Model‐based, semiquantitative and time intensity curve shape analysis of dynamic contrast‐enhanced MRI: A comparison in patients undergoing antiangiogenic treatment for recurrent glioma

Abstract

To compare time intensity curve (TIC)-shape analysis of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) data with model-based analysis and semiquantitative analysis in patients with high-grade glioma treated with the antiangiogenic drug bevacizumab. Fifteen patients had a pretreatment and at least one posttreatment DCE-MRI. We applied a pixel-by-pixel TIC shape analysis, where TICs are classified into five different types according to their shape, and calculated the occurrence of each TIC type in the region of interest (ROI). The results were compared to the pharmacokinetic model (PKM) parameters K(trans) , K(ep) , V(e) , and V(i) , and with the semiquantitative parameters maximum enhancement (ME) and initial slope of increase (ISI). The relative amount of type 2 and 4 TIC shape significantly correlated with the parameter K(ep) but not with K(trans) or V(e) . The PKM parameter V(e) and the semiquantitative parameters ME and ISI showed significant changes after treatment. None of the TIC shapes individually showed significant changes. The semiquantitative parameters ME and ISI are more sensitive to the effect of the bevacizumab than K(trans) and V(e) . The pixel-by-pixel TIC shape analysis parameters are not sensitive to the effect of bevacizumab, although they can be seen as surrogates for the PKM parameter K(ep) .