Shutter-speed dynamic contrast-enhanced MRI: Is it fit for purpose?

Purpose: To test the ability of shutter-speed dynamic contrast-enhanced (DCE) MRI to estimate water exchange (WX) using simulations and assess its performance in clinical case studies of malignant and benign breast tumors. Methods: Data were simulated using a one-compartment tracer kinetic (TK) model combined with a two-pool WX model (2PX) and with a two-compartment TK model. Typical DCE-MRI acquisition parameters were used with both WX-sensitive (8°) and insensitive (25°) flip angles. Clinical data were obtained from patients with malignant and benign breast tumors. Data were fitted using a two-compartment TK model and a one-compartment TK model combined with four WX models: fast exchange limit (FXL), no exchange, 2PX and shutter-speed. Results: Fits to the one-compartment simulated data were excellent but estimates of WX obtained using the 2PX and shutter-speed models were poor. One-compartment TK model fits to the clinical malignant tumor data were bad except for the shutter-speed model. However, that overestimated TK parameters compared to the best-fit two-compartment TK model which predicted a significant blood volume and leaky capillaries (one tracer compartment is insufficient, two are necessary). All models produced excellent fits to the clinical benign tumor data with little variation between parameter estimates (one tracer compartment is sufficient). Conclusion: The 2PX and shutter-speed models were unable to estimate WX from the DCE-MRI data. A good fit to malignant tumor data using the shutter-speed model was not explained by WX but the choice of an inappropriate TK model leading to distorted parameter estimates.