P5281Synthetic extracellular volume quantification by cardiac magnetic resonance imaging - a reliable tool in clinical practice?

Abstract

Abstract Background Extracellular volume (ECV) quantification by CMR imaging allows the measurement of changes in the extracellular space indicating diffuse edema and fibrosis, especially in early stages of myocardial disease. Usually venous hematocrit (Hct) is needed in order to calculate standard ECV. Recent findings suggest the alternative use of the blood pool T1-relaxation time (T1blood) instead of the venous Hct to generate “synthetic Hct ” and furthermore “synthetic ECV”, allowing a more routine and user-friendly application without the need of a current blood sample. Purpose Purpose of this study is to re-evaluate whether the connection between venous Hct and T1blood can be used to generate a synthetic ECV in a large number of volunteers and patients with various myocardial diseases from a tertiary care centre. Methods 945 volunteers and patients were divided into a derivation (n=490) and a validation group (n=455). Standard ECV was calculated using T1-relaxtion times acquisitioned in the septal myocardium and blood pool, each on pre- and post-contrast Images using a modified look-locker sequence at a 3 Tesla MR-scanner. In addition venous Hct was measured in a blood sample drawn prior to the scan. Synthetic Hct was derived through linear regression analysis between venous Hct and 1/T1blood and was then compared to venous Hct. Also we derived standard ECV and synthetic ECV and examined these using Lin's rho and Bland-Altman comparison. Results In the derivation group venous Hct and 1/T1blood showed a linear relationship (R2=0,22; p=ehz746.02521). This was used to calculate synthetic Hct and synthetic ECV. Synthetic ECV correlated well with standard ECV (Lin's rho= 0,903; p<0,0001). Mean difference 0,004, limits of agreement −0,038 and 0,047. See table 1 for patient characteristics and figure 1 for Bland-Altman plot on standard and synthetic ECV. Table 1. Patient characteristics Sex female (%) 312 (33%) Age, yrs 60±15 EDVi, ml/m2 86±31 ESVi, ml/m2 43±30 LVEF, % 54±16 Hematocrit, % 43±5 Figure 1 Conclusion Despite only weak correlation of synthetic and venous Hct we found excellent correlation in synthetic and standard ECV. Therefore we could confirm the value of synthetic ECV in a large cohort of volunteers and patients, providing a feasible and non-invasive marker for extracellular myocardial alterations in everyday clinical practice.