Diffusion-weighted imaging using ADC mapping as an imaging biomarker for breast cancer invasiveness.

Abstract

11093 Background: Recently, functional imaging techniques such as diffusion weighted imaging (DWI) have been added to routine MR and have shown great potential for improving breast cancer diagnosis. DWI depicts cellular diffusivity on a molecular level and can be quantified using the apparent diffusion coefficient (ADC). In malignant tumors diffusivity is restricted, leading to lower ADC values than benign tumors. The aim of this study was to proof, that DWI can be used to differentiate benign from malignant tumors and to elucidate if ADC can serve as an imaging biomarker for breast cancer invasiveness. Methods: In this IRB-approved study 250 patients with 267 suspicious breast lesions (BI-RADS IV-V) were included. All patients underwent routine MR at 3T. A DWI-sequence was added to a standard imaging protocol, increasing measurement time by 2:30 min. The lesions were identified in routine MR and DWI sequences and ADC values of the lesions were calculated. Histopathology was used as the standard of reference for all lesions. Appropriate statistical tests were used to compare the ADC values of benign and malignant tumors (cut-off value 1.25×10-3mm²/s), of invasive and non-invasive disease and between different invasive tumor subtypes. Results: There were 91 benign (mean ADC 1.58×10-3mm²/s) and 176 malignant (.94×10-3mm²/s) lesions, sensitivity and specificity were 94.3% (PPV 95.4%, CI 0.91-0.98) and 91.2% (NPV 89.2%, CI 0.81-0.94). 155 lesions were invasive cancers (median ADC .90×10-3mm²/s), while 21 were non-invasive ductal carcinoma in situ (1.22×10-3mm²/s). The invasive cancers were 130 invasive ductal (median ADC .91×10-3mm²/s) and 25 invasive lobular cancers (.83×10-3mm²/s). ADC was significantly different between benign and malignant lesions (p<.001) and between invasive and non-invasive cancers (p<.001), while no significant difference could be found between the invasive cancer subtypes (p=.163). Conclusions: Diffusion-weighted imaging reliably allows differentiation of benign and malignant breast tumors. The data suggest that ADC can be used as a non-invasive imaging biomarker for breast cancer invasiveness and may be of importance to treatment planning and outcome in breast cancer patients.