Mitral Subvalvular Apparatus
Different Functions of Primary and Secondary Chordae
Background The aim of this study was to compare the function of the primary chordae attached to the free edge with that of the secondary chordae attached to the ventricular surface of the anterior mitral leaflet.
Methods and Results An isolated working pig heart model was used. Three groups of 7 hearts were compared: Group A was the control group with intact leaflets. In group B, the primary chordae of the anterior leaflet were sectioned and the secondary chordae were left intact before assembly of the working heart model. In group C, the secondary chordae were sectioned and the primary chordae left intact. In group B, atrial and ventricular pressure evidenced dramatic mitral regurgitation. Video monitoring showed significant prolapse of the free edge of the anterior leaflet. Acute mitral regurgitation accounted for the decrease in aortic flow rate to 30 mL/min, significantly lower than in the control group (P=.006). In group C, sectioning of the secondary chordae left a competent mitral valve together with good coaptation of the anterior and posterior leaflets shown by video monitoring. However, aortic flow was lower than in the control group (P=.007), and ultrasonomicrometry evidenced impaired function (P=.009).
Conclusions This study suggests that the primary and secondary chordae of the mitral subvalvular apparatus have different functions. The primary chordae of the anterior leaflet appeared to be more involved in mitral valve competence, whereas the secondary chordae appeared to be more involved in left ventricular geometry and function.
There are several arguments in favor of preserving, insofar as possible, the internal structure of the left ventricle during mitral valve surgery.1 2 3 Such arguments favor mitral repair rather than replacement whenever possible.4 In prosthetic valve replacement, the posterior leaflet is now generally preserved by the majority of surgeons,5 6 and there is increasing emphasis on the importance of preserving the anterior mitral valve in several techniques.7 8 The dual role of left ventricular competence and function of the mitral valve is therefore acknowledged, though globally, on the basis of arguments regarding left ventricular function and postoperative course after valve replacement.9
This research aimed to investigate and determine the anatomic mechanisms enabling the mitral valve to play this dual role and to compare the different functions of primary and secondary chordae.
This investigation complied with the Guide for Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication No. 85-23, revised 1985).
The study used an isolated working heart model (Fig 1⇓). Mongrel pigs weighing 25 to 30 kg were sedated with droperidol. An intravenous line was inserted in an ear vein for Ringer’s lactate drip and infusion of thiopental and pancuronium. Breathing was controlled by tracheotomy (endotracheal intubation, tube No. 8) and mechanical ventilation (Monal; minute ventilation, 8 L; tidal volume, 450 mL; 50% oxygen). A sternotomy was then performed, and the heart was harvested by the usual method, after aortic clamping and infusion of 500 mL of crystalloid solution (St Thomas No. 2) into the aortic root. The excised hearts were placed into cardioplegic fluid at 4°C. Three groups of 7 hearts were formed by randomization (Fig 2⇓).
In group A (n=7), the control group, the left atrium was opened, but the mitral leaflets were left intact. The hearts were preserved in cold cardioplegic fluid to time 20 minutes.
In group B (n=7), after the left atrium was opened, the primary chordae of the anterior leaflet (chordae attached to the free edge) were sectioned and the secondary chordae (chordae attached to the ventricular side) left intact. This took 10 to 15 minutes. The hearts were then preserved in cold cardioplegic fluid to time 20 minutes.
In group C (n=7), after the left atrium was opened, the secondary chordae of the anterior leaflet were sectioned and the primary chordae left intact. This took 10 to 15 minutes. The hearts were then preserved in cold cardioplegic fluid to time 20 minutes.
The preparation phase therefore lasted ≈20 minutes in the three groups before assembly of the isolated heart model. This consisted of ligature of the two venae cavae and cannulation of the pulmonary artery for coronary venous return to the Krebs-Henseleit fluid recycling circuit. This model (Fig 1⇑) was directly derived from Neely’s isolated working heart model.10 The perfusion fluid (Krebs-Henseleit) was recycled with oxygenation and temperature maintenance at 37°C ensured by a membrane oxygenator (Cobe CML). The arteriovenous difference between the Po2 of the Krebs-Henseleit fluid measured in the arterial line (350 mm Hg) and the Pao2 measured in the coronary return (85 mm Hg) remained constant throughout, demonstrating good extraction, hence good-quality myocardial oxygenation. The heart was first suspended by the aortic trunk and reperfused at constant pressure (80 mm Hg) as in the conventional Langendorff model.11 During this period, the aortic valve remained closed and the heart beat empty, without ejection. The left atrium was then connected to the atrial chamber (pressure, 15 mm Hg), and air bubbles were expelled. The heart was thus in working mode, with the left ventricle being filled by the left atrium, followed by ejection toward the aorta. The left atrium and the left ventricle were then individually catheterized to allow pressure monitoring. The pressure transducer used was a rigid fluid-filled cannula connected to a pressure transducer (Medex, SX 620 449 CST).
One pair of ultrasonic crystals (sonomicrometry) was used to assess regional contractile function. The crystals were inserted via small scalpel incisions into the subepicardium in the mid anterolateral region of the left ventricle. Crystals were oriented parallel to the minor axis of the left ventricle. Segment shortening (SS) was calculated as SS=(end-diastolic length−end-systolic length)/end-diastolic length and was expressed as percentage of baseline values.12 We computed dP/dt of the left ventricular pressure and used its maximum and minimum values as end-diastolic and end-systolic time according to the reference technique.12
Three epicardial electrodes for ECG recording were positioned. These operations lasted ≈20 minutes in Langendorff mode, a phase requiring a 10-J shock to obtain defibrillation, followed by an equilibration period until a stable, regular rhythm was obtained, then another ≈20 minutes in working mode before the various recordings were initiated. All of the parameters (left atrial pressure, left ventricular pressure, and ultrasonomicrometry) were recorded continuously for 10 minutes with Lab-View software for PC. Aortic and coronary flow rates were measured directly in the aortic and pulmonary lines on three occasions during those 10 minutes. After the recording phase, a video transmitter was inserted through a left atrial bursa and a bursa located at the apex of the left ventricle in succession. Direct display of the atrial and ventricular surfaces of the mitral valve was thus possible with video recording.
Because the hearts were attached by the aortic root, modifications of the ventricular geometry could be suspected, and the hearts were also rested on a wet pad to mimic the normal anatomy of the heart in the thorax where it rests on the diaphragm.
Because of the small size of the three study groups, which did not allow any assumptions about the distribution of the studied parameters, analysis was performed with exact nonparametric tests, with statXact (version 2) software. Probability values were estimated by Monte Carlo sampling with 4000 samplings each time.13 The three groups were compared by the Kruskal-Wallis test, and in case of significant results, 2×2 comparisons were performed with the Wilcoxon test. We used Bonferroni correction of the critical probability value to take into account the number of 2×2 comparisons performed for each parameter. Consequently, differences between two groups were considered significant at P<.0167.
The mean weight of the 21 hearts selected was 175±22 g. Mean coronary flow was stable at 360±45 mL/min over the 10 minutes of measurement in the working mode. Heart rate was also stable at 75±8 bpm over these 10 minutes. There was no between-group difference in those parameters (weight, coronary output, and heart rate) of the values obtained in each of the three groups.
The results obtained during the recording phase are given in the Table⇓, and specimen curves captured and recorded with Lab-View for PC are shown for each of the three groups in Fig 3⇓. In group B, atrial and ventricular pressures showed dramatic mitral regurgitation, and video monitoring revealed marked prolapse of the free edge of the anterior leaflet (Fig 4⇓, top and middle). Acute mitral regurgitation accounted for the decrease in aortic flow rate to 30 mL/min, compared with 1100 mL/min in the control group (P=.006). However, ultrasonomicrometry showed that shortening fraction was maintained by the decrease in afterload produced by the mitral regurgitation. In group C, despite sectioning of the secondary chordae, the mitral valve remained competent, and good coaptation of the anterior and posterior leaflets was confirmed by video monitoring (Fig 4⇓, bottom). Aortic flow nonetheless decreased to 850 mL/min versus 1100 mL/min in the control group (P=.007), and ultrasonomicrometry evidenced reduced shortening fraction, 17%, versus 21% in the control group (P=.009).
Critical Assessment of the Model
The isolated working heart model has its limitations, like any experimental model; even in the control group, the hearts did not function under perfectly physiological conditions.14 15 Preload is higher with a fluid reservoir, which must be located 15 cm above the left atrium to enhance basal hemodynamic performance and, in particular, cardiac output. The circuit lines have no elastance, which increases afterload and thus myocardial work. Coronary flow is greater than the theoretical coronary flow rates (1 mL/kg myocardium). Krebs-Henseleit fluid, the rheological properties of which give rise to output greater than that obtained with blood perfusion, was actually chosen because a transparent perfusate was essential for video monitoring of the mitral valve. However, the arteriovenous difference demonstrated good perfusion quality.
Despite the above criticisms, it should be noted that when the hearts were left in working mode for longer than the study duration, the hemodynamic parameters remained stable for ≈2 hours. In Langendorff mode, the hearts functioned regularly for ≥5 hours. This relatively short study (50 minutes of reperfusion, including only 30 minutes in working mode) thus ensured the marked stability of the hearts under study. Moreover, the three groups were studied under the same conditions, and the limitations of this type of model seemed, in the final analysis, acceptable to the authors.
Primary chordae are clearly defined as the chordae inserted only on the free edge of the mitral leaflets. Conversely, secondary chordae can be inserted in various positions on the ventricular face of the leaflets, which means, theoretically, from the free edge to the annulus. In fact, on the anterior leaflet, the secondary chordae are grouped primarily in two zones, corresponding to the two papillary muscles, and located at a distance of approximately one third of the way from the free edge and two thirds from the annulus.
Discussion of the Results
It is not surprising to observe massive mitral regurgitation after rupture of the primary chordae; this actually corresponds to the classic clinical context of rupture of the chordae on dystrophic valves. This type of rupture always involves the primary chordae. Rupture induces left ventricular dysfunction in the patient. In the present experimental model, the equivalent picture is total inefficacy of left ventricular function with virtually zero output. The isolated heart model has no general adaptation resources, which explains its high susceptibility to acute mitral regurgitation. The shortening fraction remained normal in group B, but in the presence of massive mitral regurgitation, this value is difficult to interpret, so we cannot be sure that it means a conserved ventricular function. Actually, the afterload dependence of systolic myocardial shortening renders interpretation of group B data in term of left ventricular function hazardous.16 In fact, the originality of the present research lies in the analysis of the behavior of the hearts in group C, in which valvular competence remains perfect while left ventricular function is impaired.
The overall picture would seem to suggest that the two roles of the mitral valve are exercised independently. The first role, mitral valve competence, seems to be fulfilled by the peripheral part of the anterior leaflet, the free edge, and the primary chordae. The second “hemodynamic” role seems to depend on the proximal part of the anterior leaflet and the secondary chordae extending it to the mitral columns. Although the conclusions of the present study are innovative and original, the literature provides indirect evidence in support of this hypothesis. First,17 the secondary chordae are wider and stronger than the primary chordae (Fig 4⇑), probably because greater force is exerted, which would explain why they play the functional role. In addition, the histological structure of the anterior mitral leaflet18 is thicker close to the annulus than at the free edge, with a sharp decrease in thickness at the expense of the fibrous structure. This sharp decrease in thickness seems to correspond to insertion of the secondary chordae. These data are compatible with the hypothesis advanced in the present study; the stronger periannular and central part of the mitral valve appears to be part of a line of traction from the aortic annulus to the left ventricle (Fig 5⇓) via the strong secondary chordae. The second part, bordering the free edge and including the coaptation area, seems only to ensure competence under weaker mechanical stress.
The roles of the primary and secondary chordae of the posterior leaflet, not considered here, are probably very different, and the hypothesis formulated by the present study with regard to the anterior leaflet obviously does not concern the posterior leaflet.
These data may influence the choice of techniques for mitral valve repair. Differentiation between the two different roles of the primary and secondary chordae of the anterior leaflet would result in avoidance of techniques consisting of marginalizing or transferring the secondary chordae of the anterior leaflet.19 20 Transposition of chordae from the posterior leaflet21 is based on other principles not relevant to the present study.
Moreover, in the event of preservation of the mitral subvalvular apparatus during valve replacement, the various techniques could not preserve the whole of the anterior leaflet but only the proximal part. Excess preserved valvular tissue has given rise to complications.22 23 Resectioning of the peripheral edge and primary chordae should be possible without impairing function while limiting obstructive tissue.
These prospects are perhaps somewhat theoretical and to be considered in the light of clinical reality, because valve surgery24 addresses mitral valves that have often undergone marked changes and in which it may be difficult to distinguish between the different chordae. However, enhanced understanding of mitral subvalvular apparatus function must contribute to better management of mitral valve surgery while sparing the internal architecture of the left ventricle.
This study suggests that the primary and secondary chordae of the mitral subvalvular apparatus have different functions. The primary chordae of the anterior leaflet appear to be involved primarily in mitral valve competence, whereas the secondary chordae appear to be involved primarily in left ventricular geometry and function.
The authors thank Dr Michel Lièvre for statistical analysis, Dr Margueritte Perinetti for the echographic iconography, and Annie Desenfant for technical assistance.
Reprint requests to Dr Obadia, Service de Chirurgie Cardio-Thoracique, Hôpital Cardiologique, Blvd Pinel, 69003 Lyon, France.
- Received February 20, 1997.
- Revision received June 6, 1997.
- Accepted June 19, 1997.
- Copyright © 1997 by American Heart Association
David TE, Armstrong S, Sun Z. Left ventricular function after mitral valve surgery. J Heart Valve Dis. 1995;4:S175-S180.
David TE, Armstrong S, Sun Z. Left ventricular function after mitral valve surgery. J Heart Valve Dis. 1995;4:S175-S180.
Maupoil V, Rochette L. Evaluation of free radical and lipid peroxide formation during global ischemia and reperfusion in isolated perfused rat heart. Cardiovasc Drugs Ther. 1989;2:615-622.
Théroux P, Franklin D, Ross J, Kemper WS. Regional myocardial function during acute coronary artery occlusion and its modifications by pharmacologic agents in the dogs. Circ Res. 1974;35:896-908.
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. 1990;82(suppl IV):IV-58-IV-64.
Komeda M, Bolger AF, DeAnda A, Tomizawa Y, Ingels NB, Craig Miller D. Improving methods of chordal sparing mitral valve replacement, I: a new, non-distorting isovolumic balloon preparation for the left ventricle with intact mitral subvalvular apparatus. J Heart Valve Dis. 1996;5:376-382.
Shintani H, Glantz SA. Effect of disrupting the mitral apparatus on left ventricular function in dogs. Circulation. 1993;87:2001-2015.
Fontaine AA, He S, Stadter R, Ellis JT, Levine RA, Yoganathan AP. In vitro assessment of prosthetic valve function in mitral valve replacement with chordal preservation techniques. J Heart Valve Dis. 1996;5:188-198.
Smedira N, Selman R, Cosgrove D, McCarty P, Lytle B, Taylor P, Hansen C, Stewart R, Loop F. Repair of anterior leaflet prolapse: chordal transfer is superior to chordal shortening. J Thorac Cardiovasc Surg. 1996;8:287-292.
Hetzer R, Drews T, Siniawski H, Komoda T, Hofmeister J, Weng Y. Preservation of papillary muscles and chordae during mitral valve replacement: possibilities and limitations. J Heart Valve Dis. 1995;4:S115-S123.
Komeda M, David TE, Rao V, Sun Z, Weisel RD, Burns RJ. Late hemodynamic effects of the preserved papillary muscles during mitral valve replacement. Circulation. 1994;90(suppl II):II-190-II-194.