Carboplatin  high Ki 7 expression showed pCR. The difference in treat-ment response was statistically signi ant between these groups. Pathologic results showed residual invasive cancer in 7 patients . MRI detected enhanced lesions in 9 of these 7 patients . Among the remaining patients with patho-logically proven residual canc MRI showed no suspicious en-Clinical Breast Cancer April MRI of HE Breast Cancer Table The Morphologic Features of Tum Pathologic Diagnos and the Elapsed Time Between MRI and Surgery in Patients Who Were Incorrectly Diagnosed as Havingplete Response by MRI Patient Abbreviations: MRI Morphologic Features Mass NML Mass Mass NML NML Mass NML magnetic resonance imaging; NML Pathologic Findings after Surgery small scattered foci of intrating ductal carcinoma Intrating ductal carcinoma involving the skin Intrating ductal carcinoma with a micropapillary pattern Invasive ductal carcinoma with extensive fibrosis.

Multiple scattered foci of invasive ductal carcinoma Invasive ductal carcinoma presented in an area of extensive fibrosis Multiple scattered foci of invasive ductal carcinoma Intrating ductal carcinoma with ductal carcinoma in situ nonmass like enhancement lesion. Days Between MRI and Surgery hancements . Five of these patients with false-negative results had MRI performed at T and had MRI performed at T. Seven of the false-negative diagnoses belonged to patients with HR canc and only patient had HR disease. Figure Correlation Between Residual Tumor Size After NAC Determined by MRI and Pathologic Examination of Surgical Specimen Table summarizes the MRI tumor morphologic and al patho-logic dings of these patients who were falsely diagnosed as havingplete response Aurora Kinase Inhibitors by MRI. The overall sensitivi speci i and accuracy of MRI diag-nosis were found to be 8, 4, and 3, respectively. The positive predictive value was 7 and the negative predictive value was 7.

In subgroup analys the sensitivi speci i accura positive predictive val and negative predictive value of MRI diagnoses were 6fi 2fi and 6, re-spective for HR disea  respective for HR cancers. It was obvious that the negative MRI Pathology Size Correlation R = predictive value of MRI was higher in the HR group than in the HR group. Accuracy of MRI in Diagnosing Residual Tumor Size MRI and pathologic tumor size discrepancy were analyzed for each case. The overall MRI “pathologic evaluation correlation for residual tumor size in HE tumors showed Pearson R . In subgroup analys HR MRI Size Overa the diagnostic performance of MRI for predicting resid-ual tumor size was more accurate for triple-negative cancers and cancers with high Ki 7 expressi which show a better NAC re-cancers showed a higher correlation than did HR can-sponse . cers . The average discrepancy between MRI and patho-logic tumor size for the whole cohort was cm . A worse MRI “pathologic evaluation tumor sizeparison of MRI Diagnostic Accuracy Between Patients Receiving Different NAC Regimens discrepancy was found in HR than in HR Of the 8 HR patien 6 received AC and taxane sequential cancer . The 0 received only a geeks taxane-based regim and received only AC. In range of tumor size discrepancy was also signi antly different be-HR patien 0 received AC and taxane sequential and re-tween the groups .