Number 34, 2007
Picture my heart!

Cardiac interventions guided by magnetic resonance imaging

Back to the Summary

Titus Kuehne
Department of Pediatric Cardiology and Congenital Heart Disease, Deutsches Herzzentrum, Berlin

Correspondence: Titus Kuehne, Department of Pediatric Cardiology and Congenital Heart Disease, Deutsches Herzzentrum Berlin, Augustenburger Platz 1, 13353 Berlin, Germany. Tel: +49 30 4593 2800; fax: +49 30 4593 2900; e-mail: kuehne@dhzb.de

Introduction
At present there is a clear trend of replacing X-ray fluoroscopy by alternative imaging modalities, fueled mainly by the association of X-ray technology with significant exposure of patients and medical staff to ionizing radiation, its lack of visualization of soft tissue and its limited ability to acquire quantitative cardiovascular functional data. Interventional magnetic resonance imaging (iMRI) is a young, emerging technology that has already been demonstrated to have the potential to complement, or even to replace in part, X-ray-guided procedures. MRI-guided intervention can add considerable benefit in procedures that rely on good visualization of soft-tissue or 3-dimensional anatomy, in addition to immediate pre- and postinterventional assessment of physiologic parameters of cardiovascular function. As experience in the technically challenging field of iMRI is accumulating rapidly, much interesting new work can be expected in the near future.

Hardware and software considerations
Considerable efforts have recently been undertaken by the manufacturers to design magnetic resonance scanners that provide improved patient access. Siemens offers a short-bore 1.5 Tesla scanner with bores of length 125 cm and width 70 cm. Philips has recently released a fully open 1.0 Tesla system that may be suited for cardiac iMRI, although experience in using it for cardiac imaging is still at its very beginnings.
Most up-to-date scanners are commercially equipped with interactive user interfaces and fully refocused steady-state free precession sequences that are suited for real-time imaging. During imaging, the interventionalist has to compromise between temporal resolution, spatial resolution, and signal-to-noise ratio. Most groups operate at an acquisition frame rate of 1–10 frames per second and a maximum pixel matrix of 192 × 256. The implementation of ultra-fast sequences for ventricular function analysis, magnetic resonance angiograms or real-time flow measurements are an important add-on to iMRI [1,2]. Finally, investigational iMRI laboratories should be operated in conjunction with X-ray fluoroscopy, either as an adjunct to MRI or as a safety back-up.

Catheter tracking – is it finally getting there?
Extensive research in recent years has resulted in some promising solutions that allow fast and patient-safe catheter tracking. Catheter tracking is typically classified as passive, hybrid, and active approaches.
Passive tracking devices produce negative contrast (dysprosium, carbon dioxide, iron oxide) or positive contrast (gadolinium, fluorine-19), are considered safe, and are simple to manufacture. Although passive devices often generate only modest signal contrast to their background, they are, if carefully chosen in respect to the requirements of an interventional procedure, well suited for a variety of applications (Figure 1).

Endovascular implants
Endovascular implants currently in use are too numerous to be individually discussed in this report. Generally, devices made from nitinol alloys produce only slight local susceptibility artifacts on MRI (Figure 2). Devices made from platinum generate moderate artifacts that can be limited by using appropriate pulse sequences. However, the implantation of devices with high stainless steel content, which causes severe image distortion, should be avoided in patients in whom repeat MRI examinations are planned.

Experimental application and first steps towards clinical investigation

MRI-guided cardiac catheterization
Catheterization guided by magnetic resonance imaging was one of the first MRI-guided investigations to be performed in a clinical setting. This approach is appealing because it allows simultaneous derivation of pressures, blood flow, and ventricular volume data for the assessment of pulmonary arterial resistance or ventricular pressure–volume loops (Figure 1) [6–8].

Atrial and ventricular septal defects
Many large centers have partially or completely abandoned X-ray fluoroscopy for the closure of atrial septal defects. MRI-guided transcatheter closure of atrial septal defects is of potential clinical interest, because there would be no need to sedate the patient as for closure under transesophageal echocardiography guidance. Several animal studies and a clinical landmark report have demonstrated that atrial septal defects can be effectively closed under iMRI using the self-centring Amplatzer Septal Occluder [9–11]. However, to make the step into clinical application, convincing proof of the following will be necessary: appropriate positioning of the device in defects with a small rim or within large atrial aneurysms, and exclusion of minor residual shunts after implantation. To our knowledge, there are no reports of successful MRI-guided transcatheter closure of ventricular septal defects.

Aortic coarctation and aortic aneurysm
Magnetic resonance imaging is the method of choice to assess the 3-dimensional course and wall morphology of the aorta in conjunction with functional data – an essential prerequisite for successful intervention. For this reason, the aorta has been subjected to extensive iMRI research. Several animal studies have been able to demonstrate the feasibility of iMRI-guided stent placement in aortic coarctation [12,13]. Moreover, these studies showed that potential inadvertent effects, such as aneurysm formation or even wall rupture, can be detected and adequately treated under iMRI (Figure 3) [14]. Treatment of aortic aneurysms was reported to provide appropriate device apposition, restore normal lumen contour, and enable laminar luminal flow without any internal endo leaks [14].

Pulmonary artery stenosis
Similar to aortic coarctation, pulmonary stenosis can be accurately and non invasively assessed before and after intervention using MRI. However, the tortuous course of catheters through the right ventricle and pulmonary arteries requires high-end MRI catheter tracking solutions that are not yet readily available, and therefore the amount of practical experience is limited. Nevertheless, iMRI-guided stent placement and angioplasty of the pulmonary arteries have been successfully performed in animal studies [16,17].

Pulmonary and aortic valved stents
Percutaneous transcatheter placement of valved stents has evolved, in the past 5 years, into an exciting and dynamic field of research. MRI has been shown to be particularly useful, not only in postinterventional assessment of function, but also in guiding the implantation of valved stents under conditions of complex anatomy, such as that of the coronary arteries or the mitral valve (Figure 2) [18,19].

Transvascular drug delivery
Local growth factor, gene, and cell therapy have raised great expectations as effective means of treating myocardial disease. Intramyocardial injection was demonstrated to be efficient and is routinely performed under X-ray guidance. However, X-ray has distinctive disadvantages in comparison with MRI: it does not visualize soft tissue or the distribution of the substances delivered, and suffers inaccuracy in targeting the region of interest. In contrast, MRI targets dyscontractile myocardium or fibrous scar tissue, or both (Figure 1). Moreover, animal studies have demonstrated that MRI-guided myocardial drug delivery enables monitoring of the substances delivered, in multiplanar views of the dynamic distribution over time [20,21].

Electrophysiological studies
Magnetic resonance imaging-guided electrophysiological studies can be expected to have major future clinical impact that is based on the elimination of often very high doses of ionizing radiation to the patient and the medical staff, and a more comprehensive 3-dimensional view of mapping conducting pathways. Successful iMRI acquisition of intracardiac electrocardiograms and the feasibility of radiofrequency ablation, in addition to the visualization of scar tissue after ablation, have been demonstrated in animal models [22,23].

Summary
During recent years, MRI techniques have emerged for real-time visualization of anatomy, and for ultra-fast assessment of cardiovascular function. These features have made MRI into a dynamic modality that is suited to guiding endovascular interventions and to providing immediate information on the physiologic response to pharmacological or catheter-based treatment. Nevertheless, continuing efforts are needed to resolve the technically challenging problems of iMRI, which are as plentiful as the promises held out by this emerging imaging modality.

Acknowledgement
I thank A. M. Gale for her editing support.

REFERENCES