Evaluation of mitral stenosis by cine magnetic resonance imaging

Evaluation resonance

of mitral stenosis imaging

by tine magnetic

To evaluate the ability of tine magnetic resonance imaging (tine MRI) in the assessment of mitral stenosis (MS), we studied 20 patients (14 women and 6 men, mean age 60.6 & 8.5 years) with rheumatic mitral valve stenosis by using an 0.5 T magnet. Cine MRI showed several signs of MS. Mitral leaflet thickening, reduced diastolic opening, and abnormal valve motion toward the left ventricular outflow tract were all common features. MS was also characterized by an abnormal diastolic transmitral signal from blood. Both left atrial and left ventricular dimensions were similar to those obtained at two-dimensional echocardiography (2-DE) (r = 0.89 and r = 0.86, respectively; p < 0.001). A significant relationship was also found between the maximum mitral leaflet separation measured by tine MRI in diastole and the mitral valve area as calculated using the pressure half-time method and continuous wave Doppler (r = 0.81; p < 0.001). These data indicate the improved ability of MRI to detect and assess MS and also suggest that this technique may contribute to the noninvasive assessment of MS. (AM HEART J 1992;123:1252.)

Gian Carlo Casolo, MD, PhD,a Virna Zampa, MD,b Luigi Rega, MD,b Luciano Berti, MD,C Maurizio Filice, MD,C Nicola Picchione, MD,” and Loredana Poggesi, MD.a Florence, Italy

Magnetic Resonance Imaging (MRI) is a newly developed noninvasive technique capable of studying both the anatomy and the function of the heart.lr2 MRI has been used to image a variety of cardiovascular diseases and has been shown to be a reliable technique for obtaining clinically relevant diagnostic information. In its former applications this technique required excessively long acquisition times to provide functional information from the heart and was therefore mainly used to study anatomy. Moreover, the heart valves could not be explored routinely because of their weak signal on spin echo images. Recently, the development of imaging techniques based on short repetion times, partial flip angles, and gradient refocusing have permitted the acquisition within a few minutes of several frames of the cardiac cycle, which can be shown in a tine format.324 With this technique reliable measures of ejection fraction, ventricular volumes, and cardiac mass can be obtained in a short time.4-7 On tine MRI images, the strong signal determined by the flowing blood highFrom rence; Received Reprint gagni, 4/I/35742

1252

“Clinica Medica and Tkrdiology for publication requests: 85, 50123

1 and bDepartment Unit, Ospedale June

San

21, 1991;

Gian Carlo Casolo, Florence, Italy.

MD,

of Radiology, Giovanni

accepted PhD,

Nov. Clinica

University di Dio.

of Flo-

1, 1991. Medica

1, Vle Mor-

lights the internal architecture of the heart and particularly the cardiac valves, which can now be appreciated. The study of these structures is greatly enhanced by tine MRI, showing both motion and function. MRI has been used to image several valvular diseases. With the advent of tine MRI, particular attention has been directed toward the study of aortic and mitral regurgitation.4y 8lg However, there are only limited data on mitral stenosis (MS),10-12 and to date no study on the use of tine MRI has been reported for this disease. This investigation was designed to study the diagnostic features of MRI and tine MRI in MS and to relate these findings with those obtained by two-dimensional and Doppler echocardiography (2-DE). METHODS Patients

investigated. We studied 20 consecutive patients (14 women and 6 men, mean age 60.6 +- 8.5 years) having rheumatic mitral valve stenosis,who gave their informed consent to this investigation. All the patients were under twice-a-year medical control and the diagnosiswas basedupon clinical and echocardiographicdata. Nine subjects were in sinus rhythm, while the remaining 11 had atrial fibrillation with a ventricular rate below 110 beats/ min. Eight patients alsohad a mild-to-moderate degreeof mitral regurgitation. Nine subjects were known to have tricuspid regurgitation (six mild, three severe) (Table I).

Volume Number

123 5

Table

Cine MRI in mitral

1. Characteristics

of the 20 study patients Sex

1 2 3 4 5 6 I 8 9 10 11

45 54 48 67 68 58 62 65 71 66

59

12 13 14 15

68 70 64 65 54 49 45 65 68

16 17 18 19

20 Total

F F F M

F F

M F F

M F F F

F M M F

M F F 14F 6M

Rhythm

SR AF SR AF AF SR SR AF SR SR AF AF AF AF SR SR AF SR AF AF 11 AF 9 SR

with mitral Atria1

MR

TR

+ + (Appendage)

-

+

+ + + -

+ (Roof)

+

+ + (Appendage)

f + 8

AF, Atrial fibrillation; SR, sinus rhythm; MR, mitral regurgitation; TR, tricuspid regurgitation; LA, left atria1 ters); LV, left ventricular dimensions at echocardiography (in millimeters); +, present; -, absent.

Two-dimensional and Doppler echocardiography. All patients were also imaged by 2-DE. Either a 2.8 or a 3.5 MHz electronic transducer was used to perform both twodimensional and Doppler echocardiography (Sonos 1000, Hewlett-Packard Co., Medical Products Group, Andover, Mass.). For the purposes of this study left atrial (LA) and left ventricular (LV) dimensions were obtained in all the patients from the parasternal projection. We measured the mitral valve area on the continuous wave Doppler spectra of transmitral flow obtained from the apical view by using the pressure half-time method (PHT). Particular care was taken to detect atria1 thrombosis, which in fact was recognized in one patient on the roof of the left atrium. Pulsed wave Doppler was also performed in all the patients from the apical view. Both the right and left atria were carefully inspected with pulsed Doppler to detect and assess regurgitant systolic jets. Magnetic resonance imaging. MRI was performed within 10 days of the clinical and echocardiographic assessment of patients. We used a commercially available 0.5 T magnet (Philips Gyroscan, Eindhoven, The Netherlands) with an electrocardiographic gating system. The electrocardiographic signal was recorded and transmitted to the computer with the aid of a telemetry system (Hewlett-Packard Co.). Both spin echo and gradient echo imaging techniques were used. After an initial scan to localize the heart within the chest, two spin echo multislice scans were performed (matrix 256 X 256, field of view [FOV] = 450 mm, echo time [TE] = 30 msec, slice thickness = 10 mm, two averages). These multislice scans were adjusted to cover the entire heart volume.

1253

stenosis

thrombosis

3

stenosis

+ + 9

dimensions

LA

LV

43 44 45 48 43 52 48 47 45 46 43 45 50 44 51 42 46 52 46 44 46.6 -c 3.6

45 52 39 44 48 51 56 57 49 44 47 50 48 56 50 51 50 55 47 46 49.2 k 4.6

at echocardiography

(in millime-

Once completed, these scans were followed by two multislice gradient-echo series (matrix 128 X 256, TE 12 msec, (Y = 60 degrees, repetition time [TR] = 50 msec) on the transversal plane, respectively, of three and four levels, and were fitted so as to cover the mitral valve region without interruptions. According to the patient’s heart rate, 10 to 16 frames of the cardiac cycle were obtained at each level. Each slice was later displayed in a tine format on the screen of the imager. The mean imaging time was 45 + 15 minutes. All cardiac dimensions taken in this study were measured on tine MRI images. LA dimensions were measured from the posterior aspect of the aortic root to the distal part of the left atrium at end diastole. LV diameter was measured at the midventricular level from the interventricular septum to the posterolateral wall at end diastole. We also measured the maximal distance between the mitral leaflets or maximal mitral opening (MMO). This was accomplished by selecting the frame best showing the leaflet separation and then measuring MMO. Statistical analysis. Comparison between measurements of LA, LV, and mitral valve dimensions by 2-DE and MRI was performed using the linear regression analysis. Image quality for tine MRI and spin echo images was evaluated by two different observers who used the following qualitative scale: (a) good image quality, which indicated that small details were also recognizable; (b) diagnostic images, signifying main structures such as papillary muscles, the distal part of the mitral valve leaflets, and main pulmonary veins could be seen albeit good image quality was absent; (c) unsatisfactory images, indicating only that

1254

Casolo

et al.

American

?nay 1992 Heart Journal

Fig. 1. Spin echomultislice scanin a patient with mitral stenosis.Note the high intraluminal signaliin the left atrium. Also clearly visible is a thrombus in the left atria1 appendage (arrow).

the main cardiac structures were recognizable. In the latter case the scan had to be repeated until a better image

quality could be achieved. Reproducibility for MRI measurementswasalsoevaluated. Interobserver reproducibility was assessed by two different operators who independently performed the measures.Intraobserver reproducibility was assessed by the sameoperator by repeating all measurements

on 2 different

days. The result of inter-

observerand intraobserver variability waslessthan 5 % for LA and LV dimensions,and 8% and 7 % , respectively, for MMO. RESULTS Image analysis. Good image quality was achieved on spin echo images in 17 out of the 20 patients examined (85 % ). In spite of the presence of atria1 fibrillation in 11 subjects, a good image quality was achieved in 10 out of the 11. In the three patients without good image quality, diagnostic images were nevertheless obtained. The cause for a poor image quality in these three subjects was the presence of

dyspnea in two and a very irregular heart rate (HR) in one patient having slow atria1 fibrillation (mean HR <50/min). Both these factors decreased image quality significantly. In fact, respiratory movements usually do not greatly affect image quality when respiration is regular and at a normal rate. On the other hand, the presence of frequent and irregular respiratory movements introduces a source of motion that is not possible to eliminate, thus decreasing the signalto-noise ratio significantly. Low and irregular HRs represent a source of artifacts and image degradation because of triggering problems. To obtain static images of the heart, each line of the matrix is acquired in the same cardiac phase. This is relatively easy to accomplish when HR (i.e., the repetion time) is constant. If HR changes significantly during each cardiac cycle, each line of the matrix refers to a slightly different heart phase. This causesartifacts and again a low image quality. Although the tine display greatly improved the

Volume Number

123 5

Cine MRI in mitral stenosis

1255

Fig. 2. Cine MRI seriesin diastole in a patient with mitral stenosis.Note the interatrial septum bulging toward the right atrium. A bright transmitral flow-related signal can be appreciated in all the diastolic frames.

anatomic and functional

evaluation of gradient echo images, quality was much lower in these images than in spin echo images. As a matter of fact, a relatively lower signal-to-noise ratio compared with spin echo images is a peculiar characteristic of gradient echo images. Each image is in fact obtained in a very short time so that a full recovery of magnetization is not achieved, resulting in a weaker signal from the heart than that observed on spin echo images. Furthermore, gradient echo images are usually (and this is also the case in this study) obtained with an acquisition matrix of 128 X 128 pixels that is subsequently interpolated to a 256 pixel matrix. Therefore gradient echo images have a lower spatial resolution than spin echo images, that are usually obtained on a 256 X 256 matrix. Artifacts are also frequently encountered on gradient echo images. Gradient echo images are very sensitive to any kind of motion (flowing blood, respiration, involuntary patient move-

ments), small magnetic field inhomogeneities, and also require an optimal setup of the imager. To summarize, gradient echo images usually are of poorer quality than the corresponding spin echo images. In spite of all these pitfalls, gradient echo images are obtained in a very short time and can be shown in a tine format. The tine display, showing the motion of the heart and flowing blood, dramatically improves the information that each frame offers individually. For the above-described features of gradient echo imaging, a good image quality was obtained on a high percentage of frames in only 12 patients (60%). Six patients showed only diagnostic images (30% ). Two patients required a repetition of the scans because of an unacceptable image quality (10 % ), which lengthened their examination time. MRI findings in mitral stenosis with conventional spin-echo technique. Several abnormal findings were

observed in spin echo images. As opposed to normal

1256

Casolo et al.

American

May 1992 Heart Journal

Fig. 3. Spin echoimage(upper left) and tine MRI series in diastole in a patient with mitral stenosis. 1Tote the bright transmitral flow-related signal surrounded by dark side jets caused by turbulent flow.

subjects in whom the mitral valve leaflets can only be seen occasionally, in all MS patients observed the mitral valve leaflets were often partly visible. In fact, normal valve leaflets are thin, low-signal structures, with wide changes of position during the cardiac cycle, which prevents their clear recognition on spin echo images. Decreased movements of leaflets, on the other hand, allow for their detection in the case of MS. Besides the fact that mitral leaflets of MS patients were seen on spin echo images, their aspect was strikingly abnormal, as they were thick and sometimes also possessedan abnormally high signal intensity. The LA was enlarged in all the cases and an abnormal doming of the interatrial septum toward the right atrium was also often striking. In all the patients a variable degree of signal from the left atria1 cavity could be detected (Fig. 1). This abnormal intraluminal signal was variably present throughout the cardiac cycle, and when it appeared it was seen mainly in diastole. Signal intensity was sometimes so

strong that it covered the entire volume of the left atrium. In three patients MRI showed evidence for LA thrombosis. Atria1 thrombosis was localized in one patient in the atria1 appendage and in two patients on the roof of the atrium. Signs of right section overload were observed in six subjects (increased right ventricular, right atrial, and pulmonary artery dimensions). Cine MRI findings in mitral stenosis. Cine MRI permitted the detection of several abnormal features in MS. In all the patients a variable order of mitral valve leaflet thickening was constantly seen. This finding was associated with a reduced diastolic opening and an abnormal motion of the stenotic valve toward the left ventricular outflow tract. Signal from blood in the left atrium showed a lower intensity compared with that from the right atrium. Furthermore, the direction of flow within the left atrium was abnormal, as it mostly showed a circular motion within the cavity. A bulging of the interatrial septum was visible for the

Volume

123

Number

5

Cine MRI in mitral stenosis

1257

Fig. 4. Cine MRI seriesobtained in systole in a patient with mitral stenosishaving tricuspid regurgitation. Note the presenceof a dark jet originating from the tricuspid valve and directed toward the posterior aspect of the right atrium. These findings are consistent with severe tricuspid regurgitation.

entire duration of the cardiac cycle, suggesting the presence of abnormally high pressures within the left atrium. Flow acrossthe valve was also strikingly abnormal. In 14 patients, flow across the stenotic leaflets was characterized by a bright diastolic jet from the valve directed toward the apex of the left ventricle (Figs. 2 and 3). This bright jet was constantly surrounded by a small area of loss of signal. In the remaining six patients a dark jet instead of a bright one was seen in diastole. Nine patients also had tricuspid regurgitation. Tricuspid regurgitation was identified by a systolic lossof signal within the right atrium that started from the tricuspid valve (Fig. 4). The eight patients with coexistent mitral regurgitation were also readily identified. The length of the mitral and tricuspid regurgitation jet as detected by tine MRI were the same as those shown by pulsed wave Doppler mapping in the same region.

Echocardiographic and MRI measurements. A significant linear relationship was found between 2-DE and tine MRI for both LA (r = 0.89; p < 0.001) and LV (r = 0.86; p < 0.001) measurements. We also related MMU measured by tine MRI and the mitral valve area as measured by the PHT method at 2-DE (continuous wave Doppler). MM0 was 1.2 f 0.55 cm. The mitral valve area at Doppler echocardiography was 1.5 + 0.66 cm. We found a significant linear relationship between the two measures (r = 0.81; p < 0.001) (Fig. 5). MM0 was easy to calculate because of the frequent presence of the high intensity transmitral flow across the valve, which greatly enhanced the recognition of the distal part of the mitral leaflets. 2-DE showed atria1 thrombosis in only one out of the three patients identified by MRI and later confirmed at surgery. All the other main diagnostic features of MS detected by MRI were easily and rapidly obtained by 2-DE.

1258

Casolo etal.

o! 0

American

1

I

2 CINE - MRI MAXIMAL MITRAL OPENING 1

I

3

Fig. 5. Plot of the measurementof mitral valve area (in square centimeters) as obtained by Doppler echocardiography (Y axis) and of maximal mitral openingasobserved by tine MRI (in centimeters) in the 20 patients studied (X axis) (r = 0.81; p < 0.001).

DISCUSSION

This study demonstrates the improved capacity of MRI to assessMS when using fast imaging techniques. Besides the ability to detect and evaluate the abnormal movement of mitral leaflets, tine MRI showed an ability to quantitate MS. It was also possible to confirm the capability of MRI to depict associated conditions, such asLA thrombosis and signs of right section overload. Furthermore, tine MRI was able to detect the presence of mitral and tricuspid regurgitation in MS patients. In this investigation we could detect direct and indirect signs of MS. On conventional spin echo images, several indirect signs of MS could be observed. Mitral leaflets commonly appeared thickened. However, this finding was not constant. This fact was probably the result of a variable amount of both calcium deposits and fibrosis on the leaflets. While calcium does not determine any magnetic resonance signal, fibrosis usually provides a low- to medium-intensity signal. As we found that leaflets that could not be identified in their full extension by MRI showed a large amount of calcium deposits at 2-DE, we can hypothesize that purely or prevalently fibrotic leaflets are characterized by an increased thickness with a homogenous signal and are clearly visible on spin echo images. On the other hand, when mitral leaflets are not clearly depicted because of an inhomogeneous signal intensity, the presence of large calcium deposits can be suspected. However, because of these fac-

Heart

May 1992 Journal

tors, the mitral valve and its movements cannot be properly studied by conventional spin echo imaging in MS. Another feature of MS present on spin echo images was a high intraatrial signal that could be seen particularly in diastolic images. This intraatrial signal has been also reported by other authorslo, l1 and has been related to stagnant blood in the left atria. Park et a1.l’ have shown that this signal is significantly more evident in patients with a pulmonary artery wedge pressure above 20 mm Hg. In our study the presence of an LA intraluminal signal was related to the severity of MS, as it was more evident in patients with asmaller valve area. However, it should be noted that this feature is not characteristic of MS, as it can be seen in other diseases13and in normal individuals, although not as frequently.14 Other indirect signs of MS shown by MRI on spin echo images are the right ventricular and pulmonary artery dilatations that were only observed in patients with severe MS. We did not measure right ventricular and pulmonary artery pressures and therefore were not able to compare this information. However, a strict relationship between the pulmonary artery intraluminal signal in spin echo images and the value of pulmonary vascular resistance has been described.i5 Direct signs of MS were only collected when using tine MRI. Mitral leaflets were constantly observed, although the signal intensity and thickness from these structures showed wide interindividual differences. Even if calcium does not provide any signal on MRI images, its presence and extent on mitral leaflets can be suspected on tine MRI where calcified leaflets are seen (as contrasted to the high intracardiac chamber signal) as moving structures void of signal. Calcified leaflets also appeared irregularly thickened. The most interesting feature observed by tine MRI in patients with MS was the possibility of detecting a reduced opening of the mitral valve in diastole. All the patients studied showed a reduced amplitude of the mitral leaflet separation throughout the cardiac cycle. This abnormal behavior was also associated with a diastolic movement of both leaflets toward the left ventricular outflow tract. Although we found a close relationship between MM0 and mitral valve area as detected by Doppler echocardiography, it should be stressed that these measures are very different (one being a diameter, the other an area), although they may be somewhat related to each other. We performed this comparison only to evaluate the potential of MRI to assessthe severity of MS. We did not attempt to measure the area rather than the diastolic leaflet separation because of the poor spatial resolution of MRI for this application. Partial

Volume Number

123 5

volume effects, even on contiguous slices, may lead to significant errors. Another potential source of error is the weak temporal resolution of tine MRI, that in this study was set at 50 msec. Smaller slice thicknesses, better lateral resolution, and improved temporal resolution, are therefore required to precisely assess mitral valve area. Because of these limitations, it is unlikely that current MRI systems will provide information on the other important structures that are part of the mitral valve apparatus (cordae, anulus) and whose conditions need to be known when one is considering the advisability of a mitral valve surgical repair. The abnormal valve opening in MS was associated with a peculiar flow pattern. In 14 patients we could detect a transmitral blood flow pattern characterized by a high diastolic increase in signal (relative to the left ventricular intraluminal signal) originating from the distal part of the mitral leaflets and projecting toward the apex of the left ventricle. In six patients we could detect only a dark jet. This aspect was associated with more severe forms of MS and it was also associated with large amounts of calcium on the mitral leaflets. To interpret the presence and aspect of the transmitral jet that we observed in MS patients by tine MRI, it is necessary to know the relationship between signal intensity on tine MRI images and blood flow velocity. 3, 4,8 In a former investigation,16 we observed that by using our system and its parameters, signal intensity parallels velocity up to 20 to 30 cm/set. Above this velocity signal intensity does not increase.16 Furthermore, turbulent flow is always characterized by lack of signal. Therefore analysis of transmitral signal intensity may help to evaluate the kind of flow but cannot provide quantitative data with the current imaging technique. Experimental imaging techniques now under investigation may provide this information.17 MRI was able to show atria1 thrombosis in three subjects with MS; its presence was later confirmed during surgery. Although there are no studies on this matter, it is thought that MRI may be superior to 2-DE for this application. In this study only one out of three patients with atria1 thrombosis was correctly identified by 2-DE. This technique was not able to show the presence of thrombi in the left atria1 appendage, a region that is difficult to explore routinely. Recently, the advent of transesophageal echocardiography has allowed the left atria1 appendage to be studied in detail, and it is thought to be a reliable technique for the detection of atria1 thrombi in MS. la, lg Z-DE represents a standard accepted technique to assess MS.20 Besides providing the diagnosis, this

Cine MRI in mitral stenosis

1259

method can noninvasively measure the mitral valve orifice dimensions, detect the presence and amount of calcium deposition on the valve, and recognize the presence of left atria1 masses. Furthermore, it is a widespread technique, easy to repeat and relatively inexpensive. Therefore 2-DE is routinely used in the diagnosis, assessment,and follow-up of patients with MS, and since its introduction it has almost completely replaced invasive studies before surgery. In this study we have found that tine MRI can detect, evaluate, and quantitate MS. Furthermore, this technique is able to detect atria1 thrombosis and associated valve diseases. However, MRI is an expensive examination that also requires relatively long imaging times for completion. Better spatial and temporal resolution as well as quantitative information about the abnormal diastolic flow in these patients are necessary to completely assessMS and provide noninvasive information comparable with that offered by 2-DE. For these reasons MRI cannot be considered an alternative technique to 2-DE in the routine study of patients with MS. MRI may nevertheless be of clinical relevance for the noninvasive assessment of patients with MS who have poor echocardiographic windows. Also, this technique might be considered in all those patients in whom a left atria1 thrombosis may be suspected but is not demonstrated by 2-DE. REFERENCES

1.

2.

HigginsCB, Byrd BF, FarmerDW, OsakiL, SilvermanN, CheitlinM. Magneticresonance imagingin patientswith congenitalheart disease. Circulation1984;70:851-60. HigginsCB, KaufmannL, CrooksLE. Magnetic resonance imagingof the cardiovascularsystem. AM HEART J 1985;

109:136-52. 3. van Dijk P, van der Meulen P, Pettigrew RI, Bluemm R, Dannels W, Doornbos J. Dynamic studies of cardiac motion and flow with a fast multiphase MRI technique[Abstract]. J Am Co11 Cardiol 1986;7:197A. 4. Sechtem U, Pflugfelder PW, White RD, Gould R, Holt W, Lipton M, Higgins CB. Cine MR imaging: potential for the evaluation of cardiovascular function. AJR 1987;148:239-46. 5. Utz JA, H&kens RJ, Heinsimer JA, Bashore T, Califf R, Glover P, Pelt N, Shimakawa A. Cine MR determination of left ventricular ejection raction. AJR 1987:148:839-43. 6. Sechtem U, Pflugfelder PW, Gould RG, Ca&idy MM, Higgins CB. Measurement of right and left ventricular volumes in healthy individuals with tine MR imaging. Radiology 1987; 163:697-702. 7. Lotan CS, Cranney GB, Bouchard A, Bittner V, Pohost GM. The value of tine nuclear magnetic resonance imaging for assessing regional ventricular function. J Am Co11 Cardiol 1989;14:1721-9. 8. Wagner S, Auffermann W, Buser P, Lim TH, Kircher B, Pflugfelder P, Higgins CB. Diagnostic accuracy and estimation of the severity of valvular regurgitation from the signal void on tine magnetic resonance images. AM HEART J 1989;

118~760-7. 9. Aurigemma G, Reichek N, Schiebler M, Axe1 L. Evaluation mitral regurgitation by tine magnetic resonance imaging. J Cardiol 1990;66:621-5.

of Am

Gas010 et al.

American

JR, Conti CR. Mitral steno10. Hill JA, Akins EW, Fitzsimmons sis: imaeine bv nuclear magnetic resonance. Am J Cardiol 1986;57:;52-4.” 11. Park JH, Han MC, Im JG, Oh BH, Lee YW. Mitral stenosis: evaluation with MR imaging after percutaneous balloon valvuloplasty. Radiology 199O;i77:533-6. 12. Park JH, Han MC, Oh BH, Im JG, Kim SH. Magnetic resonance evaluation of mitral stenosis. Cardiovasc Intervent Radiol 1990;13:294-9. 13. Sechtem U, Higgins CB, Sommerhoff BA, Lipton MJ, Huickle EC. Magnetic resonance imaging of restrictive cardiomiopathy. Am J Cardiol 1987;59:480-3. GK, Fisher M, Crooks LE, Higgins CB. Gated 14. Von Schultess MR imaging of the heart. Intracardiac signals in patients and healthy subjects. Radiology 1985;156:125-32. GK, Fisher MR, Higgins CB. Pathologic blood 15. Von Schulthess flow in pulmonary vascular disease as shown by gated magnetic resonance imaging. Ann Intern Med 1985;103:317-23. M, Ciraoio L, Poggesi L, Renzi R. Eval16. Casolo GC, Bucciolini

Magnetic resonance atrial tumors

imaging

May 1992 Heart Journal

uation of blood flow behavior in patients with congestive heart failure by fast multiphase magnetic resonance imaging. Am J Noninvas Cardiol 1989;3:199-206. 17 Kilner PJ, Firmin DN, O’Rees RS, Martinez J, Pennell DJ, Mohiaddin RH, Underwood SR, Longmore DB. Valve and great vessel stenosis: assessment with MR jet velocity mapping. Radiology 1991;178:229-35. 18. Seward JB, Khandheria BK, Oh JK, Abel MD, Hughes WR, Edwards WD, Nichols BA, Freeman WK, Tajik J. Transesophageal echocardiography: technique, anatomic correlations, implementation, and clinical application. Mayo Clin Proc 1988;63:649-58. 19. Aschenberg W, Schluter M, Kremer P, Schroeder E, Siglow V, Bleifeld W. Transesophageal two-dimensional echocardiography for the detection of left atria1 appendage thrombus. J Am Co11 Cardiol 1986;7:163-6. H. Acquired valvular heart disease. In: Feigen20. Feigenbaum baum H, ed. Echocardiography. 4th ed. chapter 6. Philadelphia: Lea & Febiger, 1986:250.

of suspected

Two-dimensional echocardiography has become the standard technique for evaluation of cardiac and paracardiac mass lesions. We have used magnetic resonance imaging (MRI) as an independent assessment of cardiac-associated masses in patients with echocardiograms demonstrating sessile atrial tumors. MRI was performed in seven patients, ages 33 to 84, whose echocardiographic diagnoses included left atrial mass (five), right atrial mass (one), and interatrial mass (one). In four of the patients with a diagnosis of left atrial mass, MRI showed extracardiac compression of the atrium, simulating a tumor (hiatal hernia, tortuous descending aorta, bronchogenic cyst). MRI was entirely normal in one patient with an apparent left atrial mass. MRI elucidated extension of an extracavitary mass into the interatrial septum in two patients. One of these patients with an echocardiographic right atrial mass had extension of a lipoma into the interatrial septum without atrial tumor. MRI confirmed the echocardiographic diagnosis of an interatrial mass in the other patient. We conclude that MRI, because of its ability to define anatomic relationships and tissue characteristics, is a powerful noninvasive tool for evaluating suspected cardiac mass lesions. Although echocardiography remains the primary screening test for the detection of cardiac masses, MRI is a more specific modality for precise diagnosis. Correct MRI interpretation may obviate the need for invasive studies or surgery. (AM HEART J 1992;123:1260.)

Mark A. Menegus, MD, Mark A. Greenberg, MD, Hugo Spindola-France, Ayodeji Fayemi, MD. Bronx, N.Y.

From the Division of Cardiology Medical Center/Albert Einstein Received

for publication

July

and Department of Radiology, College of Medicine. 17, 1991;

Reprint requests: Mark A. Menegus, fiore Medical Center, 111 East 210th

4/l/35747

1260

accepted

Montefiore

Oct. 21, 1991.

MD, Division of Cardiology, St., Bronx, NY 10467.

Monte-

MD, and

Since its introduction in the late 196Os,echocardiography has become the standard noninvasive method for evaluation of cardiac and paracardiac mass lesions. Atria1 myxoma, the most common primary intracardiac neoplasm, has been studied extensively1