Multiparametric Evaluation of Treatment Response to Neoadjuvant Chemotherapy in Breast Cancer Using Integrated PET/MR

Abstract

The aim of this study was to investigate whether integrated PET/MR system can predict the treatment response to neoadjuvant chemotherapy (NAC) early in the course of breast cancer treatment. Fourteen women with newly diagnosed invasive breast cancer (median age, 54.5 years) were recruited. Each participant underwent 2 PET/MR studies. Study 1 was pre-NAC; study 2 was early in NAC treatment (after the first or second cycle). PET parameters included SUVmax and total lesion glycolysis (TLG). MRI parameters included choline signal-to-noise ratio (ChoSNR), peak enhancement ratio (PER), and the minimum apparent diffusion coefficient (ADCmin). The pathologic response was categorized as a pathologic complete response or residual cellularity of less than 10% (group 1) and residual cellularity of 10% or greater (group 2). The accuracy of the NAC response prediction was obtained by receiver operating characteristic analysis. Group 1 showed a greater reduction of SUVmax (percentage change, [INCREMENT]% SUVmax, P = 0.013; area under the receiver operating characteristic curve [AUC], 0.898), TLG ([INCREMENT]%TLG, P = 0.018; AUC = 0.878), and PER ([INCREMENT]% PER, P = 0.035; AUC = 0.837) than did group 2. The ChoSNR, ADCmin, [INCREMENT]%ChoSNR, and [INCREMENT]%ADCmin did not differ significantly between the 2 groups. The hybrid markers, [INCREMENT]%SUVmax/[INCREMENT]%ADCmin (AUC = 0.976) and [INCREMENT]%TLG/[INCREMENT]%ADCmin (AUC = 0.905), showed greater accuracy in predicting NAC response than the individual PET/MR parameters. The PET/MR parameters can predict the NAC response early in the course of breast cancer treatment. The hybrid markers more accurately predicted treatment response than the individual PET/MR parameters.