Introduction Cardiac Magnetic Resonance (CMR) imaging is an important imaging modality for assessing heart failure (HF) patients. In CRT knowledge of the coronary venous anatomy and myocardial scar can help in procedure planning. Studies using CMR to assess coronary veins, have mainly focused on patients with normal LV function using intravascular or no contrast agent providing no information about scar.

We aimed to evaluate coronary venous anatomy and myocardial scar in HF patients awaiting CRT in a single CMR examination with slow infusion of a high-relaxivity CA, dimeglumine-gadobenate (Gd-BOPTA).

Method Fourteen patients (EF 27±7.0%) referred for CRT and two patients with normal LV function were assessed with CMR. A 1.5T MR-scanner with a 32-element cardiac coil (12 patients) and a 5-element cardiac coil (four patients) was used. A multiple slice cine steady state free precession scan was performed in SA orientation to assess the ventricular function. For contrast enhanced MRI of the coronary vein, Gd-BOPTA was slowly infused (dose of 0.2 ml/kg at rate 0.3 ml/sec). To determine the optimal start point of the whole heart coronary vein scan, a dynamic ECG-triggered 2D-scan with IR preparation was used. For coronary vein visualisation, an ECG-triggered respiratory navigated 3D IR-SSFP MR-scan was used to acquire the whole-heart during a short interval (60–80 ms) in end systole using the following parameter: FA=50°, TI=300 ms, TR/TE=4.25/1.44 ms, resolution was 1.5×1.5×2 mm. After the coronary vein scan a delayed contrast-enhanced multi-slice IR gradient echo sequence was performed at end systole to depict areas of scar. LV volumes and volume rendering analysis was performed using Phillips Viewforum software. Multiplanar reformatting was used to measure ostial diameter and length of tributaries (abstract 083 figure 1). To correlate coronary venous anatomy to myocardial scar the 3D whole heart images were manually segmented with ITK-SNAP software. Scar was manually segmented using Osirix software and then registered to the 3D segmentation using geometry information stored in the DICOM header. Scar segmentation images were projected onto the left ventricular segmentation to provide 3D visualisation of the scar geometry (abstract 083 figure2).

• Download figure
• Open in new tab
• Download powerpoint

Abstract 083 Figure 1

• Download figure
• Open in new tab
• Download powerpoint

Abstract 083 Figure 2

Results In all subjects the coronary sinus (CS) and great cardiac vein (GCV) were visualised. Using an image quality score of 0 to 4, two independent observers assessed with an average score of 3 for the CS and 2.6 for the GCV. The number of branches visualised from the volume rendered images compared to previous studies (abstract 083 table 1, abstract 083 table 2). Seven patients with ischaemic cardiomyopathy and one with non-ischaemic cardiomyopathy had myocardial scar.

Conclusion We demonstrated that it is feasible to obtain clinically useful information in a single CMR examination in a group of patients with HF awaiting CRT. Furthermore we have shown the relationship between the coronary veins and myocardial scar.