At Duke University Medical Center, selective LABC patients were treated on a protocol using neoadjuvant Myocet/Paclitaxel (ChT) and HT. With the purpose of generating perfusion/permeability parametric maps and to use gadolinium (Gd) enhancement curves to score and predict response to neoadjuvant treatment, a study was designed to acquire 3 sets of DE-MRI images along the 4 cycles of combined ChT and HT. A T1-weighted three-dimensional fast gradient echo technique was used over 30 minutes following bolus injection of Gd-based contrast agent. Perfusion/permeability maps were generated by fitting the signal intensity to a double exponential curve that generates washin (WiP) and washout (WoP), parameters that are associated with the tumors vascularity/permeability and cellularity. Based on the values of the WiP, the tumors were divided in lowWI (WiP < 100), mediumWI (100 200). During the HT treatments temperatures in the breast were measured invasively via a catheter inserted under CT guidance. Although minimum sampled temperatures give a crude indication of the temperature distribution, several thermal dose metrics were calculated for each of the HT fractions (e.g. T90, T50, T10). As expected, tumors that were more vascularized (i.e. higher WiP) heated less than tumors with low WiP, a degree on average. The adjuvant treatment also changed the shape and inhomogeneity of the perfusion/permeability maps, with dramatic changes after the first fraction in responders. The correlation between the thermal metrics and pathological response will be discussed, as well as possible correlation with other tumor physiology parameters. In conclusion, the Gd-enhancement analysis of DE-MRI images is able to generate information related to the tumor vascularity, permeability and cellularity that can correlate with the tumor's response to the neoadjuvant treatment in general, and to HT in particular. Work supported by a grant from the NCI CA42745.