The importance of maintaining mitral annular flexibility in mitral valve surgery has long been discussed, 1–6 and valve ring prostheses that maintain such flexibility have been developed. 5–7 However, even if an ideal prosthesis, which maintains complete flexibility of the mitral annulus, were to be inserted, the expected mitral annular function would not be that of a normal heart, because the geometry of the diseased heart is different from normal. In the present study, we discuss this geometric problem by means of the study of patients after mitral repair without a prosthesis, which may be considered a substitute for repair with an ideal mitral prosthesis.
Materials and Methods
The data used in this study were obtained from examination of eight healthy men and six men who underwent mitral repair for degenerative mitral regurgitation without the insertion of a prosthetic ring. Additionally, for comparison of contraction of the mitral valve annulus (MVA) and left ventricular (LV) base, the data from seven men from a previous study, 8 who underwent mitral valve replacement (MVR) with mechanical valve prosthesis for degenerative mitral regurgitation, were used. All gave informed consent. Normal subjects in whom the left ventricular internal diameter in diastole (LVIDd) was more than 50 mm were selected for the study. The data of the normal subjects were presented in the previous study 8 except for the data on two parameters that were newly defined in the present study, i.e., proportionate share of the LVOT at the LV base (PSLVOT) and the LVB angle, which is the angle between the plane of the LVOT and the plane of the MVA.
Characteristics of the normal subjects and patients are listed in Table 1. All patients had sinus rhythm, all patients suffered from degenerative mitral regurgitation, and none had significant coronary artery disease or aortic valve disease.
The patients studied after mitral repair include one patient after Gerbode plasty, 9 one patient after Gerbode plasty and Paneth plasty 10 with our modification, 11 and four patients after mitral repair in which a pericardial strip was used for annular reinforcement after Gerbode plasty. 11 Magnetic resonance (MR) imaging was performed 6 months after the mitral repair.
Echocardiographic studies were performed before MR imaging. The diameter of the LVOT during systole in the long axis view, and the pressure gradient at the LVOT were measured in all subjects. The MR imaging, which included a four chamber view (FCV) MR scan, and a long axis scan perpendicular to the FCV scan (LAPF), was performed with a 1.5 T imager (Magnetom SP-63; Siemens Medical Systems, Erlangen, Germany).
A gradient echo sequence was used for the imaging of the MVA and the orifice of the LVOT. The LVOT orifice has been defined in detail in a previous study. 12 In brief, the LVOT orifice consists of the edge of the interventricular septum (muscular region) and the annulus of the anterior mitral leaflet (annular region). With the gradient echo sequence 13 images, one every 50 ms with a 0 ms to 600 ms delay from the electrocardiogram (ECG) R wave, were obtained in both FCV and LAPF scans. Temporal resolution was 50 ms. Parameters for spatial resolution were a matrix size of 256 × 256, a field of view of 350 mm, and a slice thickness of 5 mm for both the normal subjects and the patients. The slice interval was 5 mm in the normal subjects and 5 mm (five patients) or 6 mm (one patient) in the patients after mitral repair.
A spin echo sequence was used to image an anterior chest marker. 12 Parameters for spatial resolution were a matrix size of 256 × 256, a field of view of 350 mm, slice thickness of 5 mm, and a slice interval of 5 mm in all subjects.
The MVA, LVOT orifice, and anterior chest marker were reconstructed three dimensionally using PATRAN (MSC Software Corporation, Santa Ana, CA) application software. From the FCV and LAPF MR images, contour points for both the MVA and LVOT orifice were obtained. Incorporating these contour points, contours of the MVA and LVOT orifice were developed. The reconstructed contours of the MVA obtained from the FCV and LAPF images were superimposed by aligning the anterior chest marker. 12 Contour points for these two superimposed contours were extracted and used as contour points for a composite annular contour. Incorporating these points, the composite contour was developed. The LVOT orifice was reconstructed in the same manner as the MVA. Thus, the MVA and LVOT orifice for four cardiac phases, i.e., 0, 100, 200, and 300 ms delay from the ECG R wave, were reconstructed in both the normal subjects and patients after mitral repair.
We defined the plane of the LV base as a least squares plane of both the central portion of the posterior MVA and the central portion of the muscular segment of the LVOT orifice. 13 The images of the MVA and LVOT orifice were projected onto this plane and the area of these projected images, i.e., that for MVA (SM) and LVOT orifice (SL), were calculated. The area of the LV base (SB) was defined as (SB = SM + SL). The proportionate share of the LVOT at the LV base (PSLVOT) was defined as (SL/SB).
We defined the plane of the MVA and the plane of the LVOT as the least squares plane of the MVA and LVOT orifice, respectively. The images of MVA and LVOT orifices were projected to the least squares plane of each image, and each projected image was used for calculation of its area.
The mitrobasal angle (MB angle) was defined as the angle between the plane of the LV base and the plane of the MVA. The LVB angle was defined as the angle between the plane of the LVOT and the plane of the MVA (Figure 1).
Data are given as mean ± standard deviation unless otherwise stated. Comparisons in each group of subjects were made using Student’s t-test for paired differences. Comparisons among the groups of subjects were made using Student’s t-test for nonpaired differences. Comparisons among the groups of subjects both for characteristics of the subjects and for the change from 0 ms delay to 300 ms delay, were made using the Wilcoxon rank sum test. All comparisons were considered statistically significant when p was < 0.05.
Echocardiographic data concerning LVOT are listed in Table 2. In both the normal heart and the hearts of patients after mitral repair (postrepair hearts), the diameter of the LVOT increased during early systole to reach a maximum and decreased during mid and late systole to reach a minimum. In the hearts of patients after MVR (post-MVR hearts), at 0 ms delay the diameter of the LVOT was at its maximum and decreased during systole to reach a minimum at 306.5 ± 45.0 ms delay. The pressure gradient at the LVOT in each group was within the normal range.
The timing of the beginning of ejection flow at the LVOT in the normal and post repair hearts was 63.6 ± 8.3 ms and 101.3 ± 23.2 ms delay from ECG R wave, respectively. Timing of peak ejection flow at the LVOT in the normal and postrepair heart was 140.5 ± 16.7 ms and 180.2 ± 24.9 ms delay from ECG R wave, respectively.
Data Obtained from MR Imaging
Figures 2 and 3 show the frontal view of the LV base (i.e., the MVA and the LVOT orifice) in a patient suffering from mitral regurgitation due to degenerative mitral valve prolapse and that after mitral repair with Gerbode plasty and pericardplasty, respectively.
Systolic changes in the area of the LV base and the MVA in each group of subjects are shown in Figure 4. Both the LV base and MVA in the postrepair heart contracted to almost the same extent as the normal heart, i.e., in the normal heart and the postrepair heart, the reduction rate of the area of the LV base in systole was 23.9 ± 3.4% and 21.9 ± 4.7%, respectively, and that of the MVA in systole was 21.5 ± 3.3% and 17.9 ± 5.6%, respectively. In contrast, in the post-MVR heart, contraction of the LV base was weak (p < 0.05 vs. postrepair heart, p < 0.01 vs. normal heart, compared for the change from 0 ms delay to 300 ms delay, Wilcoxon rank sum test), and there was no contraction of the MVA (p < 0.01 vs. postrepair heart and p < 0.01 vs. normal heart).
Changes in the area of the LVOT orifice in the normal and postrepair hearts during systole are shown in Figure 5. The LVOT orifice was enlarged at 100 ms delay and then contracted during mid and late systole. The area of the LVOT orifice in the postrepair heart was approximately 1 cm2 larger than the normal value during systole.
Changes in PSLVOT during systole in the normal and postrepair hearts are shown in Figure 6. The PSLVOT increased at 100 ms and was maintained at a level a little higher than that at 0 ms delay during mid and late systole. The value remained approximately 40% in the normal heart and approximately 50% in the postrepair heart.
Changes in the MB angle in the normal and postrepair hearts are shown in Figure 7. The MB angle gradually increased during systole. This increase indicates dorsiflexion of the MVA. 14 The MB angle in the postrepair heart was 7.3 to 9.0 degrees greater than that in the normal heart during systole (p < 0.001 at each phase).
Changes in LVB angle in the normal and postrepair hearts are shown in Figure 8. The LVB angle gradually decreased during systole. The LVB angle in the postrepair heart was 10.5 to 12.5 degrees smaller than that in the normal heart during systole (p < 0.001 at each phase).
Data for the geometry of the LV base at 100 ms delay in the normal and postrepair hearts are listed in Table 3 and summarized in the schematic illustration in Figure 9. LV contraction does not yet begin at 0 ms delay, whereas the contraction of the MVA has already begun at that point. Therefore, we used data at 100 ms delay for comparison of geometry in LV systole. The data at 200 and 300 ms delay showed the same tendency as those at 100 ms delay.
Valve ring prostheses 5–7 have been developed that maintain mitral annular flexibility, i.e., both contraction/dilation 3,15–17 and dorsiflexion. 14,18,19 The effect of insertion of the prosthetic ring on contraction/dilation of the MVA has been studied in animals 3 and humans. 20 Dilation of the prosthetic ring in diastole is of significance in diastolic cardiac function. 3 Furthermore, the reduction of its size after the contraction of the heart is also important to maintain cardiac function. Insertion of the rigid prosthetic valve fixes the mitral annulus and causes suppression of contraction of the cardiac muscle of the LV inflow tract; this causes a hindrance of cardiac function. 1 Our previous study showed the decrease of contraction of the LV base due to fixation of the mitral annulus. 8 The present study shows that mitral repair without a prosthesis does not disturb contraction of the MVA or LV base.
It is important that the MVA maintains its saddle shape and shows dorsiflexion in systole 14,18,19 from the viewpoint of the prevention of LVOT narrowing, because insertion of a rigid or semirigid ring prosthesis flattens the anterior MVA, making it protrude into the LVOT, resulting in LVOT narrowing. 2–4,19
Dorsiflexion of the MVA in systole is merely the passive change of the annulus due to the discrepancy between the movement of the posterior MVA and that of the aortic root, including the anterior MVA. 14 In an experiment in sheep, dorsiflexion has been shown as the change of the angle between the anterior MVA and posterior MVA during systole by means of implanting radiopaque markers on the MVA. 19 However, as far as we have investigated this in human hearts, there is no obvious hinge point where the MVA flexes, but a gradual enhancement of the curving shape of the MVA in systole. 14 Therefore, we measured dorsiflexion as a change in the height of the anterior MVA from the approximated plane of the posterior MVA. The parameters for dorsiflexion given above merely express the extent of the dorsiflexion. Therefore, from the reduction in the value of these parameters, we cannot estimate the extent of the hindrance of the LVOT due to the disturbance of dorsiflexion.
We have defined the plane of the LV base as a least squares plane of both the central portion of the posterior MVA and the central portion of the muscular segment of the LVOT orifice. 13 The discrepancy between systolic movement of the plane of the LV base and that of the aortic root, including the anterior MVA, is expressed as the change in the MB angle. In the normal heart, the MB angle increases in systole associated with the dorsiflexion of the MVA. However, the value of the MB angle is affected not only by dorsiflexion in the MVA, but also by muscular contraction of the LVOT. When dorsiflexion of the MVA is completely disturbed, which is caused by the insertion of a rigid prosthetic valve into the mitral annulus, the decrease in MB angle in systole has been demonstrated. 8 This phenomenon suggests that the rigid prosthetic valve tilts toward the plane of the LV base causing narrowing of the LVOT. In the present study, the hearts of the patients after mitral repair showed the same gradual increase in the MB angle in systole as the normal heart, and this finding suggests that mitral annular function is maintained after mitral repair.
The PSLVOT is the parameter of LVOT narrowing. When dorsiflexion of the MVA is completely disturbed, the PSLVOT decreases in systole. 8 In the normal heart, the PSLVOT increases slightly between 0 ms and 100 ms delay from the ECG R wave, and after 100 ms delay it remains almost constant. This pattern of change in PSLVOT was also seen after mitral repair.
As described above, mitral repair without a ring does not hinder contraction of the MVA and LV base and does not cause LVOT narrowing, indicated by dorsiflexion of the MVA. Thus, mitral annular function is maintained as normal after surgical intervention, although there remains a problem originating from the geometric difference between the normal heart and the diseased heart. The differences are listed in Table 3 and are shown as a schematic drawing in Figure 9. The size of the MVA and LV base of the postrepair heart is almost the same as that of a relatively large normal heart, although the size of the LVOT orifice of the postrepair heart is larger than in the normal heart. As a result, the MB angle of the postrepair heart is larger than normal; furthermore, the LVOT orifice in the postrepair heart occupies more space in the LV base than in the normal heart.
The LVB angle, which is defined as the angle between the approximated plane of the MVA and the plane of the LVOT orifice, may not be a substitute for the mitroaortic angle, 21 which is defined as the angle between the MVA and the aortic annulus. However, the tendency of the change in the angle in systole may be the same, because both the LVOT orifice and the aortic annulus are involved in the LVOT. The LVB angle in both the normal and postrepair hearts decreased gradually during systole, and the LVB angle in the postrepair heart was always smaller than that in the normal heart during systole (Figure 8). The decrease in mitroaortic angle in systole has already been shown in an animal experiment. 19
The MB angle in both the normal and postrepair hearts gradually increased during systole, the angle in the postrepair heart being larger (Figure 7). We consider these findings to be of pathophysiologic significance. In contrast with the animal heart, the apex of the human heart is shifted to the left because of the flat thorax. Therefore, the direction of the LV ejection flow, which is at first mainly left-to-right in the LV cavity, is changed at the LVOT to mainly posteroanterior and, furthermore, at the aortic root to caudocranial. When the ejection flow changes its direction, guided by the anterior mitral leaflet at the LVOT, the shape of this leaflet becomes like that of a fully spread sail, due to the force of ejection flow, and this results in an increase in the tension on the chordae of the anterior leaflet (Figure 10). The smaller the MB angle, the greater the force generated by ejection flow at the anterior mitral leaflet becomes and the greater the tension on the chordae of the anterior leaflet becomes. Sufficient tension on the chordae prevents systolic anterior motion (SAM) of the mitral valve. 22 Therefore, it is physiologically reasonable from the point of prevention of SAM that the MB angle at 0 ms delay is smallest during systole.
In the postrepair heart, the MB angle is approximately 10° greater than that in the normal heart during systole. This finding suggests that the postrepair heart generally has the potential to induce SAM due to a higher MB angle at early systole. We assume that the anterior leaflet maintaining a fully spread configuration due to the ejection flow, resulting in sufficient tension being generated on the chordae, prevents SAM. This condition may be interfered with by several anatomic factors, such as a large redundant anterior mitral leaflet, 23 a large posterior leaflet coapting with the anterior leaflet close to the LVOT, 24,25 and slacking of the chordae of the anterior leaflet caused by displacement of the anterior annulus into the LVOT due to the prosthetic ring, 19 all of which have been considered as factors that induce SAM after mitral repair. In addition, the marked increase in the MB angle at early systole may be included in these factors as it weakens the anti-SAM function of the anterior leaflet.
Limitation of Study
To study normal hearts of relatively large size, we selected hearts of medical student athletes. Therefore, there is a significant difference in the age and heart rate between the normal group and patient groups, i.e., the patients are older than the normal subjects, and the heart rate at rest of the patients is higher than that of the normal subjects (Table 1). Although aging affects diastolic function of the left ventricle (LV) due to increase in the stiffness of the LV wall, the present study focused on the systolic change of the LV base and the systolic function of the LV indicated by the LV ejection fraction showed no differences between the normal and patient groups (Table 1). Lower heart rate in the normal group suggests a longer R-R interval, which mainly prolongs the diastolic period. When the QT interval was compared between normal and patient groups, there was little difference (Table 1). Therefore, we consider that the timing of the measurement at 300 ms delay from the ECG R wave corresponds to late systole in the normal as well as the patient groups. Furthermore, the area of the LV base at 300 ms delay showed no differences between the normal group and the postrepair group (Figure 4).
If mitral prostheses that maintain complete flexibility of the mitral annulus are developed, the dynamics of the mitral annulus may be maintained as normal after surgery. However, the problem of altered geometry of the heart, i.e., the higher MB angle at early systole, still remains to be solved.
We wish to thank Dr. Waldemar Wlodarczyk of the Department of Radiology, Charité Campus Virchow, Berlin, Germany, for his technical assistance in MR imaging, and both Dr. Naozo Sugimoto of Department of Systems Science, Graduate School of Informatics, Kyoto University, Japan, and Mr. Detlef Goesmann of the German Heart Institute Berlin, Germany, for their technical assistance in 3D imaging. We are indebted to Mrs. Julia Stein of the German Heart Institute Berlin for her valuable advice in the statistical analysis of this study. We also thank Ms. Anne M. Gale of the German Heart Institute Berlin for her editorial assistance.
1. Spence PA, Peniston CM, David TE, et al: Toward a better understanding of the etiology of left ventricular dysfunction after mitral valve replacement: An experimental study with possible clinical implications. Ann Thorac Surg 41: 363–371, 1986.
2. Schiavone WA, Cosgrove DM, Lever HM, Stewart WJ, Salcedo EE: Long-term follow-up of patients with left ventricular outflow tract obstruction after Carpentier ring mitral valvoplasty. Circulation 78 (Suppl): I60–I65, 1988.
3. van Rijk-Zwikker GL, Mast F, Schipperheyn JJ, Huysmans HA, Bruschke AVG: Comparison of rigid and flexible rings for annuloplasty of the porcine mitral valve. Circulation 82 (Suppl): IV58–IV64, 1990.
4. Duran CMG: Perspectives in reparative surgery for acquired valvular disease. Adv Card Surg 4: 1–23, 1993.
5. Cosgrove DM III, Arcidi JM, Rodriguez L, Stewart WJ, Powell K, Thomas JD: Initial experience with the Cosgrove-Edwards annuloplasty system. Ann Thorac Surg 60: 499–504, 1995.
6. Carpentier AF, Lessana A, Relland JY, et al: The “physio-ring”: an advanced concept in mitral valve annuloplasty. Ann Thorac Surg 60: 1177–1186, 1995.
7. Duran CG, Ubago JLM: Clinical and hemodynamic performance of a totally flexible prosthetic ring for atrioventricular valve reconstruction. Ann Thorac Surg 22: 458–463, 1976.
8. Komoda T, Hetzer R, Siniawski H, et al: Effects of prosthetic valve placement on mitral annular dynamics and the left ventricular base. ASAIO J 47: 60–65, 2001.
9. Gerbode F, Kerth WJ, Osborn JJ, Selzer A: Correction of mitral insufficiency by open operation. Ann Surg 155: 846–854, 1962.
10. Burr LH, Krayenbühl C, Sutton MSTJ, Paneth M: The mitral plication suture. A new technique of mitral valve repair. J Thorac Cardiovasc Surg 73: 589–595, 1977.
11. Komoda T, Hübler M, Siniawski H, Hetzer R: Annular stabilization in mitral repair without a prosthetic ring. J Heart Valve Dis 9: 776–782, 2000.
12. Komoda T, Hetzer R, Oellinger J, et al: Imaging of the orifice of the left ventricular outflow tract: technique and initial results. Technol Health Care 5: 207–217, 1997.
13. Komoda T, Hetzer R, Hofmeister J, et al: Dynamic change in the left ventricular base with or without a rigid mitral valve prosthesis. ASAIO J 43: M392–M395, 1997.
14. Komoda T, Hetzer R, Oellinger J, et al: Mitral annular flexibility. J Card Surg 12: 102–109, 1997.
15. Davis PKB, Kinmonth JB: The movements of the annulus of the mitral valve. J Cardiovasc Surg 4: 427–431, 1963.
16. Tsakiris AG, von Bernuth G, Rastelli GC, Bourgeois MJ, Titus JL, Wood EH: Size and motion of the mitral valve annulus in anesthetized intact dogs. J Appl Physiol 30: 611–618, 1971.
17. Ormiston JA, Shah PM, Tei C, Wong M: Size and motion of the mitral valve annulus in man. I. A two-dimensional echocardiographic method and findings in normal subjects. Circulation 64: 113–120, 1981.
18. Levine RA, Handschumacher MD, Sanfilippo AJ, et al: Three-dimensional echocardiographic reconstruction of the mitral valve, with implications for the diagnosis of mitral valve prolapse. Circulation 80: 589–598, 1989.
19. Dagum P, Green GR, Glasson JR, et al: Potential mechanism of left ventricular outflow tract obstruction after mitral ring annuloplasty. J Thorac Cardiovasc Surg 117: 472–480, 1999.
20. Okada Y, Shomura T, Yamamura Y, Yoshikawa J: Comparison of the Carpentier and Duran prosthetic rings used in mitral reconstruction. Ann Thorac Surg 59: 658–663, 1995.
21. Mihaileanu S, Marino JP, Chauvaud S, et al: Left ventricular outflow obstruction after mitral valve repair (Carpentier’s technique). Proposed mechanisms of disease. Circulation 78 (Suppl): I78–I84, 1988.
22. Yoganathan AP, Lemmon JD Jr, Kim YH, Levine RA, Vesier CC: A three-dimensional computational investigation of intraventricular fluid dynamics: Examination into the initiation of systolic anterior motion of the mitral valve leaflets. J Biomech Eng 117: 94–102, 1995.
23. Grossi EA, Steinberg BM, LeBoutillier M III, et al: Decreasing incidence of systolic anterior motion after mitral valve reconstruction. Circulation 90: II195-II197, 1994.
24. Lee KS, Stewart WJ, Lever HM, Underwood PL, Cosgrove DM: Mechanism of outflow tract obstruction causing failed mitral valve repair. Anterior displacement of leaflet coaptation. Circulation 88: II24–II29, 1993.
25. Jebara VA, Mihaileanu S, Acar C, et al: Left ventricular outflow tract obstruction after mitral valve repair. Results of the sliding leaflet technique. Circulation 88: II30–II34, 1993.