at our institution during January 2009 to December 2012. Their clinical, mammographic, ultrasonographic, MRI, and final histologic findings were retrospectively reviewed. Their mammographic, ultrasonographic, and MRI findings were analyzed according to the American College of Radiology Breast Imaging Reporting and Data System. CAD findings of detectability, initial (fast, medium, and slow) and delay (persistent, plateau, and washout) phase enhancement kinetic descriptor, peak enhancement percentage, and lesion size were evaluated. Continuous and categorical variables were analyzed using independent t test and χ(2) or Fisher exact test, respectively. Independent factors for predicting the presence of invasive component were evaluated by multivariate logistic regression analysis. Final histologic findings revealed that 55 breasts (39%) had DCIS with an invasive component. MRI-detected, CAD-detected, or pathologic lesion size (P = .002, P = .001, P < .001, respectively), delay washout kinetics and detectability on CAD (P < .001 and P = .004, respectively), presence of symptoms (P = .01), presence of comedonecrosis (P < .001), nuclear grade (P = .001), abnormality on mammography (P = .02), or US (P = .03) were significantly different between pure DCIS and the DCIS with an invasive component group on univariate analysis. Of those findings, multivariate analysis revealed that delay washout on CAD (odds ratio [OR], 4.36; 95% confidence interval [CI], 1.96–9.69; P = .0003) and pathologic size (OR, 1.29; 95% CI 1.05–1.57; P = .014) were independent predictive factors for the presence of an invasive component. Delay washout kinetic features measured by CAD and pathologic tumor size are potentially useful for predicting occult invasion in cases of biopsy-proven DCIS. Breast MRI including a CAD system would be helpful for predicting invasive components in cases of biopsy-proven DCIS and for selecting patients for sentinel lymph node biopsy.