Precision of Forty Slice Spiral Computed Tomography for Quantifying Aortic Valve Stenosis: Comparison With Echocardiography and Validation Against Cardiac Catheterization

Imaging of cardiac valves by computed tomography

This paper describes "how to" examine cardiac valves with computed tomography, the normal, diseased valves, and prosthetic valves. A review of current scientific literature is provided. Firstly, technical basics, "how to" perform and optimize a multislice CT scan and "how to" interpret valves on CT images are outlined. Then, diagnostic imaging of the entire spectrum of specific valvular disease by CT, including prosthetic heart valves, is highlighted. The last part gives a guide "how to" use CT for planning of transcatheter aortic valve implantation (TAVI), an emerging effective treatment option for patients with severe aortic stenosis. A special focus is placed on clinical applications of cardiac CT in the context of valvular disease.

Comparison of cardiac computed tomographic angiography to transesophageal echocardiography for evaluation of patients with native valvular heart disease

Retrospectively gated helical cardiac computed tomographic angiography (CCTA) has been reported accurate in the evaluation of isolated valvular abnormalities, but its ability to provide comprehensive assessment of common valvular lesions is not established. We evaluated 56 consecutive patients undergoing 64-detector retrospective electrocardiogram-gated CCTA and transesophageal echocardiography for the presence of aortic and mitral stenoses, aortic and mitral regurgitations, mitral valve prolapse, and tricuspid regurgitation. Two cardiac computed tomographic angiographic readers measured maximum aortic and mitral valve opening areas, assessed for aortic or mitral valve regurgitant area, and evaluated for mitral valve prolapse. Tricuspid regurgitation was assessed by the contrast ratio of the inferior vena cava to the right heart. After excluding nondiagnostic valves on CCTA (mitral valve n = 4, aortic valve n = 2), the sensitivity, specificity, positive predictive values, and negative predictive values of CCTA compared to transesophageal echocardiography were 100%, 96%, 50%, and 100% for aortic stenosis, 44%, 96%, 67%, and 90% for aortic regurgitation, 100% each for mitral stenosis, 13%, 95%, 80%, and 45% for mitral regurgitation, and 50%, 98%, 80%, and 91% for mitral valve prolapse. There was no relation between tricuspid regurgitation grade and contrast ratio (p = 0.53). There was excellent interobserver agreement for aortic and mitral stenoses (kappa = 1.0 for each), and good agreement for aortic regurgitation, mitral regurgitation, and mitral valve prolapse (kappa = 0.81, 0.78, and 0.88, respectively). In conclusion, CCTA exhibited high diagnostic performance for detection of aortic and mitral stenoses and limited diagnostic performance for aortic regurgitation, mitral regurgitation, and mitral valve prolapse; tricuspid regurgitation could not be evaluated. The ability of CCTA to provide comprehensive assessment of valvular function is variable.

Planimetry of the aortic valve orifice area: comparison of multislice spiral computed tomography and magnetic resonance imaging

OBJECTIVE

We sought to determine the comparability of multislice computed tomography (MSCT) and magnetic resonance imaging (MRI) for measuring the aortic valve orifice area (AVA) and grading aortic valve stenosis.

MATERIALS AND METHODS

Twenty-seven individuals, among them 18 patients with valvular stenosis, underwent AVA planimetry by both MSCT and MRI. In the subset of patients with valvular stenosis, AVA was also calculated from transthoracic Doppler echocardiography (TTE) using the continuity equation.

RESULTS

There was excellent correlation between MSCT and MRI (r = 0.99) and limits of agreement were in an acceptable range (± 0.42 cm(2)) although MSCT yielded a slightly smaller mean AVA than MRI (1.57 ± 0.83 cm(2) vs. 1.67 ± 0.98 cm(2), p < 0.05). However, in the subset of patients with valvular stenosis, the mean AVA was not different between MSCT and MRI (1.05 ± 0.30 cm(2) vs. 1.04 ± 0.39 cm(2); p > 0.05). The mean AVAs on both MSCT and MRI were systematically larger than on TTE (0.88 ± 0.28 cm(2), p < 0.001 each). Using an AVA of 1.0 cm(2) on TTE as reference, the best threshold for detecting severe-to-critical stenosis on MSCT and MRI was an AVA of 1.25 cm(2) and 1.30 cm(2), respectively, resulting in an accuracy of 96% each.

CONCLUSION

Our study specifies recent reports on the suitability of MSCT for quantifying AVA. The data presented here suggest that certain methodical discrepancies of AVA measurements exist between MSCT, MRI and TTE. However, MSCT and MRI have shown excellent correlation in AVA planimetry and similar accuracy in grading aortic valve stenosis.

Evaluation of cardiac valves using multidetector CT

Cardiac CT is an accurate and reasonable alternative modality for valvular imaging. It is used primarily for the evaluation of coronary artery disease; however, important information regarding valvular anatomy and function can be derived from CT. Calcification is a common CT finding in various valvular abnormalities and carries important diagnostic and prognostic value. In addition, valvular morphology, stenosis, and regurgitation also are detected on contrast enhanced scans, with good correlation with trans-thoracic echocardiography and other techniques.

Mitral annular shape, size, and motion in normals and in patients with cardiomyopathy: evaluation with computed tomography

OBJECTIVE

To assess prospectively, in healthy subjects and in patients with dilated cardiomyopathy (DCM) and hypertrophic obstructive cardiomyopathy (HOCM), the 3-dimensional (3D) shape, size, and motion of the mitral annulus (MA) using computed tomography (CT).

MATERIALS AND METHODS

Twenty patients with no cardiac abnormalities (referred to as normals), 15 with DCM, and 15 with HOCM as determined by echocardiography underwent contrast-enhanced, retrospectively electrocardiography (ECG)-gated 64-slice CT of the heart. The MA was manually segmented in 10% steps of the RR interval with dedicated 3D software employing the point-wrap algorithm. The MA shape, area size, change of the MA area, and apicobasal MA motion throughout the cardiac cycle was determined and compared between the groups. Intercommissural distances were measured with CT and compared with findings during surgery in 9 patients undergoing ring annuloplasty.

RESULTS

The MA was nonplanar in all phases and subjects, being largest in diastole and smallest in systole. The MA area was significantly (P < 0.001) larger in patients with DCM (11.5 +/- 4.1 cm/m) as compared with normals (5.5 +/- 0.9 cm/m) and HOCM (4.7 +/- 0.9 cm/m). The change of MA area throughout the cardiac cycle was significantly (P < 0.017) smaller in patients with DCM (12.2 +/- 3.3%/m) as compared with normals (20.0 +/- 7.9%/m) and HOCM (20.5 +/- 7.7%/m). The mean apicobasal motion was significantly (P < 0.017) smaller in patients with DCM (2.2 +/- 1.0 mm/m) as compared with normals (3.6 +/- 0.8 mm/m) and HOCM (2.7 +/- 0.7 mm/m). Intercommissural distances as determined by CT showed a good correlation (r = 0.68, P < 0.05) with intraoperative measurements (mean difference, 0.44 mm; limits of agreement, -2.73-3.62 mm).

CONCLUSION

Our study provides in vivo human data on the 3D shape, size, and motion of the MA in healthy subjects. Significant changes in size and motion of the MA were noted in patients with HOCM.