Angular Displacement of the Papillary Muscles During the Cardiac Cycle MICHAEL J. MIRRO, M.D., EDWIN W. ROGERS, M.D., ARTHUR E. WEYMAN, M.D., AND HARVEY FEIGENBAUM, M.D. SUMMARY Left ventricular rotation about the longitudinal axis of the heart has been qualitatively described but difficult to quantitate. Cross-sectional echocardiography offers a simple, noninvasive method for determining both the presence and degree of left ventricular rotation. In this study we examined normals and patients with various forms of heart disease for angular displacement of the papillary muscles during the cardiac cycle. Assuming that the papillary muscles represent fixed internal reference points, angular displacement of these points represents left ventricular rotation about the longitudinal axis. Short-axis, cross-sectional echocardiograms were recorded in 100 patients 34 normals, 16 patients with secundum atrial septal defect and 50 patients with various forms of valvular, congenital and ischemic heart disease. Angular displacement of the papillary muscles was determined by measuring the angle subtended between a line drawn parallel to the chest wall and a second line drawn through the tips of the papillary muscles. The degree of angular displacement of the papillary muscles was then derived by subtracting the diastolic angle from the systolic angle. This value was interpreted to represent the degree of left ventricular rotation about the longitudinal axis during the cardiac cycle. The normal group had a minimal degree of counterclockwise left ventricular rotation (mean 3.00; range 0-6°). Exaggerated counterclockwise left ventricular rotation was noted in patients with secundum atrial septal defect (mean 17.20; range 9-29°; p < 0.001). Eight of these patients were studied after surgical closure of the defect, and the pattern of left ventricular rotation was not significantly different from that in normals (mean 3.3°; range 2-5°). The third group of patients, who had various forms of heart disease, including right ventricular volume and pressure overload, had minimal left ventricular rotation (mean 3.20; range 0-7°). We conclude that minimal counterclockwise left ventricular rotation occurs in normals and patients with heart disease. Exaggerated left ventricular rotation is characteristically observed in patients with secundum atrial septal defect, and is abolished after surgical closure of the defect.
of the left ventricle in patients with secundum atrial septal defects, we observed exaggerated angular displacement of the papillary muscles during the contraction-relaxation sequence. Assuming that the papillary muscles represent fixed internal reference points, the angular rotation of these internal structures reflects angular rotation of the left ventricle about its longitudinal axis during the cardiac cycle. Both M-mode and cross-sectional echocardiographic features of this congenital lesion have been extensively described; however, exaggerated rotation of the left ventricle about its longitudinal axis has not been included in these descriptions.10-19 Therefore, we used cross-sectional echocardiograms, to examine the left ventricle for rotational motion about its longitudinal axis during the contraction-relaxation sequence, and quantitated the degree of left ventricular rotation in normals, patients with atrial septal defects, and patients with various forms of heart disease.
SIR WILLIAM HARVEY, in 1628, first qualitatively described rotation of the heart during the contraction-relaxation sequence.' He observed, in the open-chested animal, a twisting motion of the ventricles that accompanied the onset of ventricular systole. Since that time, numerous investigators have observed that the heart rotates about its longitudinal axis during the cardiac cycle.2`9 The degree of left ventricular rotation, however, has been difficult to measure, and only recently, with the use of epicardial markers, has this phenomenon been quantitated.7 These studies have revealed a minimal degree (3-7°) of angular, counterclockwise rotation of the left ventricle about its longitudinal axis during the cardiac cycle. Unfortunately, invasive methods of studying left ventricular motion and geometry (cineangiography and epicardial markers) have limited the assessment of left ventricular rotation to a small group of patients. Cross-sectional echocardiography provides a noninvasive method of recording and studying the dynamic spatial geometry of the heart, as well as left ventricular motion. While studying the geometric shape
Methods Patient Population
Short-axis, cross-sectional echocardiographic studies of the left ventricle were performed in 100 patients by means of an 800 focused, phased-array imaging system and a 300 or 820 real-time mechanical sector scanner. We studied three groups of patients. The first group consisted of 34 uncatheterized patients who were normal by physical examination and echocardiography. There were 18 females and 16 males, average age was 28 years (range 16-84 years). The second group consisted of 16 patients with secun-
From the Krannert Institute of Cardiology, the Department of Medicine, Indiana University School of Medicine, and the Veterans Administration Hospital, Indianapolis, Indiana. Supported in part by the Herman C. Krannert Fund, and by grants HL 06308 and HL 07182 from the NHLBI, NIH, USPHS, and the American Heart Association, Indiana Affiliate, Inc. Address for reprints: Michael J. Mirro, M.D., University of Iowa Hospitals, Iowa City, Iowa 52242. Received October 30, 1978; revision accepted February 6, 1979.
Circulation 60, No. 2, 1979.
327
328
CIRCULATION
VOL 60, No 2, AUGUST 1979
TABLE 1. Summary of Patient Data
Age PA pressures LVR Patient (years) Sex Diagnosis (mm Hg) Qp/Qs PVR (preop) (postop) 1 36 F S-ASD 46/13, 26 3.8:1 0.67 290 30 2 50 F S-ASD 38/18, 26 4.1:1 2.7 160 30 3 F 6 S-ASD 23/4, 13 2:1 1.33 190 20 4 44 M S-ASD + CAD 33/13, 17 3.4:1 0.6 90 20 5 22 F S-ASD 5:1 0.59 210 50 27/5, 15 4 6 F S-ASD 22/4, 10 2.8:1 1.0 200 40 7 20 F S-ASD 18/7, 13 3.2:1 1.19 210 40 8 F 8 S-ASD 28/8, 18 2.1:1 2.0 10° 30 9 9 M S-ASD 20/5, 12 2:1 1.0 110 4 10 F S-ASD 26/7, 13 2.78:1 0.88 180 11 F 58 S-ASD 27/12, 17 2.6:1 2.0 110 12 F 5 S-ASD 1.94:1 18/6, 10 0.77 200 13 50 F S-ASD 36/14, 24 3.5:1 1.4 140 14 22 M SV-ASD 28/10, 18 1.7:1 1.05 290 15 F 5 S-ASD 22/5, 12 3.8:1 1.0 100 M 16 23 S-ASD 19/9, 10 2.6:1 0.63 170 Abbreviations: S-ASD = secundum atrial septal defect; SV-ASD = sinus venosus atrial septal defect; CAD = coronary artery disease; PA = pulmonaary artery; Qp/Qs = pulmonary-to-systemic flow ratio; PVR pulmonary vascular resistance; LVR = lefft ventricular rotation.
dum atrial septal defect documented by cardiac catheterization. Eight of these 16 patients were studied both before and after surgical closure of the defect. A summary of the clinical and preoperative hemodynamic data of this group is presented in table 1. The third group consisted of 50 patients with various forms of valvular, congenital and ischemic heart disease documented by cardiac catheterization and echocardiography. There were 28 males and 22 females, average age 32 years (range 1-72 years). Of these 50 patients, 25 had valvular heart disease, nine had mitral valve prolapse with mitral regurgitation, six had aortic regurgitation, four had tricuspid regurgitation (with pulmonary hypertension), four had aortic stenosis and two had mitral stenosis. There were 15 patients with congenital heart disease: five had ventricular septal defect, three had atrial septal defect associated with moderate-to-severe pulmonic stenosis, two had atrial septal defect and ventricular septal defect, four had tetralogy of Fallot and one had single ventricle. In the subgroup of patients with ischemic heart disease, two patients had inferior wall myocardial infarction, four had anterior wall myocardial infarction and four had significant (> 80%) coronary lesions by angiography without evidence of myocardial infarction. Echocardiographic Techniques
Cross-sectional echocardiographic studies were performed with patients in a supine or 30° left lateral decubitus position. The sweep of the cross-sectional probe was aligned with the plane of the scan perpendicular to the long axis or parallel to the short axis of
the left ventricle. Angular displacement of the papillary muscles was measured by first obtaining a short-axis study of the left ventricular cavity at the level of the papillary muscles. Individual frames were then obtained in systole and diastole and converted to a hard copy display with a standard Polaroid photographic system. Next, a line was drawn through the tips of the papillary muscles and its angle measured with respect to a second line drawn parallel to the chest wall (fig. 1). Finally, the degree of angular displacement of the papillary muscles was obtained by subtracting the angle measured in the diastolic frame from the angle measured in the following systolic frame. The degree of angular displacement of the papillary muscles was assumed to represent the degree of angular rotation of the left ventricle about its longitudinal axis; therefore, in the present communication, angular rotation of the papillary muscles has been equated with left ventricular rotation, and hereafter is referred to as left ventricular rotation. The echocardiographic studies were performed with an 800 phased-array sector scanner (Varian 3000) or a 300 and/or 82° mechanical sector scanner (Smith Kline). Statistical Analysis
Statistical significance between the means of each group was determined by the unpaired t test. Analysis of variance was also used to confirm statistical differences between the means. Significant differences between the means in patients with secundum atrial septal defects were determined by the paired t test. All results are expressed as the mean ± SEM.
LEFT VENTRICULAR ROTATION/Mirro et al.
329
FIGURE 1. Short-axis, cross-sectional recording of the left ventricle at the level of the papillary muscles in a normal subject. A) A diastolic frame. A line drawn through the tips of the papillary muscles intersects at an angle of 25 ° with a plane drawn parallel to the chest wall. B) A systolic frame. A line drawn through the tips of the papillary muscles intersects at an angle of 300 with the plane parallel to the chest wall. Thus, the left ventricle has rotated 5° from diastole to systole about its longitudinal axis in a counterclockwise direction.
Results Left Ventricular Rotation in Normals
We observed minimal counterclockwise left ventricular rotation about the longitudinal axis in a consecutive series of 34 normals (mean 2.80; range 0-60). Figure 1 shows a typical short-axis study recorded in a normal subject at the level of the papillary muscles. Panel A shows a late diastolic frame in which the plane of the papillary muscles was 250 with respect to a plane parallel to the chest wall. In panel B, with the onset of left ventricular ejection, the plane of the papillary muscles rotated counterclockwise to 300 with respect to the plane parallel to the chest wall. Thus, the left ventricle rotated 50 about its longitudinal axis during the contraction-relaxation sequence. Left ventricular rotation always proceeded in a counterclockwise direction and was an early systolic event. These data confirm previous reports of counterclockwise rotation of the heart during the cardiac cycle.9
Exaggerated Left Ventricular Rotation in Secundum Atrial Septal Defect Patients with secundum atrial septal defects studied
preoperatively had exaggerated counterclockwise left ventricular rotation compared with normal subjects. In the 16 patients studied, the mean left ventricular rotation about the longitudinal axis was 17.20 (range 9-29°). Figure 2 shows a typical study from a patient with a secundum atrial septal defect (short-axis left ventricle at the level of the papillary muscles). In panel A, the plane of the papillary muscles in late diastole was 10° with respect to a plane parallel to the chest wall. With the onset of left ventricular ejection (panel B), the plane of the papillary muscles has rotated in a counterclockwise direction to 280. Thus, the left ventricle rotated 180 about its longitudinal axis during the cardiac cycle. As in the normal subjects, counterclockwise rotation of the left ventricle during early systole coincided with left ventricular ejection. Eight of the patients with secundum atrial septal defects were studied 1-20 weeks after surgical closure of the defect. Exaggerated counterclockwise left ventricular rotation was no longer present and the pattern of left ventricular rotation was similar to that in normals. The mean rotation was 3.3° (range 2-5°). Figure 3 shows a typical cross-sectional study from a patient after surgical closure of the defect. In panel A, the plane of the papillary muscles in late diastole was 240 with respect to the plane parallel to the chest wall.
CI RCULATION
330
VOL 60, No 2, AUGUST 1979
FIGURE 2. Short-axis, cross-sectional recording of the left ventricle at the level of the papillary muscles in a patient with a secundum atrial septal defect. A) A diastolic frame. The plane parallel to the papillary muscles intersects at an angle of 10° with the plane parallel to the chest wall. B) A systolic frame, in which this angle has increased to 280. Thus, the left ventricle rotated 18° from diastole to systole about its longitudinal axis.
In early systole (panel B), the plane of the papillary muscles rotated counterclockwise to 29°. Therefore, left ventricular rotation about the longitudinal axis was 50, and similar to the pattern in normal subjects. Thus, secundum atrial septal defect (preoperatively) was associated with a significantly greater degree of counterclockwise left ventricular rotation than in normals or patients with heart disease (fig. 4). Furthermore, after surgical closure of the defect, the pattern of rotation was not significantly different from that in normal subjects. We were unable to find any correlation between the degree of left ventricular rotation in patients with secundum atrial septal defect and their preoperative catheterization data (table 1). However, this negative correlation should be interpreted with caution, since all the patients in this group had relatively high-flow, low pulmonary vascular resistance atrial septal defects. If a comparable number of low-flow, high pulmonary vascular resistance atrial septal defects were included, the correlation between hemodynamics and left ventricular rotation might improve. Left Ventricular Rotation in Patients
with Other Forms of Heart Disease A group of patients with valvular, congenital and ischemic heart disease were also studied, and the
pattern of left ventricular rotation was compared with that in normal subjects and in patients with secundum atrial septal defects. The patients with a variety of heart diseases had a mean of 3.2° (range 0-7°) of counterclockwise left ventricular rotation about the longitudinal axis (fig. 4). There was no significant difference in the degree of left ventricular rotation between this group and the normal subjects. Preoperatively, however, the patients with secundum atrial septal defects had a significantly greater degree of left ventricular rotation than this group (p < 0.001). Furthermore, the degree of left ventricular rotation was analyzed for each subgroup of patients with heart disease, and no significant difference from normals was detected. The subgroup of patients with valvular disease included four patients with tricuspid insufficiency secondary to pulmonary hypertension, and the subgroup of patients with congenital heart disease included three patients with atrial septal defect associated with moderate-to-severe pulmonic stenosis and two patients with atrial septal defect combined with ventricular septal defect. These patients with right ventricular volume overload combined with right ventricular pressure overload or left ventricular volume overload did not have exaggerated left ventricular rotation. Furthermore, isolated left ventricular volume overload (nine patients with mitral regurgitation and six patients with aortic regurgita-
LEFT VENTRICULAR ROTATION/Mirro et a3.
331
FIGURE 3. Short-axis, cross-sectional recording of the left ventricle at the level of the papillary muscles in a patient after surgical closure of an atrial septal defect. A) A diastolic frame. The plane parallel to the papillary muscles intersects at an angle of 24° with the plane parallel to the chest wall. B) A systolic frame, in which this angle has increased to 290. Therefore, the left ventricle has rotated 5' about its longitudinal axis from diastole to systole and resembles the pattern seen in normals.
z 0-
200
12
0 16
FIGURE 4. Histogram comparing the degree of angular displacement of the papillary muscles during the cardiac cycle (left ventricular rotation) in normal subjects, patients with secundum atrial septal defect (preoperative = open bar; postoperative = stippled bar) and patients with various forms ofheart disease. Values are mean ± SEM (in degrees). Asterisk indicates p < 0.001 when compared with normals and the group with heart disease; plus sign indicates p < 0.01 when compared with preoperative study (paired t test).
ck
a: D
52 120
z uJ
V
8
Lu
0 C] L 4 LU.0 ua
0
FNORMALS
16 80 SECUNDUM
HEART
ATRIAL SEPTAL DEFECT
DISEASE
332
CIRCULATION
tion) also did not have exaggerated left ventricular rotation.
Discussion The present study demonstrates that left ventricular rotation occurs in normal subjects and can be quantitated by cross-sectional echocardiography. Furthermore, this rotation of the left ventricle about its longitudinal axis is counterclockwise (viewed from the apex) and coincides with the onset of left ventricular ejection. In addition, exaggerated counterclockwise left ventricular rotation is a uniform feature of secundum atrial septal defect (without right-sided pressure overload) that can be abolished after surgical closure of the defect. Finally, we have not observed exaggerated left ventricular rotation in combined right ventricular pressure and volume overload or in other disease states. Thus, this study confirms earlier observations that left ventricular rotation occurs during the cardiac cycle and agrees quantitatively with the values reported on normal subjects.7-" This study extends previous observations concerning the cross-sectional echocardiographic features of secundum atrial septal defects, and we hope this finding will improve diagnostic accuracy by noninvasive methods.17-19 In this report, angular displacement of the papillary muscles during the cardiac cycle represents the actual measurement obtained, and we assume that this represents rotation of the left ventricle about its longitudinal axis. Although this assumption appears to be sound, there are a few theoretical mechanisms by which the measurement of angular displacement of the papillary muscles might not truly reflect axial rotation of the left ventricle. First, if the papillary muscles were anatomically distorted in a particular disease state (i.e., atrial septal defect) and represented large, elongated structures that coursed in an exaggerated spiral fashion within the left ventricular cavity, and if these unusual anatomic relationships were viewed from the apex in a short-axis projection of the left ventricle, left ventricular shortening (from apex to base) might appear to be angular displacement of the papillary muscles. Second, if one of the papillary muscles exhibited exaggerated motion due to a localized area of hyperdynamic left ventricular wall motion, as might occur in disease states where dramatic changes in mitral valvular (and chordal structures) and left ventricular geometry occur, the longitudinal axis (long axis) of one papillary muscle might drastically change in relation to the long axis of the other papillary muscle (and left ventricle). Finally, if in a particular disease state, the various muscular sheaths, which are layered within the left ventricular cavity, could undergo angular rotation independent of each other, angular displacement of the papillary muscles would not represent true left ventricular rotation. This appears to be unlikely, however, since each muscular sheath is intimately intertwined with the various other muscular layers.3 Although there are a variety of hypothetical explanations for angular displacement of the papillary muscles, the most plausible
VOL 60, No 2, AUGUST 1979
explanation is that this finding is a reflection of left ventricular rotation about its longitudinal axis. The mechanism for left ventricular rotation in normals and exaggerated rotation in secundum atrial septal defect is open to speculation. However, from the observations reported by other investigators and the data acquired from this study, a reasonable hypothesis can be postulated.2 Motion of specific points of the ventricular myocardium must be interpreted with knowledge of the architecture and functional orientation of the muscular sheaths that make up the ventricular walls. Functionally, the ventricles are formed by two sets of muscles: 1) the spiral muscles, which form the internal and external layers of myocardium, and 2) the deep ventricular constrictor muscles, which lie between the inner and outer layers of spiral muscle. Rushmer et al. stated that simultaneous shortening of the internal and external layers of the spiral muscle would not normally produce ventricular rotation, because the oblique tensions exerted by each layer would be counterbalanced, and the result of their combined work is to shorten the long axis of the ventricular chambers.3 However, the deep constrictor muscles, being much more powerful on the left than the right, function to reduce ventricular circumference. Normally, the combined effort of these two groups of muscle fibers ejects an equal volume of blood from both ventricles. In order for the left ventricular chamber to rotate counterclockwise in early systole, there must be asymmetric shortening of the spiral or constrictor bundles. Since the spiral muscles are arranged clockwise epicardially and counterclockwise endocardially, it is possible that the endocardial spiral layer shortens to a greater degree and thus produces a minimal degree of left ventricular counterclockwise rotation in normal subjects. If this hypothesis is correct, isolated right ventricular volume overload exaggerates this disparity in fiber shortening between the external and internal spiral muscles. However, if this indeed represents the mechanism for exaggerated left ventricular rotation in isolated right ventricular volume overload, why have we not observed exaggerated clockwise left ventricular rotation in isolated left ventricular volume overload? In chronic left ventricular volume overload, it is possible that the deep constrictor muscles completely predominate over the external and internal spiral muscles so that any disparity in fiber shortening between these layers is not reflected in left ventricular motion. Analysis of acute vs chronic left ventricular volume overload might possibly further define the pathophysiology of this phenomenon. Further investigative efforts will be necessary to fully elucidate the mechanism of left ventricular rotation; however, this phenomenon appears to extend our knowledge of myocardial contraction. In addition to providing useful information concerning myocardial muscle function, the observation that exaggerated left ventricular rotation occurs in secundum atrial septal defects may explain the mechanism by which certain physical findings occur in this disorder. The clinical features of secundum atrial
LEFT VENTRICULAR ROTATION/Mirro et al.
septal defect have been extensively described,20 22 and the diagnostic value. of precordial movement in this congenital lesion has been emphasized. The right ventricular impulse, palpated at the left sternal border, has been described as hyperdynamic and of brief duration in early systole. Furthermore, if pulmonary hypertension develops or right ventricular pressure overload is superimposed, the right ventricular impulse changes dramatically to a less dynamic, more sustained systolic phenomenon.20 23 It appears reasonable that, at least in part, the counterclockwise rotation of the left ventricle may aid in thrusting the right ventricle against the sternum in early systole and exaggerate the intensity of the right ventricular impulse. This phenomenon might occur because, with chronic right ventricular volume overload, the right ventricle dilates and develops hypertrophy, shifting the area normally occupied by the right ventricle laterally toward the area of the apex.23 Thus, with the onset of left ventricular ejection, the exaggerated counterclockwise rotation of the left ventricle (viewed from the apex) would thrust the bulging right ventricle anteriorly and toward the left sternal border. Because left ventricular rotation is an early and brief systolic event, this phenomenon coincides with the fact that the hyperdynamic right ventricular impulse in right ventricular volume overload is also a brief, early systolic event. Further investigative efforts (combining two-dimensional echocardiography with phonocardiography) should disclose the precise cardiac movements responsible for these findings. In summary, left ventricular rotation occurs in normal subjects and in patients with various heart diseases. Exaggerated left ventricular rotation is a uniform feature of secundum atrial septal defect. Appreciation of this phenomenon is important for interpreting data that rely on a constant relationship between epicardial surface and chest wall (i.e., STsegment mapping). We hope that recognition of exaggerated patterns of left ventricular rotation will improve our diagnostic accuracy in echocardiographically identifying patients with secundum atrial septal defects. References 1. Harvey W: Exercitatis Anatomica de Motu Cordis et Sanguinis in Animalibus. Frankfurt, 1628, ch 5
333
2. Hamilton WF, Romph JH: Movements of the base of the ventricle and the relative constancy of the cardiac volume. J Physiol 102: 559, 1932 3. Rushmer RF, Crystal DK, Wagner C: The functional anatomy of ventricular contraction. Circ Res 1: 162, 1963 4. Harrison DC, Goldblatt A, Braunwald E: Studies on cardiac dimensions in intact, unanesthetized man; description of techniques and their validation. Circ Res 13: 448, 1973 5. Rushmer RF: Physical characteristics of myocardial performance. Am J Cardiol 18: 61, 1966 6. Lynch RP, Bove AA: Geometry of the left ventricle as studied by a high speed cineradiographic technique. Fed Proc 28: 1330, 1969 7. Sandler H, Dodge HT: The use of single plane angiocardiograms for the calculation of left ventricular volume in man. Am Heart J 75: 325, 1968 8. Klein MD, Herman MV, Gorlin R: A hemodynamic study of left ventricular aneurysm. Circulation 35: 614, 1967 9. McDonald IG: The shape and movements of the human left ventricle during systole. Am J Cardiol 26: 221, 1970 10. Diamond MA, Dillon JC, Haine JC, Chang S, Feigenbaum H: Echocardiographic features of atrial septal defect. Circulation 43: 129, 1971 11. Nimura Y, Matsuo H, Matsumoto M, Kitabatake A, Abe H: Interatrial septum in ultrasonocardiotomogram and ultrasound cardiogram. Medical Ultrasound 9: 58, 1971 12. Tajik AJ, Gau GT, Ritter DG, Schattenberg TT: Echocardiographic pattern of right ventricular diastolic volume overload in children. Circulation 46: 36, 1972 13. McCann WD, Harbold NB, Giuliani BR: The echocardiogram in right ventricular overload. JAMA 221: 1243, 1972 14. Matsumoto M: Ultrasonic features of interatrial septum: its motion analysis and detection of its defect. Jpn Circ J 37: 1383, 1973 15. Radtke WE, Tajik AJ, Gau GT, Schattenberg TT, Giluiani ER, Tancredi RG: Atrial septal defect: echocardiographic observations. Ann Intern Med 84: 246, 1976 16. Feigenbaum H: Echocardiography, 2nd ed. Philadelphia, Lea and Febiger, 1976, ch 14 17. Weyman AE, Wann LS, Feigenbaum H, Dillon JC: Mechanism of abnormal septal motion in patients with right ventricular volume overload. Circulation 54: 179, 1976 18. Lieppe W, Scallion R, Behar VS, Kisslo JA: Two-dimensional echocardiographic findings in atrial septal defect. Circulation 56: 447, 1977 19. Weyman AE, Wann LS, Caldwell RL, Hurwitz RA, Dillon JC, Feigenbaum H: Negative contrast echocardiography: a new method for detecting left-to-right shunts. Circulation 59: 498, 1979 20. Perloff JK: The Clinical Recognition of Congenital Heart Disease. Philadelphia, WB Saunders, 1970, ch 15 21. Roesler H: Interatrial septal defect. Arch Intern Med 54: 339, 1934 22. Bedford DE, Papp C, Parkinson J: Atrial septal defect. Br Heart J 3: 37, 1941 23. Tavel ME: Clinical phonocardiography and external pulse recording, 2nd ed. Chicago, Year Book Medical, 1973, ch 8
of the left ventricle in patients with secundum atrial septal defects, we observed exaggerated angular displacement of the papillary muscles during the contraction-relaxation sequence. Assuming that the papillary muscles represent fixed internal reference points, the angular rotation of these internal structures reflects angular rotation of the left ventricle about its longitudinal axis during the cardiac cycle. Both M-mode and cross-sectional echocardiographic features of this congenital lesion have been extensively described; however, exaggerated rotation of the left ventricle about its longitudinal axis has not been included in these descriptions.10-19 Therefore, we used cross-sectional echocardiograms, to examine the left ventricle for rotational motion about its longitudinal axis during the contraction-relaxation sequence, and quantitated the degree of left ventricular rotation in normals, patients with atrial septal defects, and patients with various forms of heart disease.
SIR WILLIAM HARVEY, in 1628, first qualitatively described rotation of the heart during the contraction-relaxation sequence.' He observed, in the open-chested animal, a twisting motion of the ventricles that accompanied the onset of ventricular systole. Since that time, numerous investigators have observed that the heart rotates about its longitudinal axis during the cardiac cycle.2`9 The degree of left ventricular rotation, however, has been difficult to measure, and only recently, with the use of epicardial markers, has this phenomenon been quantitated.7 These studies have revealed a minimal degree (3-7°) of angular, counterclockwise rotation of the left ventricle about its longitudinal axis during the cardiac cycle. Unfortunately, invasive methods of studying left ventricular motion and geometry (cineangiography and epicardial markers) have limited the assessment of left ventricular rotation to a small group of patients. Cross-sectional echocardiography provides a noninvasive method of recording and studying the dynamic spatial geometry of the heart, as well as left ventricular motion. While studying the geometric shape
Methods Patient Population
Short-axis, cross-sectional echocardiographic studies of the left ventricle were performed in 100 patients by means of an 800 focused, phased-array imaging system and a 300 or 820 real-time mechanical sector scanner. We studied three groups of patients. The first group consisted of 34 uncatheterized patients who were normal by physical examination and echocardiography. There were 18 females and 16 males, average age was 28 years (range 16-84 years). The second group consisted of 16 patients with secun-
From the Krannert Institute of Cardiology, the Department of Medicine, Indiana University School of Medicine, and the Veterans Administration Hospital, Indianapolis, Indiana. Supported in part by the Herman C. Krannert Fund, and by grants HL 06308 and HL 07182 from the NHLBI, NIH, USPHS, and the American Heart Association, Indiana Affiliate, Inc. Address for reprints: Michael J. Mirro, M.D., University of Iowa Hospitals, Iowa City, Iowa 52242. Received October 30, 1978; revision accepted February 6, 1979.
Circulation 60, No. 2, 1979.
327
328
CIRCULATION
VOL 60, No 2, AUGUST 1979
TABLE 1. Summary of Patient Data
Age PA pressures LVR Patient (years) Sex Diagnosis (mm Hg) Qp/Qs PVR (preop) (postop) 1 36 F S-ASD 46/13, 26 3.8:1 0.67 290 30 2 50 F S-ASD 38/18, 26 4.1:1 2.7 160 30 3 F 6 S-ASD 23/4, 13 2:1 1.33 190 20 4 44 M S-ASD + CAD 33/13, 17 3.4:1 0.6 90 20 5 22 F S-ASD 5:1 0.59 210 50 27/5, 15 4 6 F S-ASD 22/4, 10 2.8:1 1.0 200 40 7 20 F S-ASD 18/7, 13 3.2:1 1.19 210 40 8 F 8 S-ASD 28/8, 18 2.1:1 2.0 10° 30 9 9 M S-ASD 20/5, 12 2:1 1.0 110 4 10 F S-ASD 26/7, 13 2.78:1 0.88 180 11 F 58 S-ASD 27/12, 17 2.6:1 2.0 110 12 F 5 S-ASD 1.94:1 18/6, 10 0.77 200 13 50 F S-ASD 36/14, 24 3.5:1 1.4 140 14 22 M SV-ASD 28/10, 18 1.7:1 1.05 290 15 F 5 S-ASD 22/5, 12 3.8:1 1.0 100 M 16 23 S-ASD 19/9, 10 2.6:1 0.63 170 Abbreviations: S-ASD = secundum atrial septal defect; SV-ASD = sinus venosus atrial septal defect; CAD = coronary artery disease; PA = pulmonaary artery; Qp/Qs = pulmonary-to-systemic flow ratio; PVR pulmonary vascular resistance; LVR = lefft ventricular rotation.
dum atrial septal defect documented by cardiac catheterization. Eight of these 16 patients were studied both before and after surgical closure of the defect. A summary of the clinical and preoperative hemodynamic data of this group is presented in table 1. The third group consisted of 50 patients with various forms of valvular, congenital and ischemic heart disease documented by cardiac catheterization and echocardiography. There were 28 males and 22 females, average age 32 years (range 1-72 years). Of these 50 patients, 25 had valvular heart disease, nine had mitral valve prolapse with mitral regurgitation, six had aortic regurgitation, four had tricuspid regurgitation (with pulmonary hypertension), four had aortic stenosis and two had mitral stenosis. There were 15 patients with congenital heart disease: five had ventricular septal defect, three had atrial septal defect associated with moderate-to-severe pulmonic stenosis, two had atrial septal defect and ventricular septal defect, four had tetralogy of Fallot and one had single ventricle. In the subgroup of patients with ischemic heart disease, two patients had inferior wall myocardial infarction, four had anterior wall myocardial infarction and four had significant (> 80%) coronary lesions by angiography without evidence of myocardial infarction. Echocardiographic Techniques
Cross-sectional echocardiographic studies were performed with patients in a supine or 30° left lateral decubitus position. The sweep of the cross-sectional probe was aligned with the plane of the scan perpendicular to the long axis or parallel to the short axis of
the left ventricle. Angular displacement of the papillary muscles was measured by first obtaining a short-axis study of the left ventricular cavity at the level of the papillary muscles. Individual frames were then obtained in systole and diastole and converted to a hard copy display with a standard Polaroid photographic system. Next, a line was drawn through the tips of the papillary muscles and its angle measured with respect to a second line drawn parallel to the chest wall (fig. 1). Finally, the degree of angular displacement of the papillary muscles was obtained by subtracting the angle measured in the diastolic frame from the angle measured in the following systolic frame. The degree of angular displacement of the papillary muscles was assumed to represent the degree of angular rotation of the left ventricle about its longitudinal axis; therefore, in the present communication, angular rotation of the papillary muscles has been equated with left ventricular rotation, and hereafter is referred to as left ventricular rotation. The echocardiographic studies were performed with an 800 phased-array sector scanner (Varian 3000) or a 300 and/or 82° mechanical sector scanner (Smith Kline). Statistical Analysis
Statistical significance between the means of each group was determined by the unpaired t test. Analysis of variance was also used to confirm statistical differences between the means. Significant differences between the means in patients with secundum atrial septal defects were determined by the paired t test. All results are expressed as the mean ± SEM.
LEFT VENTRICULAR ROTATION/Mirro et al.
329
FIGURE 1. Short-axis, cross-sectional recording of the left ventricle at the level of the papillary muscles in a normal subject. A) A diastolic frame. A line drawn through the tips of the papillary muscles intersects at an angle of 25 ° with a plane drawn parallel to the chest wall. B) A systolic frame. A line drawn through the tips of the papillary muscles intersects at an angle of 300 with the plane parallel to the chest wall. Thus, the left ventricle has rotated 5° from diastole to systole about its longitudinal axis in a counterclockwise direction.
Results Left Ventricular Rotation in Normals
We observed minimal counterclockwise left ventricular rotation about the longitudinal axis in a consecutive series of 34 normals (mean 2.80; range 0-60). Figure 1 shows a typical short-axis study recorded in a normal subject at the level of the papillary muscles. Panel A shows a late diastolic frame in which the plane of the papillary muscles was 250 with respect to a plane parallel to the chest wall. In panel B, with the onset of left ventricular ejection, the plane of the papillary muscles rotated counterclockwise to 300 with respect to the plane parallel to the chest wall. Thus, the left ventricle rotated 50 about its longitudinal axis during the contraction-relaxation sequence. Left ventricular rotation always proceeded in a counterclockwise direction and was an early systolic event. These data confirm previous reports of counterclockwise rotation of the heart during the cardiac cycle.9
Exaggerated Left Ventricular Rotation in Secundum Atrial Septal Defect Patients with secundum atrial septal defects studied
preoperatively had exaggerated counterclockwise left ventricular rotation compared with normal subjects. In the 16 patients studied, the mean left ventricular rotation about the longitudinal axis was 17.20 (range 9-29°). Figure 2 shows a typical study from a patient with a secundum atrial septal defect (short-axis left ventricle at the level of the papillary muscles). In panel A, the plane of the papillary muscles in late diastole was 10° with respect to a plane parallel to the chest wall. With the onset of left ventricular ejection (panel B), the plane of the papillary muscles has rotated in a counterclockwise direction to 280. Thus, the left ventricle rotated 180 about its longitudinal axis during the cardiac cycle. As in the normal subjects, counterclockwise rotation of the left ventricle during early systole coincided with left ventricular ejection. Eight of the patients with secundum atrial septal defects were studied 1-20 weeks after surgical closure of the defect. Exaggerated counterclockwise left ventricular rotation was no longer present and the pattern of left ventricular rotation was similar to that in normals. The mean rotation was 3.3° (range 2-5°). Figure 3 shows a typical cross-sectional study from a patient after surgical closure of the defect. In panel A, the plane of the papillary muscles in late diastole was 240 with respect to the plane parallel to the chest wall.
CI RCULATION
330
VOL 60, No 2, AUGUST 1979
FIGURE 2. Short-axis, cross-sectional recording of the left ventricle at the level of the papillary muscles in a patient with a secundum atrial septal defect. A) A diastolic frame. The plane parallel to the papillary muscles intersects at an angle of 10° with the plane parallel to the chest wall. B) A systolic frame, in which this angle has increased to 280. Thus, the left ventricle rotated 18° from diastole to systole about its longitudinal axis.
In early systole (panel B), the plane of the papillary muscles rotated counterclockwise to 29°. Therefore, left ventricular rotation about the longitudinal axis was 50, and similar to the pattern in normal subjects. Thus, secundum atrial septal defect (preoperatively) was associated with a significantly greater degree of counterclockwise left ventricular rotation than in normals or patients with heart disease (fig. 4). Furthermore, after surgical closure of the defect, the pattern of rotation was not significantly different from that in normal subjects. We were unable to find any correlation between the degree of left ventricular rotation in patients with secundum atrial septal defect and their preoperative catheterization data (table 1). However, this negative correlation should be interpreted with caution, since all the patients in this group had relatively high-flow, low pulmonary vascular resistance atrial septal defects. If a comparable number of low-flow, high pulmonary vascular resistance atrial septal defects were included, the correlation between hemodynamics and left ventricular rotation might improve. Left Ventricular Rotation in Patients
with Other Forms of Heart Disease A group of patients with valvular, congenital and ischemic heart disease were also studied, and the
pattern of left ventricular rotation was compared with that in normal subjects and in patients with secundum atrial septal defects. The patients with a variety of heart diseases had a mean of 3.2° (range 0-7°) of counterclockwise left ventricular rotation about the longitudinal axis (fig. 4). There was no significant difference in the degree of left ventricular rotation between this group and the normal subjects. Preoperatively, however, the patients with secundum atrial septal defects had a significantly greater degree of left ventricular rotation than this group (p < 0.001). Furthermore, the degree of left ventricular rotation was analyzed for each subgroup of patients with heart disease, and no significant difference from normals was detected. The subgroup of patients with valvular disease included four patients with tricuspid insufficiency secondary to pulmonary hypertension, and the subgroup of patients with congenital heart disease included three patients with atrial septal defect associated with moderate-to-severe pulmonic stenosis and two patients with atrial septal defect combined with ventricular septal defect. These patients with right ventricular volume overload combined with right ventricular pressure overload or left ventricular volume overload did not have exaggerated left ventricular rotation. Furthermore, isolated left ventricular volume overload (nine patients with mitral regurgitation and six patients with aortic regurgita-
LEFT VENTRICULAR ROTATION/Mirro et a3.
331
FIGURE 3. Short-axis, cross-sectional recording of the left ventricle at the level of the papillary muscles in a patient after surgical closure of an atrial septal defect. A) A diastolic frame. The plane parallel to the papillary muscles intersects at an angle of 24° with the plane parallel to the chest wall. B) A systolic frame, in which this angle has increased to 290. Therefore, the left ventricle has rotated 5' about its longitudinal axis from diastole to systole and resembles the pattern seen in normals.
z 0-
200
12
0 16
FIGURE 4. Histogram comparing the degree of angular displacement of the papillary muscles during the cardiac cycle (left ventricular rotation) in normal subjects, patients with secundum atrial septal defect (preoperative = open bar; postoperative = stippled bar) and patients with various forms ofheart disease. Values are mean ± SEM (in degrees). Asterisk indicates p < 0.001 when compared with normals and the group with heart disease; plus sign indicates p < 0.01 when compared with preoperative study (paired t test).
ck
a: D
52 120
z uJ
V
8
Lu
0 C] L 4 LU.0 ua
0
FNORMALS
16 80 SECUNDUM
HEART
ATRIAL SEPTAL DEFECT
DISEASE
332
CIRCULATION
tion) also did not have exaggerated left ventricular rotation.
Discussion The present study demonstrates that left ventricular rotation occurs in normal subjects and can be quantitated by cross-sectional echocardiography. Furthermore, this rotation of the left ventricle about its longitudinal axis is counterclockwise (viewed from the apex) and coincides with the onset of left ventricular ejection. In addition, exaggerated counterclockwise left ventricular rotation is a uniform feature of secundum atrial septal defect (without right-sided pressure overload) that can be abolished after surgical closure of the defect. Finally, we have not observed exaggerated left ventricular rotation in combined right ventricular pressure and volume overload or in other disease states. Thus, this study confirms earlier observations that left ventricular rotation occurs during the cardiac cycle and agrees quantitatively with the values reported on normal subjects.7-" This study extends previous observations concerning the cross-sectional echocardiographic features of secundum atrial septal defects, and we hope this finding will improve diagnostic accuracy by noninvasive methods.17-19 In this report, angular displacement of the papillary muscles during the cardiac cycle represents the actual measurement obtained, and we assume that this represents rotation of the left ventricle about its longitudinal axis. Although this assumption appears to be sound, there are a few theoretical mechanisms by which the measurement of angular displacement of the papillary muscles might not truly reflect axial rotation of the left ventricle. First, if the papillary muscles were anatomically distorted in a particular disease state (i.e., atrial septal defect) and represented large, elongated structures that coursed in an exaggerated spiral fashion within the left ventricular cavity, and if these unusual anatomic relationships were viewed from the apex in a short-axis projection of the left ventricle, left ventricular shortening (from apex to base) might appear to be angular displacement of the papillary muscles. Second, if one of the papillary muscles exhibited exaggerated motion due to a localized area of hyperdynamic left ventricular wall motion, as might occur in disease states where dramatic changes in mitral valvular (and chordal structures) and left ventricular geometry occur, the longitudinal axis (long axis) of one papillary muscle might drastically change in relation to the long axis of the other papillary muscle (and left ventricle). Finally, if in a particular disease state, the various muscular sheaths, which are layered within the left ventricular cavity, could undergo angular rotation independent of each other, angular displacement of the papillary muscles would not represent true left ventricular rotation. This appears to be unlikely, however, since each muscular sheath is intimately intertwined with the various other muscular layers.3 Although there are a variety of hypothetical explanations for angular displacement of the papillary muscles, the most plausible
VOL 60, No 2, AUGUST 1979
explanation is that this finding is a reflection of left ventricular rotation about its longitudinal axis. The mechanism for left ventricular rotation in normals and exaggerated rotation in secundum atrial septal defect is open to speculation. However, from the observations reported by other investigators and the data acquired from this study, a reasonable hypothesis can be postulated.2 Motion of specific points of the ventricular myocardium must be interpreted with knowledge of the architecture and functional orientation of the muscular sheaths that make up the ventricular walls. Functionally, the ventricles are formed by two sets of muscles: 1) the spiral muscles, which form the internal and external layers of myocardium, and 2) the deep ventricular constrictor muscles, which lie between the inner and outer layers of spiral muscle. Rushmer et al. stated that simultaneous shortening of the internal and external layers of the spiral muscle would not normally produce ventricular rotation, because the oblique tensions exerted by each layer would be counterbalanced, and the result of their combined work is to shorten the long axis of the ventricular chambers.3 However, the deep constrictor muscles, being much more powerful on the left than the right, function to reduce ventricular circumference. Normally, the combined effort of these two groups of muscle fibers ejects an equal volume of blood from both ventricles. In order for the left ventricular chamber to rotate counterclockwise in early systole, there must be asymmetric shortening of the spiral or constrictor bundles. Since the spiral muscles are arranged clockwise epicardially and counterclockwise endocardially, it is possible that the endocardial spiral layer shortens to a greater degree and thus produces a minimal degree of left ventricular counterclockwise rotation in normal subjects. If this hypothesis is correct, isolated right ventricular volume overload exaggerates this disparity in fiber shortening between the external and internal spiral muscles. However, if this indeed represents the mechanism for exaggerated left ventricular rotation in isolated right ventricular volume overload, why have we not observed exaggerated clockwise left ventricular rotation in isolated left ventricular volume overload? In chronic left ventricular volume overload, it is possible that the deep constrictor muscles completely predominate over the external and internal spiral muscles so that any disparity in fiber shortening between these layers is not reflected in left ventricular motion. Analysis of acute vs chronic left ventricular volume overload might possibly further define the pathophysiology of this phenomenon. Further investigative efforts will be necessary to fully elucidate the mechanism of left ventricular rotation; however, this phenomenon appears to extend our knowledge of myocardial contraction. In addition to providing useful information concerning myocardial muscle function, the observation that exaggerated left ventricular rotation occurs in secundum atrial septal defects may explain the mechanism by which certain physical findings occur in this disorder. The clinical features of secundum atrial
LEFT VENTRICULAR ROTATION/Mirro et al.
septal defect have been extensively described,20 22 and the diagnostic value. of precordial movement in this congenital lesion has been emphasized. The right ventricular impulse, palpated at the left sternal border, has been described as hyperdynamic and of brief duration in early systole. Furthermore, if pulmonary hypertension develops or right ventricular pressure overload is superimposed, the right ventricular impulse changes dramatically to a less dynamic, more sustained systolic phenomenon.20 23 It appears reasonable that, at least in part, the counterclockwise rotation of the left ventricle may aid in thrusting the right ventricle against the sternum in early systole and exaggerate the intensity of the right ventricular impulse. This phenomenon might occur because, with chronic right ventricular volume overload, the right ventricle dilates and develops hypertrophy, shifting the area normally occupied by the right ventricle laterally toward the area of the apex.23 Thus, with the onset of left ventricular ejection, the exaggerated counterclockwise rotation of the left ventricle (viewed from the apex) would thrust the bulging right ventricle anteriorly and toward the left sternal border. Because left ventricular rotation is an early and brief systolic event, this phenomenon coincides with the fact that the hyperdynamic right ventricular impulse in right ventricular volume overload is also a brief, early systolic event. Further investigative efforts (combining two-dimensional echocardiography with phonocardiography) should disclose the precise cardiac movements responsible for these findings. In summary, left ventricular rotation occurs in normal subjects and in patients with various heart diseases. Exaggerated left ventricular rotation is a uniform feature of secundum atrial septal defect. Appreciation of this phenomenon is important for interpreting data that rely on a constant relationship between epicardial surface and chest wall (i.e., STsegment mapping). We hope that recognition of exaggerated patterns of left ventricular rotation will improve our diagnostic accuracy in echocardiographically identifying patients with secundum atrial septal defects. References 1. Harvey W: Exercitatis Anatomica de Motu Cordis et Sanguinis in Animalibus. Frankfurt, 1628, ch 5
333
2. Hamilton WF, Romph JH: Movements of the base of the ventricle and the relative constancy of the cardiac volume. J Physiol 102: 559, 1932 3. Rushmer RF, Crystal DK, Wagner C: The functional anatomy of ventricular contraction. Circ Res 1: 162, 1963 4. Harrison DC, Goldblatt A, Braunwald E: Studies on cardiac dimensions in intact, unanesthetized man; description of techniques and their validation. Circ Res 13: 448, 1973 5. Rushmer RF: Physical characteristics of myocardial performance. Am J Cardiol 18: 61, 1966 6. Lynch RP, Bove AA: Geometry of the left ventricle as studied by a high speed cineradiographic technique. Fed Proc 28: 1330, 1969 7. Sandler H, Dodge HT: The use of single plane angiocardiograms for the calculation of left ventricular volume in man. Am Heart J 75: 325, 1968 8. Klein MD, Herman MV, Gorlin R: A hemodynamic study of left ventricular aneurysm. Circulation 35: 614, 1967 9. McDonald IG: The shape and movements of the human left ventricle during systole. Am J Cardiol 26: 221, 1970 10. Diamond MA, Dillon JC, Haine JC, Chang S, Feigenbaum H: Echocardiographic features of atrial septal defect. Circulation 43: 129, 1971 11. Nimura Y, Matsuo H, Matsumoto M, Kitabatake A, Abe H: Interatrial septum in ultrasonocardiotomogram and ultrasound cardiogram. Medical Ultrasound 9: 58, 1971 12. Tajik AJ, Gau GT, Ritter DG, Schattenberg TT: Echocardiographic pattern of right ventricular diastolic volume overload in children. Circulation 46: 36, 1972 13. McCann WD, Harbold NB, Giuliani BR: The echocardiogram in right ventricular overload. JAMA 221: 1243, 1972 14. Matsumoto M: Ultrasonic features of interatrial septum: its motion analysis and detection of its defect. Jpn Circ J 37: 1383, 1973 15. Radtke WE, Tajik AJ, Gau GT, Schattenberg TT, Giluiani ER, Tancredi RG: Atrial septal defect: echocardiographic observations. Ann Intern Med 84: 246, 1976 16. Feigenbaum H: Echocardiography, 2nd ed. Philadelphia, Lea and Febiger, 1976, ch 14 17. Weyman AE, Wann LS, Feigenbaum H, Dillon JC: Mechanism of abnormal septal motion in patients with right ventricular volume overload. Circulation 54: 179, 1976 18. Lieppe W, Scallion R, Behar VS, Kisslo JA: Two-dimensional echocardiographic findings in atrial septal defect. Circulation 56: 447, 1977 19. Weyman AE, Wann LS, Caldwell RL, Hurwitz RA, Dillon JC, Feigenbaum H: Negative contrast echocardiography: a new method for detecting left-to-right shunts. Circulation 59: 498, 1979 20. Perloff JK: The Clinical Recognition of Congenital Heart Disease. Philadelphia, WB Saunders, 1970, ch 15 21. Roesler H: Interatrial septal defect. Arch Intern Med 54: 339, 1934 22. Bedford DE, Papp C, Parkinson J: Atrial septal defect. Br Heart J 3: 37, 1941 23. Tavel ME: Clinical phonocardiography and external pulse recording, 2nd ed. Chicago, Year Book Medical, 1973, ch 8
