Abstract 9604: Pulmonary Artery Systolic Blood Pressure Predicted From Cardiac MR Image Data Without Requiring Calibration; A Virtual RHC?

Abstract

Introduction: Despite decades of initiatives, assessment of absolute pulmonary arterial pressures (PAP) by CMR has remained an elusive goal. We introduce a means to predict the PAP from routine image data easily obtained by CMR. The image data is directly used to calculate the impedance between the LV and the aorta (IMP LV ) and between the RV and the main PA (IMP RV ) which are combined to predict PAP. Methods: Patients (n=33, 82% F), 51 ±12 yrs) with a diagnosis of (PHTN) underwent evaluation by CMR to assess the blood flow through the aorta and PA at the interface to the LV and RV, respectively. Phase velocity mapping flow data at single plane (acquisition time 2min) was used to calculate the previously described and validated impedance for the LV and RV: IMP LV and IMP RV . IMP was calculated using the formula:IMP = (end systolic ejection time) x (mean blood velocity) / (vessel diameter). Systolic PAP was measured during a right heart cath examination performed within one week of the CMR. A multiple linear regression model was generated to predict systolic PAP, with parameters retained with a significance <0.05. Results: The parameters that best predicted the systolic PAP were IMP LV , IMP RV and the distance of the measured IMP LV from the center position of the data range of IMP LV (r = 0.69, p<0.001), Fig 1. The standard major-axis line was fitted to the data (p<0.001). Conclusions: This is the first demonstration of the utility of a CMR-measured left and right ventricular impedance value to directly predict systolic PAP. Critically, the correlation coefficient, identical to Echo, holds over a very wide physiologic range, and has never been demonstrated before by CMR. Of central importance is the distinguishing feature that no calibration of the data is required, thus, unlike typical Echo-derived estimates of PAP or newer implanted pressure monitors, no estimate of CVP or RAP is required. Thus, this is the first-in-man CMR demonstration of a virtual RHC.