Virtually all prosthetic heart valves (PHV) are considered safe in the magnetic resonance (MR) environment at field strengths of up to 1.5 T (Figure 1 A-H).
(A) Starr-Edwards ball-and-cage valve.
(B)Medtronic-Hall monoleaflet mechanical valve.
(C) St Jude bileaflet mechanical valve (LVOT and short axis views).
(D) Medtronic Mosaic bioprosthetic valve (LVOT and short axis views).
(E) Medtronic Freestyle subcoronary stentless bioprosthetic valve (LVOT and short axis views).
(F) Transcatheter aortic Medtronic Core Valve .
(G) Transcatheter aortic Edwards Sapien Valve.
(H) Transcatheter pulmonic Edwards Sapien Valve
(Figure 1) - Examples of different types of PHV in aortic position (except for H) as imaged by CMR (balanced SSFP sequence). Virtually all PHV are safe in the MRI environment at field strengths up to 1.5T.
There is no evidence to justify withholding a cardiac or extracardiac MR study at 1.5T in such a patient just due to the presence of a PHV. Most of the PHV are also safe at 3T, but testing is less widely available at 3T. In case of any doubt or for documentation purposes, online resources (e.g., www.mrisafety.com) are available for checking on safety of any devices for MR imaging.
The quality of images and therefore the diagnostic capability of a cardiovascular magnetic resonance (CMR) study in a patient with PHV depend on artefacts from the metallic components of the PHV. Cine sequences and phase contrast velocity mapping sequence are essential for assessment of valve function and quantification of velocity and flow. These sequences are gradient echo sequences and thus susceptible to artefacts caused by ferromagnetic metals. Spin echo sequences are less prone to these artefacts but provide only static images.
The daily workhorse cine sequence in CMR is balanced steady state free precession (SSFP) which is very sensitive to artefacts from ferromagnetic objects. The severity of artefacts in cine images caused by PHV may range from no artefact to severe artefact obscuring not only the valve but also the adjacent structures. Therefore from a CMR perspective the ferromagnetic metallic component of the PHV is the main factor determining cine image quality. While one may think that mechanical PHV will cause the more severe artefacts and bioprosthetic PHV will be friendlier with the image quality, this is not always the case. The metallic component of a mechanical valve may not be highly ferromagnetic whereas the very small amount of metal used in the supporting stent of a bioprosthetic valve may cause significant image distortion (Figure 2 A, B).
(A) Two different types of mechanical heart valves on the right and left side displaying how the artefact caused by the metallic heart valves may differ in severity. On the left side are 2 chamber and short axis images of a bileaflet mechanical valve with a minimal artefact from the metallic component. Even the details of the leaflet opening and closing can be clearly seen. The valve on the left side however causes significant artefact which not only impairs the image of the valve itself but also extends to the aortic flow.
(B) Image quality of bioprosthetic heart valves also differ depending on the ferromagnetic metal component used in the stents. Two different aortic stented bioprosthetic valves in LVOT and short axis views are displayed; one with no image artefact (left side) and the other with significant image impairment (right side) almost as severe as seen with a mechanical prosthetic valve (right side).
Even the annuloplasty rings can cause significant image distortion when they contain metal in their structure (Figure 3).
(Figure 3) - Even the annuloplasty rings cause significant image artefacts when metals are included in their structure. An annuloplasty ring implanted for functional mitral regurgitation causing significant artefact (LVOT and short axis views).
Stentless bioprosthetic valves are essentially metal free and do not cause any artefacts, hence their appearance is not much different than native valves (Figure 4).
(Figure 4) - Stentless bioprosthetic valves are not different than native valves from an imaging perspective and can be assessed reliably by CMR. A Toronto stentless valve is shown, the images are devoid of any artefact and the bioprosthetic valve appears almost like a native aortic valve.
The image quality of CMR scans for the new percutaneous valves are reasonably good as well (Figure 1 F, G, H). When the cine images with balanced SSFP are significantly deteriorated due to artefacts from the PHV, a spoiled gradient echo sequence which is less prone to these artefacts can be used to acquire cine images, albeit at the expense of a lower signal to noise ratio (Figure 5).
Spoiled gradient echo
(Figure 5) - Balanced SSFP sequence is very susceptible to artefacts caused by ferromagnetic metals. A spoiled gradient echo sequence can also be used to acquire cine images which is relatively less susceptible to such artefacts and may yield better image quality.
Cine images are also essential for reliable assessment and monitoring of ventricular size and function after valve replacement (Figure 6 A, B).
(A)This figure shows CMR images of a patient with congenitally corrected transposition of great arteries who had severe tricuspid regurgitation and dilatation of the systemic right ventricle. The patient had tricuspid valve replacement with significant improvement of the systemic right ventricular volumes.
(B) Quantification of ventricular volumes revealed significant improvement after tricuspid valve replacement.
(Figure 6) - CMR is a robust technique for assessment of ventricular volumes and ejection fraction. Accurate calculation of ventricular volumes is not only important for timing of valve replacement but also for follow-up after surgery.
Phase contrast (PC) imaging:
These images are essential to acquire velocity and flow data. Depending on the type of the PHV, there may be a signal void region below and above the valve. Imaging slice for through plane images is positioned just above this signal void level to avoid the artefact. It is easier to use through plane images to quantify flow and velocity for the aortic and pulmonary valves as the borders of flow through these valves (the ascending aorta and pulmonary artery) can be traced (Figure 7 A, B, C and D).
Phase contrast image
(A) In this image a mechanical aortic valve with minimal artefact is seen, the imaging plane for through plane can be placed just above the valve.
(B) A bioprosthetic valve with no artefact is seen. Since the valve does not cause any artefact reliable PC images can be acquired even at the valve level.
(C-a)Magnitude and PC images of a mechanical and bioprosthetic valve.
(C-b)Magnitude and PC images of a mechanical with artefacts from the metallic components of the valve.
(C)Magnitude and PC images of a mechanical (a) and bioprosthetic valve (b) with artefacts from the metallic components of the valve. The through plane PC image level should be just above the valve level to avoid the artefact area. Since the maximum velocity through a stenotic valve is downstream of the valve level, this generally does not cause much problem.
(D)A systolic still frame from the PC image of the bioprosthetic valve (b) showing the appropriate level of through plane image prescription just above the valve level.
(Figure 7) - PC sequence is used to derive velocity and flow data through any valve or vessel. For the aortic and pulmonary valves a through plane PC image just above the valve is most reliable to detect maximum velocity. This should be just above the level of the metallic artefact. For mitral and tricuspid valves through plane images are less useful and therefore not commonly acquired, however in-plane PC images can still provide important information on valve function (see Figure-15 C).
For the mitral and tricuspid valves however phase contrast images are mainly used for qualitative assessment.
Spin echo sequences:
These are least prone to artefacts from ferromagnetic objects, but they can only be used to acquire static images. As such they cannot be used to assess function of the PHV; however when any structural abnormality related to a PHV is suspected spin echo images can be very useful (Figure 8 A, B).
Cine (gradient echo)
(A) This patient had submitral pseudoaneurysm (*) following mitral valve replacement. The cine image is affected from the metallic artefact whereas spin echo image provides good anatomical image without any artefact.
Cine (gradient echo)
(B) Another example showing usefulness of spin echo images for assesment of aortic root in a patient with mechanical aortic valve prosthesis.
(Figure 8) - Spin echo sequences are less effected from artefacts caused metallic structures of the valves but can only provide static images. Nevertheless these may be valuable when structural abnormalities are assessed.
Usefulness of CMR for assessment of patients with PHV
No established guidelines or recommendations exist about when to use CMR for the assessment of PHV. Doppler echocardiography is essentially the first line of imaging modality for assessment of patients with PHV. However, CMR can provide invaluable information in most of the patients with PHV. This may include information on one or more of the following:
Prosthetic valve itself
Quantitative data that can be derived from CMR for assessment of valvular heart disease are summarized below (see relevant native valve disease sections). All these approaches are essentially valid for the assessment of PHV function, and planning of the images are the same except for minor alterations to avoid artefacts. The reliability of the assessments, however, will depend finally on the quality of images that can be acquired.
Valve stenosis (Figure 9-10):
Valve area planimetry can be calculated from the cine images and the maximum velocity through the valve can be derived from PC velocity mapping.
Bioprost SEvere AS LVOT
Bioprost SEvere AS LVOT PCC
Bioprost SEvere AS Planimetry
(Figure 10) - Another example of a bioprosthetic valve with moderate stenosis.A pulmonary tissue valve in a 27 years-old female patient. RVOT cine and in-plane PC images (a and b) suggest stenosis of the valve, the maximum velocity was 3.4 m/s. A cross cut through the valve leaflets level (c) showing the opening of the valve leaflet. Valve area by planimetry was 1.2 cm2.
Valvular regurgitation (Figure 11-14):
Regurgitation volume and fraction can be calculated from PC velocity mapping of aortic and pulmonary flow (for aortic and pulmonary regurgitation, respectively), from difference between left and right ventricular strokes volumes or from difference between ventricular stroke volumes and the corresponding outflow (such as difference between LV stroke volume and aortic flow derived from aortic PC image to calculate mitral regurgitant volume). Also secondary findings of valvular regurgitation can be imaged (e.g., appearance of regurgitation jet severity, diastolic flow reversal in descending aorta in aortic regurgitation)
(B) In-plane and through plane magnitude and PC images. Aortic flow curve derived from the through plane PC image demonstrated regurgitation volume of 55ml and regurgitation fraction of 45% consistent (aortic valve flow-time graph on bottom left)with moderate to severe aortic regurgitation.
(C) Flow reversal in descending aorta is readily seen in cine images suggesting significant aortic regurgitation.
(Figure 11) - An aortic bioprosthetic valve with degeneration, central coaptation defect and severe aortic regurgitation.
paravalvular LVOT XC regurgitataion 2
paravalvular regurgitataion-through PC
paravalvular SAX regurgitataion 1
Mechanical aortic valve and the ascending aorta were normal (a). There was severe regurgitation of the pulmonary homograft valve (b). In-plane and through plane PC images showing significant, almost free pulmonary regurgitation (c and d). Pulmonary flow curve derived from the through plane PC image showing moderate to severe pulmonary regurgitation, regurgitation fraction 40% (e).
(Figure 13) - A 45 years-old male who had Ross procedure for bicuspid aortic valve. He then had mechanical aortic valve replacement for failing aortic homograft. Patient was referred for a CMR study for assessment of mechanical prosthetic aortic valve and pulmonary homograft valve.
(A)Cine images showing dehiscence and paravalvular leak on the aortic side of the mitral annulus. LV stroke volume was 121ml and RV stroke volume was 65 ml resulting in a regurgitation volume of 56ml.
(B)Magnitude and the corresponding in plane PC images showing a large jet suggesting a s