To evaluate the influence of bolus arrival time (BAT) in pharmacokinetic analysis of breast cancer.Dynamic contrast-enhanced (DCE) MRI has emerged as a promising tool in the characterization of breast cancer, due to its ability to assess tissue vascularity¹. The analysis of data obtained from DCE-MRI with high temporal resolution is conventionally performed using a two-compartment model, as presented by Tofts et al.², and quantification is performed by deconvolving the images with a corresponding arterial input function (AIF). The Tofts model assumes simultaneously BAT of contrast agent (CA) in the arteries and in the volume of interest (VOI). In reality, however, a delay may occur between the two arrival times, affecting the estimation of kinetic parameters³. The purpose of this work was to evaluate the influence of BAT in pharmacokinetic analysis of breast cancer, by estimating the kinetic parameters both with and without delay correction.The study was approved by the regional ethics committee. Thirty-eigth patients with a total of 39 verified lesions underwent MR examination using Philips Achieva (1,5T) system with NOVA gradients. The protocol consisted of a high temporal resolution 3D T1 multi-shot dual-echo EPI sequence with the following key parameters: TR = 42ms, TE = 5.5ms/23ms, flip angle = 28°, voxel size = 1.69*1.48*4mm³, number of slices = 25, temporal resolution = 2.8s/image volume with a total of 77 dynamic series acquired. A PROSET fat suppression technique was applied along with a SENSE factor of 2.5 in the AP direction. Image post-processing was performed using the nordicICE software package (NordicNeuroLab, Bergen, Norway). The images were evaluated on a voxel-by-voxel basis by fitting the acquired signal-intensity-time curves and a corresponding AIF to a two-compartment tracer kinetic model, yielding 4 different kinetic parameters; K^trans, k_ep, v_e and v_p. For each patient, the delay in onset time between the AIF- and VOI-signal was manually measured, followed by the implementation of the delay as a free variable in the model-based analysis. In order to evaluate how delayed BAT influences the reliability of parameter estimation, the kinetic analysis was also performed without delay correction. For each lesion, a VOI was manually drawn by an experienced radiologist. The mean parametric values, with and without delay correction, were extracted by using the delineated VOI. Finally, the relative error of the estimated kinetic parameters was calculated, assuming that the delayed-corrected parametric values represent a more correct characterization of the pharmacokinetics. The statistic analysis was performed by estimating the Pearson’s correlation coefficient, yielding the degree of relationship between the relative parametric error and BAT-delay. A p-value of 0.05 was used as the threshold for statistical significance.Figure 1 show the association between the corrected and noncorrected parameter estimate of K^trans, k_ep, v_e and v_p. The result indicates that neglecting the delayed BAT leads to an overestimation of the volume transfere rates, K^trans and k_ep, and the extravascular ectracellular volume, v_e, in addition to a consistent underestimation of the plasma volume, v_p. Assuming that the delayed-corrected parametric value represents a more correct characterization of the pharmacokinetics, the kinetic parameters K^trans (p=0,04) v_p (p<0,01) and v_e (p=0,02), all showed a significant correlation between the relative error and the measured BAT-delay (Figure 2).The result suggests that the two-compartment model is highly sensitive to the delay between BAT in the arteries and in the volume of interest. The result also suggests that neglecting this delay leads to a significant increased uncertainty in the estimated parameters.In order to achieve an accurate and precise quantification of the contrast kinetics, the delayed BAT needs to be accounted for in the model-based analysis.