Application Value Of ~1H-MRS Associated With DCE-MRI In Breast Cancer

Purpose: (1)To investigate the clinical values of 1H magnetic resonance spectroscopy (1H-MRS) associated with dynamic contrast- enhanced MRI (DCE-MRI) in breast cancer.(2)To investigate the influencing factors of 1H-MRS in breast cancer.Materials and Methods: Thirty-eight patients with breast tumors (all tumors were non-treated primary breast tumors confirmed by pathology and with strong enhancement) examined by conventional MRI, DCE-MRI and 1H-MRS were studied.1.5T MRI (GE)was employed to examine all the cases using special breast coil. Patients were examined prone and kept still during the process with 2 breast glands falling naturally. All patients were performed with conventional MRI, which include bilateral breast axial T2WI(STIR), axial T1WI(FSE) and sagittal T2WI(FSE).DCE-MRI was performed by using axial LAVA T1WI with fat sat to scan bilateral breast. Using LAVA, precontrast axial images covering bilateral breasts were acquired with acquisition time about 55 seconds. Contrast medium (Gd-DTPA, 20ml) was administered intravenously with a flow rate of 2.5ml/s. 7-9 continuous phases scanning with the same parameters and acquisition time used before contrast began simultaneously when adminstration. All patients underwent single-voxel 1H-MRS(SS- PRESS)after DCE-MRI. Slice shown lesion clearest was chosen as the localization image with 3D voxel no less than 2.0cm3.Substraction was performed by using the Add/Sub built in the AW4.1 work station. The morphological and hemodynamic analysis of substracted images were done. MIP and VR were used to investigate whether the devious and widened blood vessels existed nearby the lesion. ROI(5- 10mm2) was set in the most enhanced area of the lesion. we measured the signal intensity and generated the time-signal intensity curve using Functool. MRS analysis was performed by using single-voxel spectroscopy analysis software . We mainly observed the Choline peaks.Results: Of 38 lesions, 20 were pathologically cancer and 18 were pathologically benign. 20 breast cancers had annular enhancement, of which 2 showed type I curve, 6 showed type II curve, and 12 showed type III curve. In 20 breast cancers, only 11 demonstrated increased choline peak, of which 8 were Invasive ductal carcinomas(6 showed type III curve, 2 showed type II curve), 3 were Invasive lobular carcinomas(2 showed type I curve, 1 showed type II curve). Of 9 lesions without choline peak, 3 showed type II curve and 6 demonstrated type III curve.18 benign tumors demonstrated nodular enhancement, among them, 12 showed type I curve, 6 showed type II curve. Only 3 had increased choline peak and all of them showed type II curve, of which 1 was fibroadenoma with chronic mastitis, 2 were great fibroadenoma in adolescent.The sensitivity, specificity and accuracy of time-signal intensity curve for differentiation were 90.0%, 66.7% and 78.9% respectively; the sensitivity, specificity and accuracy of choline peak were 55.0%, 83.3%and 68.4% respectively; the sensitivity, specificity and accuracy of the combination criteria were 100%, 66.7% and 84.2% respectively.Conclusion: Many factors affect choline peak checking. Diagnosing breast cancer only by using Choline peak has low sensitivity due to technological factors or lesions. The combination of Choline peak and patterns of time-signal intensive curve can improve the sensitivity and accuracy of diagnosing breast cancer.