Kinetic information - particularly the delayed washout phase kinetic curve - obtained by breast dynamic contrast enhanced (DCE) MRI is crucial in diagnosing breast lesions ¹. However, low specificity is a problem in DCE-MRI ². Some studies suggested the value of perfusion analysis using very early phase of DCE MRI ³. Mann et al. showed the usefulness of the Maximum slope (MS) as a novel dynamic parameter for ultrafast DCE MRI. Recently, image acquisition with higher temporal / spatial resolution becomes possible using K-space weighted Imaging Contrast (KWIC: prototype sequence), which is one of the K-space acquisition and reconstruction schemes⁴. With this radial KWIC technique, a spatial resolution of 1.0 x 1.0 x 2.5 mm ³, a temporal resolution of 3.75 seconds can be achieved. The purpose of this study is to evaluate diagnostic performance of MS obtained by ultrafast DCE MRI using KWIC acquisition compared to standard DCE MRI kinetic parameter in breast lesions.

Study population: In total 94 consecutive patients with 106 enhancing lesions (60 malignant / 46 benign lesions). Pathological diagnosis of malignant lesions consisted of 54 Invasive cancer NST, four DCIS and two Invasive lobular cancers (ILC). Benign lesions consisted of 27 fibrocystic changes (FCC), 10 fibroadenomas (FA), three intraductal papillomas, two inflammatory lesions, and four other benign lesions.

MR Image acquisition: Breast MRI studies were performed with a 3.0T scanner (MAGNETOM Trio, A Tim System, Siemens AG) with 16ch dedicated bilateral breast coil. Routinely, T2-weighted, T1-weighted, diffusion weighted and fat-suppressed T1-weighted DCE images were obtained. Acquisitions of a non-product KWIC sequence were obtained 0-1 min after gadolinium injection (TR/TE 3.57/1.68, FA 15, 1.0x1.0x2.5 mm3/voxel, 3.75 sec/frame, 16 frames). Two post contrast standard DCE images were obtained immediately after KWIC protocol. (TR/TE 3.85/1.47, FA 15, 0.86x0.95x1.0mm3/voxel, 60 sec/frame).

Image analysis and statistical analysis: The Maximum Slope (MS), a parameter to quantify contrast inflow in ultrafast DCE MRI, was evaluated using dedicated software (TWIST Breast Viewer prototype) and calculated its percentage relative enhancement/second [%/s]. For the KWIC acquisitions, the curve was obtained from a 3x3x2.5voxel region of interest (ROI),that showed the highest MS value. MS map and the corresponding time intensity curve (TIC) of typical lesion (Invasive cancer NST) are shown in Fig 1. As a semi-quantitative parameter for conventional DCE, we evaluated the washout index (WI) in ROI with highest washout with same voxel size as for MS. WI was defined as WI=(SI delay-SI early)/ SI pre x100 (%).Diagnostic performance (malignant/benign differentiation) of MS and WI were initially compared using ROC analysis. Optimal cut-off point for diagnosing malignant lesions were determined as the value that maximize Youden index to calculate sensitivity and specificity of MS and WI. Subanalysis for 78 mass / 28 non-mass lesions were performed.

MS demonstrated high diagnostic performance with area under the curve (AUC) of 0.91. Comparing diagnostic performance of MS and WI, MS shows significantly higher AUC than WI (0.91 vs 0.80; P=0.03 )(Fig 2).

With the optimal cut-off value of above 12.2 % for MS, and below -25 for WI, sensitivity and specificity of MS and WI were calculated (Tab 1). MS showed high specificity (87%) with slightly low sensitivity (83%). WI showed high sensitivity (94%) but low specificity (59%). Cases with discrepant diagnosis, i.e.false positive / negative cases for MS and WI were further analyzed for its association with pathological diagnosis (Tab 2). Of note, 12 lesions of fibrocystic changes were falsely diagnosed as malignant on WI while correctly diagnosed as benign based on low MS value. MS map and the corresponding TIC of WI false positive case are shown in Fig 3.

Our results of high AUC (0.91) is better than the previous report using MS (AUC 0.83) 3 and support the value of MS as a promising novel kinetic parameters. MS can compensate for the weakness of low specificity in standard DCE MRI. Better specificity (=low false positive rate) of MS may be explained by more accurate diagnosis of fibrocystic changes, which often show malignant wash-out pattern. Although diagnosis using MS only can shorten scanning time, MS has a drawbacks of underdiagnosis in some invasive cancer NST, ILC, and DCIS with relatively low perfusion. Combining MS with other parameters including WI, margin or other BI-RADS descriptor and ADC may help to further improve diagnostic accuracyMS obtained with ultrafast DCE breast MRI using KWIC is a promising kinetic parameter in diagnosing malignant breast lesions showing higher AUC. MS can compensate for low specificity in WI in standard DCE.