The implantation of bioresorbable Vascular Scaffolds (BVS) is a rapidly evolving technique in the treatment of coronary artery disease. BVS consist of a polylactate-based backbone which is completely dissolved after 2-3 years. Potential advantages over bare metal stents in the mid- or long term include restoration of the vessel function, reduction in angina, and avoiding very late stent thrombosis as the vessel is free of any metal caging after resorption of the BVS. Furthermore, due to the organic backbone, coronary MRI can be used as a non-invasive technique for post-interventional imaging of arteries treated with BVS, which is not possible for metal stents [1]. Artifact-free imaging has been shown in a small number of cases at 1.5T [2,3]. In this study, we further evaluated the use of 3T MRI in assessing patency of coronary arteries treated with BVS, as well as flow measurements for myocardial blood flow quantification as a possible indicator of post-interventional angina.

Coronary MR-imaging was performed in 10 patients at a 3T system (Siemens Prisma) within two days post intervention. The patients included in this study received one or more BVS in proximal segments of the left anterior descending (LAD), left circumflex (LCX) or right coronary artery (RCA). The complete MR protocol was acquired using ECG triggering. The measurements started with localization of the scaffolded artery using a low-resolution 3D navigator-gated bSSFP sequence with fat-saturation covering the whole heart. The gating acceptance window was set to the end-expiratory phase for all measurements.

For the assessment of coronary patency a 3D bright-blood navigator-gated FLASH sequence with 1.08 mm isotropic resolution, fat saturation and T2 preparation (TE/TR: 1.9/3.1 ms, FoV: 28x277x295 mm³, TE_T2prep: 40 ms, R = 2) was acquired along the course of the vessel. This was followed by a 3D navigator-gated T2-weighted variable flip angle TSE sequence with 1.15 mm isotropic resolution, dual-inversion recovery black-blood preparation and elliptical k-space coverage (TE_eff: 93 ms, TR: 2*(average duration of cardiac cycle), FoV: 16x234x294 mm³).

In a subgroup of 6 patients, time-resolved blood flow velocity quantification was done via phase-contrast MRI, using a cross-sectional 2D single slice with through-plane velocity-encoding positioned proximal to the scaffolded vessel area. Images were acquired with spiral sampling, prospective ECG-triggering and respiratory navigator-gating as well as a 1-2-1 binomial water selective RF excitation pulse for fat signal suppression (TE/TR: 4.8/15.2 ms, 16 interleaves, SL: 8 mm, FoV: 200x200 mm³, VENC: 35 cm/s). As an additional functional parameter, the ejection fraction was measured with a multi-slice cine-bSSFP sequence in short-axis view.

Reformatted 3D bright-blood images show prominent and wide open luminal areas in all patients (cf. fig. 1). Luminal characteristics corresponded well to digital subtraction angiography images obtained during the coronary intervention. As previously shown at 1.5T, no susceptibility or RF-shielding artifacts arising from the BVS are seen in both gradient-echo and spin-echo images. 3D TSE imaging was successful in only 5 patients. This could be explained by high and unstable heart rates seen in some patients which can lead to severe image artifacts in TSE acquisitions due to their high susceptibility to motion. Average acquisition durations of the 3D sequences were: 8:47 ± 3:00 min (FLASH) and 6:56 ± 3:21 min (TSE).

Figure 2 shows an example of the coronary flow measurements in a patient treated with a BVS in the RCA as well as a patient treated with a BVS in the LAD. Flow velocities are averaged over the cross-section of the artery and plotted against the time after the R-wave. In general, the measured flow velocity curves agree with previously published coronary flow MRI [4] showing higher peak velocities in the LAD compared to the RCA that occur also later in the cardiac cycle, indicating that blood flow through the left coronary artery occurs mainly during diastole.

In this study, we show that artifact-free 3T coronary MRI for the judgement of luminal patency of scaffolded coronary arteries is feasible and safe in patients treated with bioresorbable vascular scaffolds. Furthermore, measurement of blood flow and blood flow velocity could be performed in segments proximal to BVS implantation. In a follow-up, patients will be examined with the same MR protocol, so that potential changes in patency and blood flow pattern of the treated artery as well as the ejection fraction could be assessed. This might help to explain clinical findings such as post-interventional reduction of angina.