Treatment of Ruptured or Elon ated Anterior Mitral Valve Chordae bv Parti 3 Transposition of the Posterior Leaflet: Experience with 29 Patients

Transcription

1 Treatment of Ruptured or Elon ated Anterior Mitral Valve Chordae bv Parti 3 Transposition of the Posterior Leaflet: Experience with 29 Patients Arrigo Lessana, M.D., Mauro Romano, M.D., Genevieve Lutfalla, M.D., Carmine Carbone, M.D., Evelyne Palsky, M.D., Sid-Ali Amalou, M.D., and Marcio Escorsin, M.D. ABSTRACT We report a series of 29 patients, 5 to 75 years of age (mean age, 31.8 f 21.4 [SD] years), with pure mitral regurgitation caused by ruptured or elongated chordae of the anterior mitral leaflet. These patients underwent mitral valve repair by segmental transposition of the posterior leaflet with its attached chordae sutured to the free edge of the flail anterior leaflet. There were 2 hospital deaths. Follow-up ranged from 1 to 35 months (mean follow-up, 14.9 f 8.5 months). One patient is lost to follow-up. Two patients are in New York Heart Association Functional Class 11; all others are in Class I. In 17 patients there is no detectable murmur; in 5 patients a mild to moderate systolic murmur can be detected, while 4 have a marked systolic murmur. The adequacy of the repair could be confirmed by Doppler echocardiography, which has shown no evidence of prolapse in 22 patients. A mild regurgitation jet is present in 4 patients, and a marked jet, in 3. Postoperative cardiac catheterization performed in 5 patients has confirmed the Doppler echocardiographic findings. Although longer follow-up is necessary, this technique appears adequate for repairing a major prolapse of the anterior leaflet caused by multiple ruptured or elongated chordae, therefore obviating the need for a prosthetic valve substitute. Prolapse of the anterior leaflet of the mitral valve caused by rupture or elongation of the chordae has long been considered a contraindication to conservative surgical procedures [l-31. Although some experimental attempts have been made to transfer mitral valve chordae, they have never been put into practice for clinical use [4]. In 1983, Carpentier [5] proposed a different technique, in which a ruptured chorda of the anterior leaflet was replaced by a chorda transferred from the posterior leaflet. Extending this concept, we have operated on 29 patients in whom the prolapsed anterior mitral valve leaflet resulting from ruptured or elongated chordae, or From the Service de Chirurgie Cardio-Vasculaire, CHPA "La Roseraie,"Aubervilliers, France, the Service de Cardiologie, HBpital Cochin, Paris, France, and the Clinique Pediatrique, Bou Ismail, Algerie. Presented at the Twenty-third Annual Meeting of The Society of Thoracic Surgeons, Toronto, Ont, Canada, Sept 21-23, Address reprints requests to Dr. Lessana, CHPA "La Roseraie," Service de Chirurgie Cardio-Vasculaire, 120 Avenue de la Republique, Aubervilliers. France. both, has been repaired by a transposition of part of the posterior leaflet with its attendant chordae [3], the bundle of the posterior leaflet being sutured to the prolapsing anterior leaflet. This experience forms the basis of the present report. Patients and Methods Twenty-nine consecutive patients were operated on between June, 1984, and April, Their age ranged between 5 and 75 years (mean age, 31.8 f 21.4 [SD] years). Ten children were younger than 14 years of age. There were 6 male and 23 female patients. Twenty-one patients had pure mitral insufficiency, while 8 had mitral stenosis as well. The cause was rheumatic fever in 16 cases, bacterial endocarditis in 5, degenerative disease of the mitral valve in 7, and traumatic rupture in I. Twelve patients were in New York Heart Association (NYHA) Functional Class 11, while 16 were in Class I11 and 1 child was in Class IV. The average cardiothoracic ratio was Mean pulmonary artery pressure was (SD) mm Hg, mean pulmonary capillary pressure was 17.9 f 6.8 mm Hg, and cardiac index was L/min/m*. Seventeen patients had a preoperative angiographic study, which showed that the mitral insufficiency was Stage 3 or 4 of the classification of Bjork and co-workers [6] (Fig 1A). In addition, 10 patilents had marked functional tricuspid insufficiency and 2 patients had aortic insufficiency. One patient had aortic vegetations without stenosis or insufficiency. Operative Technique The operation was done with the patient under moderate hypothermia (24 C) and with the aid of crystalloid cardioplegic arrest. The average aortic cross-clamp time was minutes. The average myocardial septa1 temperature was "C. Evaluation of the mitral valve apparatus was performed before arresting the heart. This method allows a more accurate assessment of the amplitude of valve motion. All patients were found to have prolapse of the anterior mitral leaflet caused by multiple elongated chordae in 15 patients, multiple ruptured chordae (notably principal and paramedial [7]) in 8, and both ruptured and elongated chordae in 6. In addition, all patients had marked dilatation of the rnitral valve annulus. The technique of partial transposition of the posterior leaflet was done in the following fashion 181: Two chordae of the posterior leaflet, which face the prcilaps- 404 Ann Thorac Surg 45:404408, Apr Copyright by The Society of Thoracic Surgeons

2 405 Lessana et al: Mitral Valve Repair ing part of the anterior leaflet, are selected. A strip measuring 3 x 8 to 3 x 12 mm is cut from the free edge of the posterior leaflet so as to include these chordae (Fig 2A). The transposed posterior leaflet is sutured with a series of separate 5-0 polypropylene stitches to the free edge of the anterior leaflet (Fig 28). The procedure is completed by a quadrangular resection of the remaining posterior leaflet (Fig 2C) [2, 91 and by the insertion of a Carpentier prosthetic ring. The aortic cross-clamp is then removed, and the competency of the mitral valve is assessed after the left ventricle has been filled (Fig 2D). Treatment of the Associated Lesions In 8 patients, fibrosis and retraction of the posterior leaflet necessitated its mobilization by resection of basal chordae, commissurotomy, and sometimes chordal fenestration. In 4 patients, in addition to the transposition, it was necessary to shorten chordae by plication and insertion into the previously split corresponding papillary muscle [l, 91. In 19 patients with tricuspid insufficiency caused by an annular dilatation, we inserted a Fig 1. Left ventricular angiograms showing severe mitral regurgitation before repuir (A) and mitral competence after repair (B). Fig 2. (A) A strip of the posterior leaflet is separated (SPL) with its chordae. (B) The transposed strip of the posterior leaflet is then sutured to the free edge of the anterior leaflet (AL). (C) Next, the quadrangular resection (SQR) of the posterior leaflet (PL) is sutured. (D) Finally, the reconstructed mitral apparatus is tested after insertion of the Carpentier ring.

3 406 The Annals of Thoracic Surgery Vol45 No 4 April 1988 Fig 3. Postoperative pulsed Doppler recording of mitral flow velocity. The signal Doppler shows two flow peaks, one in early diastole (e) and one in late diastole (a). Note the absence of systolic flow Carpentier prosthetic ring [l]. Two patients had an aortic valve replacement. In 1 patient, only ablation of aortic valve vegetations was performed. Results There were 2 hospital deaths, 1 due to severe arrhythmia and 1 due to respiratory failure. Seventeen patients needed inotropic support in the early postoperative hours, and 1 required intraaortic balloon pumping. Follow-up ranged from 1 to 35 months (mean follow-up, months). One patient living abroad was lost to follow-up. The 26 remaining patients have all exhibited functional improvement: 24 are now in NYHA Class I, and only 2 remain in NYHA Class 11. The cardiothoracic ratio has decreased from 0.57 * 0.7 to Of the 8 patients who were in atrial fibrillation before operation, 3 returned to sinus rhythm after the procedure. Seventeen patients have no audible heart murmur. Five patients have Grade 116 systolic murmurs, and 4 patients have Grade 216 to 316 murmurs. There have been no late deaths, nor has a thromboembolic episode been documented during follow-up. Only 5 patients (19.2%) are receiving anticoagulant therapy, 4 because of atrial fibrillation, 1 because of an associated mechanical aortic valve. None of the patients has undergone reoperation. Five random patients had cardiac catheterization performed one month after the mitral valve repair. The hemodynamic values had returned to normal. Mean pulmonary artery pressure was 24.0 * 8.09 mni Hg preoperatively and * 6.08 mm Hg postoperatively. Mean pulmonary capillary pressure was mm Hg preoperatively and 12 * 3.81 mm Hg postoperatively. Mean cardiac index was I,/min/m2 preoperatively and 2.8 * 0.68 L/min/m2 postoperatively. The left ventricular angiography showed a totally competent mitral valve (Fig 1B) in 4 patients, while 1 patient had mild mitral insufficiency. The latter had a Grade 2/6 systolic murmur. All patients underwent a postolperative pulsed Doppler echocardiogram during the tlnreemonth period of data collection. Pulsed Doppler echocardiograms confirmed the absence of any mitral regurgitation in 19 patients (Fig 3) and the presence of a moderate regurgitant jet in 4 patients. There is marked regurgitation in 3 patients, 2 of whom remain in NYHA Class 11. Comment There is now an increasing awareness that valve replacement yields less than optimal long-term results, regardless of whether a mechanical device or a bioprosthesis is used. This realization has provided a new impetus to the search for surgical alternatives in patients operated on for valve disease, among which conservative procedures are prominent. In this setting, the treatment of multiple ruptured (or elongated) chordae (notably principal and paramedial [7]) of the anterior leaflet still lacks an acceptable solution. Triangular resection of the anterior leaflet must be very small to compensate for the lack of tissue and can be used only for minimal prolapse of the anterior leaflet [l-31. The attempts at replacing the chordae [4] have

4 407 Lessana et al: Mitral Valve Repair yielded mixed results, as the exact length of the chordal substitute (regardless of whether it is pericardium or prosthetic material) is difficult to assess. When multiple chordae are ruptured, it is even more difficult to determine the length of each chordal substitute. We have extended the concept of transposing a group of chordae of the posterior cusp [5, 101 to that of transposition of a bundle of the posterior cusp with its multiple chordal attachments. Since dilatation of the mitral annulus is consistently present, it is always possible to remove a good chunk of the cusp and still approximate the edges without undue tension. In this series we have consistently been able to remove 8 to 12 mm of the cusp. We have found that it is always necessary to extend the wedge excision of the posterior cusp into a quadrangular resection [2] to avoid undue tension on the posterior leaflet. The natural length of the posterior chordae enables them, once transposed to the anterior cusp, to ensure adequate motion of the latter without prolapse. Further, they remain viable, soft, and pliable without retraction. The use of the Carpentier prosthetic ring allows the remodeling of the mitral annulus and, above all, reduces the tension on the sutures by evenly dividing the traction forces. The cause of the mitral valve disease seems to affect the results [3]. In this small series, 6 of the 7 patients with a mild or a marked regurgitation at Doppler echocardiography had rheumatic disease. The best results were obtained in the degenerative lesions (5 patients) and in the bacterial ruptures of chordae (5 patients). Mitral valve repair in children is still difficult [3]; thus, although the overall functional results in this subset of patients were satisfactory, of the 7 patients who had mild or marked regurgitation demonstrated by Doppler echocardiography, 5 were under the age of 14 years. In conclusion, this technique allows the prolapsed anterior leaflet to be equilibrated with several new chordae and, in this small series, yields good mediumterm clinical and echocardiographic results. We thank Mrs. Sylvie Gallone and Mrs. Sylvie Brade for their expert secretarial assistance. References 1. Carpentier A, Chauvaud S, Fabiani JN, et al: Reconstructive surgery of mitral valve incompetence: ten year appraisal. J Thorac Cardiovasc Surg 79:338, Carpentier A, Relland J, Deloche A, et al: Conservative management of the prolapsed mitral valve. Ann Thorac Surg 26:294, Lessana A, Tran Viet T, Ades F, et al: Mitral reconstructive operations: a series of 130 consecutive cases. J Thorac Cardiovasc Surg 86:553, Frater RWN, Gabbay S, Shore D, et al: Reproducible replacement of elongated or ruptured mitral valve chordae. Ann Thorac Surg 35:14, Carpentier A: Cardiac valve surgery: the "French correction." J Thorac Cardiovasc Surg 86:323, Bjork VO, Lodin H, Malers E: The evaluation of the degree of mitral insufficiency by selective left ventricular angiocardiography. Am Heart J 60:691, Acar C: Anatomie de la valve mitrale. L'information cardiologique. 8:787, Lessana A, Escorsin M, Romano M, et al: Transposition of posterior leaflet for treatment of ruptured main chordae of the anterior mitral leaflet. J Thorac Cardiovasc Surg 89:804, Antunes MJ, Magalhaes MP, Colsen PR, Kinsley RH: Valvuloplasty for rheumatic mitral valve disease: a surgical challenge. J Thorac Cardiovasc Surg 94:44, Spencer FC, Colvin SB, Culliford AT, Isom OW: Experiences with the Carpentier techniques of mitral valve reconstruction in 103 patients ( ). J Thorac Cardiovasc Surg 90:341, 1985 Discussion DR. STEPHEN B. COLVIN (New York, NY): Chordal elongation without rupture should be treated by one of a variety of shortening techniques, usually at the level of the papillary muscle. A flail leaflet can be fixed by resuspension of the area on the free margin of the anterior leaflet to adjacent secondary chordae, or with a segment of mural leaflet transposed and sutured to the free margin of the anterior leaflet. Resuspension is the first choice if there are good adjacent secondary chordae. Occasionally the chordae rupture at the level of the valve leaflet, and in that case it is easier to resuspend to a secondary chordae that inserts into the middle of the leaflet. If the chordal rupture is near the tip of the papillary muscles such as seen with ischemia, then it is usually necessary to transpose a segment of the mural leaflet up to the free margin. The Carpentier ring is used routinely. We now have over 250 cases of mitral reconstructions. Forty-three patients had flail anterior leaflets with ruptured chordae. In most cases we resuspend the flail margin to secondary chordae, and in the remaining cases we transpose mural leaflet tissue up to the free margin of the anterior leaflet. Leaflet resection is rarely used. There were no hospital deaths in this group of 43 patients as opposed to 1% mortality in a group of about 150 patients studied, and at 5 years the survival free from cardiac death appears to be greater than 95%. Freedom from the need of further mitral valve surgery is also about 95%. Dr. Lessana, have you done any of the chordal shortening procedures when you have just had chordal elongation and how have the results been; and have you tried any resuspension to secondary chordae adjacent to these flail segments? DR. DOMINIQUE ROMAIN METRAS (Marseille, France): I compliment Dr. Lessana for an outstanding experience and support what is really a clever and wonderful operation derived from the Carpentier procedure. We have some experience with this procedure in the pediatric group, mainly in patients coming from developing countries such as those patients in Dr. Lessana's series. We have done 53 conservative procedures in children with mitral insufficiency, mostly rheumatic, but also in patients with endomyocardial fibrosis during a Black African experience. Among these, 46 had prolapse of the anterior leaflet mostly caused by elongation of the chordae. We have performed 48 annuloplasties, mostly without a prosthetic ring, using a kind of DeVega annuloplasty in most cases. The prolapsed leaflet was treated by a chordae-short-

5 408 The Annals of Thoracic Surgery Vol 45 No 4 April 1988 ening Carpentier procedure. However, 5 of the patients had an operation similar to the one described by Dr. Lessana. The lesions were identical in these 5 children: elongation or rupture of the chordae and annular dilatation, always at the same level-the posterior medial part of the posterior leaflet and commissural leaflet. The procedure consisted of quadrangular resection of the posterior mitral valve of the dilated area, transfer to the anterior leaflet, and suture of the resected portion. The so-called annulus was reinforced at that place by a strip of Gore-Tex. Thus, we have avoided the use of a prosthetic ring in a child. The last 2 patients were 4 and 5 years old. I think this nice operation of valve leaflet transfer is a safer operation than chordae shortening in some children in whom the elongated chordae are very thin and can secondarily rupture, as occurred after chordae shortening in 2 of our patients. This operation of leaflet transfer is very useful in children. My questions to Dr. Lessana concern the use of a prosthetic ring. Do you really think it is mandatory, and especially in children when you expect a growth of the orifice? What would you advocate to reinforce the annular repair if no prosthetic ring is used? DR. ROBERT w. M. FRATER (Bronx, NY): I too would like to congratulate Dr. Lessana for an elegant piece of surgery following on his mentor's example. There is, of course, a theoretical argument against the use of the posterior cusp to solve an anterior cusp problem. If there is a great deal of posterior cusp available, then it is quite reasonable to sacrifice some of it. However, normally functioning tissue is being taken from the posterior cusp which must then in turn be corrected by shortening the mural annulus. For this reason we became interested some years ago in using artificial chordae as an alternative. In using artificial chordae, clearly what we have to prove is that the durability of an artificial chord is as good as the durability of natural living tissue. Treated pericardium has been used for artificial chordae. This was done in an experimental animal who survives a year and a half later. Obviously, the real crux comes in human subjects. Of those patients in whom we have used chordae made from pericardium, we now have 9 patients who have gone beyond the magical 5-year mark, and all are continuing to maintain constant chordal length and competence. Thus, there are alternative techniques besides posterior cusp transposition, although, indeed, this is an elegant technique and Dr. Lessana's results are admirable. DR. TIRONE E. DAVID (Toronto, Ont, Canada): This operation is simpler to describe than to perform. Those of you who have done it, know how complex it can be to transfer a segment of the posterior leaflet to correct a prolapsing anterior leaflet, and at the end of the operation have a competent and nonstenotic mitral valve. We have performed this operation in a small number of patients and it worked fairly well. I was, however, very concerned that the anterior leaflet did not open as much at the end of the repair and that I had made these valves somewhat stenotic. Indeed, postoperative echocardiography and pulsed Doppler revealed that the repaired portion of the anterior leaflet had limited excursion, and these patients had a mean diastolic gradient of about 5 mm Hg. Since 1985, we have treated 6 patients with flail anterior leaflet by replacing the ruptured chordae tendineae with Gore- Tex sutures. A 5-0 Gore-Tex suture with a needle on each end was passed through the papillary muscle head that anchored the ruptured chordae tendineae and then tied down. The two arms of the suture were then brought up to the edge of the anterior leaflet and passed through it precisely at the point where the native ruptured chordae was attached. The two arms of the suture were then passed again through the leaflet and the two ends tied down. These patients have been followed up from 3 to 28 months and all repaired valves remain competent and nonstenotic. I have two questions for the authors: have you used synthetic or biomaterials to replace chordae tendineae? Have you examined the repaired valves by echocardiography and pulsed Doppler, and if so, are the valves not stenotic? I enjoyed your paper very much. DR. LESSANA: I thank the discussants for their stimulating comments. Dr. David, we have not seen any mitral stenosis at Doppler echocardiography or at angiography. Dr. Metras, I agree that, ideally, a prosthetic ring should not be used in children; however, in this series, all the children required such a device. Only one had a 26-mm ring. The others had 28-mm rings or above. A 28-mm ring can be used in an adult with a medium body surface area. My experience with not using prosthetic rings in children comes from patients with posterior valve prolapse who had quadrangular resection. Of 4 patients, 1 had to undergo reoperation for rupture of the :suture line on the mitral annulus, and the others had systolic murmurs. Actually, my tendency is not to use any prosthetic ring under 28-mm, and in some rare instances 26 mm. In any event, I would like to stress the absolute necessity of reinhorcing the annulus on its posterior side with Gore-Tex, Dacron, or something else. Dr. Frater, I congratulate you for your technique of replacing chordae, the difficulty of which is assessing the exact length of the chordae. From this standpoint, our technique is more easily reproducible and the natural length of the posterior chordae enables them to ensure adequate motion without prolapse or retraction. Dr. Colvin, I do not have experience with resuspension of the secondary chordae. I did many chordae shortenings that worked well in isolated elongations, but again the major drawback of this technique is the difficulty of precisely assessing the length of the chordae and the extent of the shortening. So, it seems easier to make those transpositions, because the length is naturally the good one.