952 BRIEF REPORTS

and D points during either late systole or throughout systole (pansystolic) as suggested by previous investi- gators.2*4 M-mode echocardiography was conducted by using standard echocardiographic equipment. Par- ticular attention was given to place the transducer perpendicular to the chest wall at the left parasternal area to avoid technically related false-positive results as cautioned by Markiewicz et al.5 Levels of statistical significance were determined by paired t test.

In the 16 patients with auscultatory evidence of ‘MVP, mean LV end-diastolic dimensions (mean f standard error of the mean) were 45.6 f 0.9 mm in the supine position and 47.5 f 0.9 mm in the left lateral position. The difference of mean LV end-diastolic di- mensions in these 2 positions was statistically signifi- cant (p <O.OOl). In the left lateral position, M-mode echocardiographic findings were considered unequi- vocally negative for MVP in 7 patients and nondiag- nostic in the remaining 9 patients because of insuffi- cient degree of posterior displacement of the leaflets during the systole.

Figure 3 illustrates M-mode echocardiographic tracings recorded at the mitral and LV levels in pa- tient 1. As shown in the right panel, mitral valvular motion appeared normal during the systole when the measured LV end-diastolic dimension was 46 mm in the left lateral position. The left panel shows the re- cords obtained in the supine position, and diagnostic pansystolic posterior displacement was shown when the LV end-diastolic dimension was reduced to 43 mm in this position.

Figure 2 depicts the result in patient 4. Mitral val- vular motion obtained in the left lateral position pre- sents uncertain and nondiagnostic posterior displace- ment when the LV dimension measured 44 mm. In the supine position, diagnostic pansystolic posterior dis- placement of the mitral leaflets was demonstrated when the LV dimension was reduced to 40 mm.

In the 10 patients without any auscultatory evi- dence of MVP, M-mode echocardiographic findings

were negative for MVP in both left lateral and supine positions.

In our study, a positional change from the left later- al position to the supine position appeared to have provoked MVP or altered equivocal results to un- equivocal outcomes. To compare the LV size on 2 dif- ferent positions, LV end-diastolic dimensions were measured in both supine and left lateral positions. Al- though the difference in LV dimensions in these 2 positions was small in a given patient, LV end-diastolic dimensions were either unchanged or, in most pa- tients, increased in the left lateral position. These dif- ferences in our 16 patients were statistically significant (p KO.001). In our patients, even a small magnitude of reduction in LV size could have provoked echocardio- ,graphic evidence of MVP. In this study, particular at- tention was given to placing the transducer perpendic- ular to the chest wall to avoid artifactual induc’tion of MVP in both supine and left lateral positions. Our observation was based on a few selected patients, and it is not certain that measured LV end-diastolic dimen- sions in 2 different positions actually reflect a change in the LV size. Further studies in a large number of patients and preferably in conjunction with %dimen- sional study could substantiate the validity of our observation.

References 1. Ruwitch JF. Weiss AN, Fleg JL, McKntght RC. Ludbrook PA. Insensitivity of echocardiography in detecting mitral valve prolapse in older patients with chest pain. Am J Cordid 1977;40:686-690. 2. Noble LM, Dabesteni A, Child JS, Krtvokepich J. Mitml valve prolapse. Cross sectional and provocative M-mode echocardiography. Chest 1982; 82:158-163. 3. Sabn DJ, DeMaria A, Kiss10 J, Weyman A. Recommendations regarding quontitotion in M-mode echocardiogmphy: results of a survey of echocardio- graphic measurements. Circulation 1676;58:1072-1083. 4. Popp RL. Brown OR, Silverman JF. Harrison DC. Echocardiographic ab- normalities in the mitral valve proiopse syndrome. Circulation 1974; 49:428-433. 5. Marktewicz W, London E. Popp RL. Effect of transducer placement on echocordiographic mitral valve motion. Am Heart J 1978;98:555-556.

Mitral Stenosis: Imaging by Nuclear Magnetic Resonance

JAMES A. HILL, MD E. WILLIAM AKINS, MD

JEFFREY Ft. FITZSIMMDNS, PhD C. RICHARD CDNTI, MD

N uclear magnetic resonance (NMR) imaging of the’ heart has been reported1 as being capable of imaging

From the Division of Cardiology and Department of Radiology, University of Florida and Gainesville Veterans Administration Medical Center, Gainesville, Florida 32610. This work was sup- ported in part by Grant P41 RR02278 from the National Institutes of Health, Bethesda, Maryland. Manuscript received February 20, 1985; revised manuscript received June 21, 1985, accepted June 24,1985.

various abnormalities of cardiac anatomy. Cardiac chambers and great vessels can be imaged in most patients using electrocardiographic gating. There are several advantages of NMR that make it potentially useful for imaging patients with valvular heart disease. NMR is not limited by air or bone conduction as is echocardiography, and potentially more patients can be imaged. NMR has a reasonably high spatial resolu- tion and has been shown to be capable of imaging the valve leaflets2 In addition, the potential exists for as- sessing relative blood flow using differential signal intensity.3 Van Dijks has shown how flow velocity im- ages may allow assessment of blood flow. This report develops evidence that NMR imaging can provide useful information about cardiac anatomy and phys- iology in patients with proved mitral stenosis (MS).

Five patients with ‘clinical and hemodynamic evidence of MS were studied. Some clinical and he- modynamic data are listed in Table I. A 0.15Tesla

February 1, 1986 *HE AMERICAN ,‘O~J?NAL OF CAPDlO!BGY Volume 57 353

TABLE I Ciinicai and Hemodynamic Parameters in Five Patients with Millrai Stenosis ._.- ---_.-_- ._.____-__

Pressure:; (mm ,Hg) Ventric- ._.- -- -_.____.______ PAW-L.‘/

Age (~0 ular Rate RA PA PAW MDG Cl (liters/ MVA Pt 8 sex Rhythm (beatslmin) (mean) RV (mean) (mean) LV Aorta (man) min/m*) (cm2)

_-.-

1 49M Sinus 105 (7) 13817 138/54(80) (30) 14018 140/90 (25) 2.9 1.7 2 58M Sinus 75 (5) 5317 53/20(35) (20) 128112 128/12 (14) 2.9 1.5 3 59M Sinus 108 (6) 5518 55/25(35) (2V 13018 130175 (13) 3.4 1.8 4 82F AF 111 (10) 85112 85/35(40) (‘W 140115 140/80 (1% 3.1 1.3 5 32M Sinus 88 (5) 5517 55/30(41) cw 110112 llOI70 (17) 3.0 1.2

AF = atrial fibrillation; Cl = cardiac index; LV = left ventricle: MDG = mean diastolic gradient; MVA = mitral valve area; PA = pulmonary artery: PAW = pul- monary artery wedge; RA = right atrium: RV = right ventricle.

resistive magnet (TechnicareQ) was used with a high- performance, half-saddle coil in the receive-only mode. A spin echo pulse sequence was used and a single echo was obtained at 30 ms. Electrocardio- graphic gating was used and a multislice (IO-slice) ac- quisition obtained. images were acquired on a 128 X 256 matrix providing spatial resolution of 1.4 X 2.8 mm with a slice thickness of 10 mm. Each acquisition step was repeated 4 times to improve the signal-to-noise ratio. Planes modified to image the heart along its axes were used to simulate more conventional cardiac im- aging techniques and produce images best depicting the left ventricle and mitral valve.5

AII patients were successfully imaged. The average acquisition times were II minutes per view with a total time in the magnet of approximately 60 minutes per patient. Two patients had 2 views (sagittal long axis and coronal right anterior oblique), 2 patients had 3 views (sagittal long axis, transverse short axis and cor- onal right anterior oblique), and z patient 3 other views

(transverse, sagittal long axis and coronal right anteri- or oblique).

AII patients had cardiac abnormalities visualized by NMR but not all abnormalities were present in all patients in all views. The sagittal long-axis view, which corresponds to the apical $-chamber view of 2- dimensional echocardiography, gave the best anatom- ic definition. The mitral valve, which is not always well visualized in normal subjects, was easily seen in all patients (Fig. I). In addition, clear chordal struc- tures were present in 4 of5 patients (Fig. 2). Short-axis planes through the mitral valve (transverse short axis) were used in 2 patients but did not show valvular struc- tures clearly enough to measure valve area presum- ably because of slice thickness (Fig. 3). The left atrium was large in all patients in relation to the aorta (Fig. 4). with bulging of the atrial septum in 2 patients (Fig. 1). Zncreased signal intensity of the contents of the left atrium during diastole was seen in all patients (Fig. 1). However, no patient had atrial thrombi at operation. The pulmonary artery was enlarged compared with the aorta in 2 patients and the right ventricle was thickened in 1 patient in relation to the left (Fig. 5).

FiGURE 1. Sagtttal long-axis view showing the 4 chambers and clear mitral valve structures (arrow). There is a -8 in- crease in signal intensity comparing the iett atrtum (LA) with the right atrium (RA).

FlGURE 2. Sagittai long-axis view showing chordal structures (arrows).

354 BRIEF REPORTS

FIGURE 3. Transverse short-axis view showing mitral valve leaflets (arrows). RV = right ventricle.

FIGURE 5. Saglttal lofig-axis vlew showlng thlckened rlght ventricle (RV) In relatlon to the left ventricle.

FIGURE 4. Transverse short-axis vlew of the aorta (Ao) and left atrlum (LA) showlng enlarged left atrium In relation to the aorta.

NMR imaging offers a new approach to diagnosis in patients with MS. This study demonstrates that mor- phologic features known to exist in such patients can be imaged by NMR. Quantitation, although not per- formed in these patients, is possible. High signal inten- sity within the left atrium during diastole was always observed and may provide a means of measuring flow, although we did not use special techniques that would

have better demonstrated this6 The ability of NMR to assess flow quantitatively in vivo has not yet been shown. There are several disadvantages in NMR and its application to MS that deserve mention. When a slice thickness of 10 mm is used, partial volume effects are significant. This may hinder assessment of valve area. Current NMR imaging technology does not pro- vide real-time images, making assessment of valve motion difficult. As NMR technology improves and other acquisition techniques are evaluated and per- fected, NMR may be a useful alternative to echocardi- ography in select patients in the noninvasive assess- ment of MS.

References 1. Higgins CB, Kaufman L, Crooks LE. Magnetic resonance imaging of the cardiovascular system. Am Heart J 1985;109:136-152, 2. Higgins CB, Stark D, McNamere M, Lenzer P, Crooks LE. Kaufman L. Multiplane magnetic resonance imagfng of the heart and major vessels stud- ies in normal vohrnteers. AJR 1984;142:661-667. 3. Bredley WG, Waluch V. Blood flow: magnetic resonance imaging. Radiolo- my 1985;154[2]:443-450. 4. Van Dijk P. Direct cardiac NMR imaging of heart wall and blood flow vefocity. J Comput Assist Tomogr 1984;8(3)429-436. 5. Akins EW, Hill JA, Fitzsbnmons JR, Pepine CJ, Williams CM. Magnetic resonance imaging of the left ventricle: the importance of imaging plane. Am J Cardiol 1985;56:366-372. 6. Wehrli FW. MacFall JR, Axe1 L, Shutts D. Glover GH, Herfkins RJ. Ap- proaches to in-plane and out-of-plane flow imaging. Noninvasive Medical Imaging 1984;1(2]:127-136.