P3-14-06: Combined Use of 18F-FDG PET/CT and MRI To Monitor Breast Cancer Response during Neoadjuvant Chemotherapy.

Abstract

Abstract Background and aim: Contrast-enhanced breast magnetic resonance imaging (MRI) is commonly used for the evaluation of the therapeutic effect of neoadjuvant chemotherapy (NAC). A recent study demonstrated the relevance of breast cancer subtype with regard to the accuracy of MRI to predict response during NAC. The role of 18F-fluorodeoxyglucose (FDG) positron emission tomography (FDG PET/CT) is under investigation, and varying results have been reported. MRI and 18F-FDG PET/CT, visualize different functional characteristics of the tumor (perfusion and glucose metabolism), but it is unclear if they complement one another to monitor response to NAC, and whether breast cancer subtype plays an independent role. Method and material: The study group consisted of 65 women with stage II or III breast cancer and measurable FDG tumor uptake, who were treated with NAC and participated in an ongoing MRI/FDG PET/CT study. MRI and FDG PET/CT scans were acquired prior to and during treatment, using prone patient positioning. Written informed consent was obtained. Prior to the start of NAC a biopsy was taken and tumors were divided into three subtypes, using immunohistochemistry: human epidermal growth factor receptor 2 (HER2) amplified (HER2+) (N=18), estrogen receptor positive/HER2 non-amplified (ER+/HER2−) (N=29), and triple negative (TN) tumors (N=18). MRI interpretation included lesion morphology at baseline, changes in morphology, tumor size, and contrast uptake kinetics (initial and late enhancement). At FDG PET/CT the maximum standardized uptake value (SUVmax) at baseline and during treatment was obtained. Tumor response at pathology was stratified to “incomplete response” (substantial presence of vital invasive tumor cells) and “favourable response” (complete absence of invasive residual tumor or only a small number of scattered tumor cells). Appropriate statistical analyses including T-tests, multivariate logistic regression and receiver operating characteristic (ROC) curves were employed to indentify factors associated with incomplete response at pathology. The following imaging and clinical features were candidates for forward feature selection in the multivariate analysis: lesion morphology, largest tumor diameter, pattern of reduction, absolute and relative change in largest tumor diameter at initial and at late enhancement during NAC, SUVmax at baseline; absolute and relative change in SUVmax during NAC; breast cancer subtype; chemotherapy regimen, and age. Results: At pathology after NAC, 32 tumors (49 %) showed favourable response and 33 (51 %) had residual disease. At multivariate analysis, breast cancer subtype, relative change in the diameter of late enhancement on MRI and relative change in SUVmax on FDG PET/CT were independent markers of tumor response at final pathology. The area under the ROC curve increased from 0.86 for MRI alone to 0.94 (p<0.01) in combination with FDG PET/CT and breast cancer subtype, yielding 84% sensitivity at 95 % specificity. Conclusion: MRI and FDG PET/CT show complementary potential to predict response during NAC, in addition to breast cancer subtype. Ackowledgement: This study was supported by the Center for Translational Molecular Medicine, project Breast CARE (grant ***03O-104). Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-14-06.