Rationale and Objectives Breast dynamic contrast enhanced (DCE) MRI scanning protocols vary widely. The purpose of this study was to determine the effects of post-contrast timing on delayed-phase lesion kinetics assessment and ability to discriminate malignant from benign lesions. Materials and Methods Following IRB approval, we retrospectively reviewed all lesions assessed on MR examinations from 4/2005 to 6/2006. DCE-MRI was performed with 90 seconds temporal resolution. Delayed-phase kinetic parameters including percentages of persistent, plateau, and washout, and categorizations of predominant and worst curve type were compared between 4.5 and 7.5 minutes post-contrast. Ability to discriminate benign and malignant lesions based on delayed-phase kinetic parameters was compared between post-contrast timings by receiver operating characteristic (ROC) analysis. Results 280 consecutive breast lesions (206 malignant, 74 benign) were evaluated in 228 women. Comparing kinetics assessments at 7.5 versus 4.5 minutes: volume percentage of washout increased in malignancies by a mean of 9.4% (p<0.0001) and increased slightly in benign lesions (mean 3.2%, p=0.007); predominant curve type categorizations changed significantly only for malignancies (p<0.0001); and worst curve categorizations did not change significantly for either benign or malignant lesions (p>0.05). There were no significant differences between timings in area under ROC curves for delayed-phase kinetic parameters. Conclusion The choice of delayed post-contrast timing more strongly affects the kinetics assessments for malignancies than benign breast lesions, but our results suggest a shortened breast DCE-MRI protocol may not significantly impact diagnostic accuracy. Furthermore, worst curve type classifications are least affected by post-contrast timing and may provide reliable assessment of delayed phase kinetics across protocols.

Although it has lower temporal resolution, a 180-second acquisition may be preferable because it allows higher spatial resolution and captures higher initial-phase peak enhancement without loss of delayed-phase kinetic information.

No abstract