Anterior Mitral Leaflet Prolapse as a Primary Cause of Pure Rheumatic Mitral Insufficiency

Transcription

1 Anterior Mitral Leaflet Prolapse as a Primary Cause of Pure Rheumatic Mitral Insufficiency Afksendiyos Kalangos, MD, PhD, Maurice Beghetti, MD, Dominique Vala, MD, Edgar Jaeggi, MD, Gürkan Kaya, MD, Vildan Karpuz, MD, Nicolas Murith, MD, and Bernard Faidutti, MD Clinic for Cardiovascular Surgery, Clinic for Pediatric Cardiology, and Department of Pathology, University Cantonal Hospital of Geneva, Geneva, Switzerland Background. This study was designed to revise the mechanisms and repair techniques of anterior mitral leaflet prolapse observed during the correction of pure rheumatic mitral regurgitation in children. Methods. From March 1993 to May 1998, 36 children suffering from pure rheumatic mitral regurgitation due to anterior leaflet prolapse underwent mitral valve repair. The mean age was 12.5 years (range, 6 to 16 years). Anterior leaflet prolapse was due to chordal elongation in 25 patients (group A), chordal rupture in 6 patients (group B), and retraction of anterior secondary chordae tendineae, creating a V-shaped deformity in the middle of the anterior leaflet, thus moving the free edge of the anterior leaflet away from the coaptation plane, in 5 patients (group C). Chordal shortening, transposition, and resection of anterior secondary chordae tendineae were used to correct anterior leaflet prolapse according to the predominantly responsible mechanism. Results. All patients were available for clinical followup, which ranged from 6 months to 5 years (mean follow-up, 3 years). Echocardiographic studies were obtained until the 3rd postoperative month, and all patients showed significant improvement in their left ventricular and atrial dimensions. There was one late death related to endocarditis. Two patients in group C who had mitral valve repair underwent mitral valve replacement on the 19th and 24th postoperative months, respectively, because of failure of mitral valve repair. Conclusions. Mitral valve repair for pure mitral regurgitation due to rheumatic anterior leaflet prolapse can be performed safely for all types of mechanisms. Although the techniques we used provide stable short-term results in each of these groups, midterm results are better in groups A and B, where tissue thickening is less important, recurrences of rheumatic carditis are lower, and the interval between the first rheumatic attack and the surgical procedure is shorter than in group C. (Ann Thorac Surg 2000;69:755 61) 2000 by The Society of Thoracic Surgeons In patients with rheumatic carditis, the mitral valve (MV) is the primary focus of endocardial involvement, and mitral regurgitation (MR) is the most common finding on echocardiographic examination [1]. Mitral valve leaflet prolapse due to elongated chordae or chordal rupture has been described previously as a cause of rheumatic MR. The well-established techniques of MV repair offer the satisfactory correction of rheumatic MR, although mid- and long-term performance of MV repair is less stable than in degenerative lesions [2 4]. Our retrospective study comprises a detailed evaluation of anterior leaflet prolapse mechanisms in pure rheumatic MR assessed by direct examination during the surgical procedure, outlines the techniques we use to correct them, and reports the midterm results of MV repair according to the responsible mechanism of anterior leaflet prolapse. Accepted for publication Aug 9, Address reprint requests to Dr Kalangos, Clinic for Cardiovascular Surgery, University Cantonal Hospital of Geneva, 24, rue Micheli-du- Crest 1211 Geneva 14, Switzerland; afksendyios.kalangos@ hcuge.ch. Material and Methods Patients Between March 1993 and May 1998, 98 children with a mean age of years (range, 6 to 16 years) underwent MV repair by the same surgeon (A.K.) for isolated rheumatic MR at our institution. All of these children were from African countries (Morocco, Senegal, Madagascar, Guinea, Mauritania) and had a welldocumented history of rheumatic fever determined by the revised Jones criteria [5]. The diagnosis of MV disease was established by preoperative transthoracic two-dimensional (TTE) and Doppler echocardiograms in the parasternal long-axis and apical four-chamber views. In all patients, indications for operation included the presence of grade III or IV MR resulting in progressive increase in left ventricular dimensions. In addition, all patients had intraoperative transesophageal echocardiography (TEE) before skin incision and after termination of cardiopulmonary bypass (CPB). Systematic examination of all cardiac valves was performed with particular regard to the annulus, thickness and mobility of leaflets and chordae tendineae, areas and mechanisms of prolapse, and direction of the regurgitant jet. Of these by The Society of Thoracic Surgeons /00/$20.00 Published by Elsevier Science Inc PII S (99)01396-X

2 756 KALANGOS ET AL Ann Thorac Surg ANTERIOR MITRAL LEAFLET PROLAPSE 2000;69: Fig 1. (A) Four-chamber view showing a severely dilated left atrium and prolapse of the anterior mitral leaflet. (B) Four-chamber view showing pseudoprolapse of the anterior mitral leaflet. Note that the posterior leaflet is immobile and the free edge of the anterior leaflet remains at the mitral annular plane and does not override it during systole. (C) Anterior leaflet pseudoprolapse in the presence of a retracted immobile posterior leaflet. patients, 38 had predominant MR with a varying component of mitral stenosis and 60 had pure MR. Mitral regurgitation was considered to be pure in cases associated with unrestricted leaflet excursion and a normal mitral orifice area, whereas its combination with commissural fusion requiring limited commissurotomy was considered to be a predominantly regurgitant lesion. Of the 60 patients with pure MR, echocardiographic examination demonstrated anterior leaflet prolapse in 36 patients (Fig 1A). The two criteria required for the diagnosis of anterior leaflet prolapse on two-dimensional echocardiography were (1) systolic displacement of the coaptation point of the anterior leaflet above the mitral annular plane in the parasternal or apical long-axis views and (2) similar displacement in the apical four-chamber view. Of the remaining 24 patients with pure MR, 19 had echocardiographic findings consistent with pseudoprolapse of the anterior leaflet in the presence of annulus dilatation. A severely retracted immobile posterior leaflet, usually in the presence of annulus dilatation and a more mobile anterior leaflet, can give the impression, particularly in the apical four-chamber view, of anterior leaflet prolapse during systole. Contrary to true anterior leaflet prolapse, the free edge of the anterior leaflet remains in the mitral annular plane during systole and does not override it (Figs 1B, 1C). In the remaining 5 patients, pure MR was due to severe annular dilatation with mild to moderately thickened but mobile anterior and posterior leaflets whose free edges remained in the same coaptation plane on echocardiographic examination despite the lack of

3 Ann Thorac Surg KALANGOS ET AL 2000;69: ANTERIOR MITRAL LEAFLET PROLAPSE 757 Fig 2. (A) Anterior leaflet prolapse due to primary chordal elongation. (B) Anterior leaflet prolapse due to ruptured primary chordae tendineae. (C) Anterior leaflet prolapse due to the retraction of secondary chordae tendineae causing V-shaped deformity in the middle of the anterior leaflet. coaptation between them. Among the 36 patients with true anterior leaflet prolapse, 19 were boys and 17 were girls. Their ages ranged from 6 to 16 years (mean age, years). The mean age of the patients was years when acute rheumatic fever was diagnosed. Five patients had severe tricuspid insufficiency as an associated valvular lesion, and in all of these cases MV repair was accompanied by De Vega annuloplasty of the tricuspid valve. Twenty-three patients were in New York Heart Association (NYHA) class III and 13 were in class IV. All patients had sinus rhythm. None of the children had active rheumatic carditis at the time of operation based on clinical examination, serologic criteria, and surgical macroscopic evaluation. No patient had a history of bacterial endocarditis. All patients were on monthly prophylaxis with intramuscular benzathine penicillin and all were on cardiac medication. Surgical Technique Cardiopulmonary bypass was established by ascending aortic and bicaval cannulation. In all patients, myocardial protection consisted of systemic hypothermia to 28 C, topical hypothermia, and antegrade hyperkalemic crystalloid cardioplegia. Mean cross-clamp time was minutes (range, 30 to 80 minutes). In five cases, De Vega annuloplasty was performed on a beating heart after having finished MV repair. The MV was explored through left atriotomy behind the interatrial groove. Each structure of the MV was analyzed systematically. None of the patients had commissural fusion. Anterior leaflet prolapse was confirmed when the free edge of the anterior leaflet pulled toward the left atrium using a curved hook, overriding the plane of the MV orifice. A variety of surgical techniques were then applied according to the mechanism of anterior leaflet prolapse to restore a competent MV. At surgery, three mechanisms of anterior leaflet prolapse were confirmed. Group A: chordal elongation (Fig 2A). In 25 patients, prolapse of the anterior leaflet was caused by elongation of primary chordae tendineae. Group B: ruptured chordae (Fig 2B). In 6 patients, anterior leaflet prolapse was caused by ruptured paramedian primary chordae tendineae. In 3 patients, ruptured chordae tendineae were misinterpreted as vegetations on two-dimensional echocardiographic examination. In 2 patients, segmental analysis of the anterior leaflet showed a mild to moderate degree of elongation in chordae adjacent to the ruptured paramedian one. Group C: retraction of anterior secondary chordae tendineae (Fig 2C). In 5 patients, retraction of anterior secondary chordae tendineae created a flexion deformity in the shape of a V in the middle of the anterior leaflet, hence moving the free edge of the anterior leaflet away from the posterior leaflet during systole. This mechanism may be well recognized by echocardiography (Fig 3). In group A, thickening of the anterior leaflet and its chordae was judged as mild in 12 patients and moderate in 13 patients, and that of the posterior leaflet was judged as moderate in 19 patients and severe in 6 patients. In group B, thickening of the anterior leaflet was judged as mild in all patients and that of the posterior leaflet as moderate in 5 patients and severe in 1 patient. In group C, thickening of both leaflets and their chordae was judged as moderate in 4 patients and severe in 1 patient. In group A, prolapse of the anterior leaflet was localized on the median segment in 15 patients, on the

4 758 KALANGOS ET AL Ann Thorac Surg ANTERIOR MITRAL LEAFLET PROLAPSE 2000;69: Fig 3. Parasternal long-axis view of anterior leaflet prolapse due to retraction of secondary chordae tendineae. Note the prominent hypertrophied fibrotic tissue on the free edge of the anterior leaflet. posterior paracommissural segment in 3 patients, on both median and posterior paracommissural segments in 3 patients, on the anterior paracommissural segment in 3 patients, and along all three segments of the anterior leaflet in 1 patient. In 15 patients, prolapse of the median segment was corrected by transposing one or two secondary chordae tendineae on its free edge close to this segment. Prolapse of the posterior paracommissural segment was corrected by transposition of a secondary chordae on the free edge of this segment in 1 patient and by a shortening plasty, which consisted of invaginating the excess length of the chordae into a trench created in the papillary muscle, in 2 patients. Prolapse of both median and posterior paracommissural segments was corrected by a shortening plasty of both median and posterior paramedian primary chordae of the anterior leaflet in all 4 patients. In 3 patients, prolapse of the anterior paracommissural segment was corrected by a sliding plasty of the anterior papillary muscle from which elongated paracommissural chordae arose. In addition, 20 patients in group A underwent resection of multiple secondary chordae tendineae of the posterior leaflet. In group B, in all cases, rupture of anterior or posterior paramedian chordae of the anterior leaflet was the primary cause of prolapse of the median segment. This was corrected by transposition of close secondary chordae on the free edge of the median segment. In 2 patients who presented a moderate degree of chordal elongation on the adjacent segments, chordal shortening was performed at the cusp level as previously described by Frater [6]. Four patients in this group underwent resection of posterior secondary chordae tendineae. In groups A and B, the degree of chordal shortening or the length of transposed chordae was determined according to the degree of posterior leaflet retraction in order to achieve an adequate coaptation surface situated in a lower plane than that of a normal MV. In group C, prolapse of the median segment of the anterior leaflet was due to shortened retracted secondary chordae creating the flexion deformity. None of these patients had any degree of commissural fusion. In these cases, leaflet thickening was more important than in the other two types, and there was prominent fibrotic tissue, 3 to 4 mm large, along the free edge of the anterior leaflet, leaving the primary chordae tendinea, which were not elongated, in a more internal position. During the surgical procedure, analysis of the lesions showed us that the prolapsing part of the anterior leaflet was in fact this prominent tissue, thereby constituting an intermediate type of prolapse between that of pseudo and true prolapse. The coaptation surface between the anterior and posterior leaflets was improved after resecting the retracted secondary chordae and after shaving the free edge of the anterior leaflet so that primary chordae might regain their appropriate position. Annulus remodeling using a rigid Carpentier-Edwards prosthetic ring was necessary in all patients due to associated annulus deformation and dilatation. The mean prosthetic ring size used in this series was (range, 26 to 34). The adequacy of repair was assessed by injection of saline solution with a bulb syringe into the left ventricle directly through the MV. All patients received oral antivitamin K treatment for 3 months after operation. Anticoagulation at an international normalized ratio of 2 to 2.5 was maintained during this period. Follow-up Patients were followed up in the outpatient clinic during the first 3 postoperative months. In addition to clinical examination, all patients underwent TTE with color Doppler flow mapping at discharge and at 3 months postoperatively before returning to their countries of origin. Doppler echocardiographic grade of MR was measured with color Doppler flow and graded according to the width and length of the regurgitant jet in the left atrium (grades I to IV). Mean gradient across the MV was estimated by measuring peak diastolic velocity from Doppler studies in the four-chamber view. Left ventricular end-diastolic (LVEDD) and end-systolic diameters (LVESD), and left ventricular shortening fraction were measured at midpapillary level in the standardized parasternal transthoracic long-axis or short-axis position. All patients were then followed up in outpatient clinics by cardiologists in their countries of origin, who periodically informed us of the patients evolution by filling out a questionnaire enquiring about clinical, echocardiographic, and medication details. Statistical Analysis Variables are reported as the mean 1 SD and the range. Temporal changes in echocardiographic measurements were compared by Student s t tests. A Kruskal-Wallis test was used for comparisons between groups A, B, and C regarding the mean interval between the diagnosis of

5 Ann Thorac Surg KALANGOS ET AL 2000;69: ANTERIOR MITRAL LEAFLET PROLAPSE 759 acute rheumatic fever and the surgical procedure. Multiple comparisons were then made using Bonferroni corrections. Results Clinical Follow-up Clinical follow-up was complete, ranging from 6 months to 5 years (mean, 3 years). There were no operative and there have been no early postoperative deaths. The mean postoperative hospital stay was 9 2 days. One patient died in the 9th postoperative month of septicemia and multiple organ failure in another hospital with apparently competent aortic and mitral valves. Late follow-up revealed that functional improvement was observed in all patients. Twenty-nine patients are now in NYHA class I and 6 in class II, according to the last follow-up control. Mean Interval Between the Diagnosis of Acute Rheumatic Fever and Surgical Procedure In group A, the mean time interval between the diagnosis of acute rheumatic fever and the time of surgical procedure was months (range, 15 months to 3 years). Sixteen patients had one recurrence of rheumatic carditis during this interval. In group B, this interval was months (range, 15 to 22 months). All patients had one recurrence of rheumatic carditis during this interval. In group C, the interval was months (range, 37 to 66 months). The interval was statistically longer in group C than in the other groups ( p 0.005). There was no statistical difference between groups A and B. All patients had at least three recurrent rheumatic attacks during this interval. Reoperation Two patients who had had MV repair in group C were reoperated due to failure of the MV repair on the 19th and 24th postoperative months, respectively. Both of them had their MV replaced by mechanical prostheses. The first patient developed severe residual MR due to infective endocarditis with Staphylococcus aureus. Multiple ruptured anterior chordae with healed vegetations were noted at operation. The second patient developed predominantly severe mitral stenosis and demonstrated clinical and serologic evidence of recurrence of rheumatic heart disease during the follow-up period. Fusion of the leaflets and chordae at both commissural sites and tissue retraction were noted in the 2 patients at operation. Both of them were in NYHA class I after MV replacement and were doing well up to the last follow-up control. Complications There were few perioperative complications. One patient required mediastinal reexploration for the control of postoperative hemorrhage. There were no wound complications or neurologic events. Postoperative atrial fibrillation, later converted into sinus rhythm, occurred in 2 patients who then continued to remain in sinus rhythm. Fig 4. Left ventricular end-diastolic diameter (LVEDD) preoperatively (Preop) and postoperatively on the 1st week and the 3rd month. *p vs before operation. Residual Mitral Regurgitation and Transvalvular Gradient Twenty-five patients had grade IV MR and 11 patients grade III MR, as detected by TTE before operation. Transesophageal echocardiography upon termination of CPB showed grade 0 MR in 7 patients and grade I MR in 29 patients. At discharge, the number of grades 0 and I by TTE was the same. Aggravation of MR from grade I to grade II was observed in 1 patient on the 3rd postoperative month in whom no further significant change in the degree of residual MR was found thereafter, throughout 36 postoperative months. The mean transmitral gradient was estimated at mm Hg (range, 1.3 to 7 mm Hg) at discharge and mm Hg (range, 1.6 to 7 mm Hg) on the 3rd postoperative month. Left Ventricular and Left Atrial Dimensions and Shortening Fraction Because of the increased preload due to MR, LVEDD and LVESD were enlarged preoperatively with values of cm (range, 4.3 to 8.2 cm) and cm (range, 2 to 6.3 cm), respectively. Mean LVEDD and LVESD dimensions were reduced significantly at discharge with values of cm (range, 3.2 to 7.1 cm; p 0.001) and cm (range, 2.1 to 6.5 cm; p 0.001), respectively (Fig 4). Mean left atrial diameter was mm preoperatively and was reduced significantly at discharge with values of mm (range, 17 to 58 mm; p 0.001) and mm (range, 15 to 55 mm; p 0.001) at 3 months. Twenty-six patients had normal preoperative left ventricular shortening fraction ( 30%) with a mean value of 38% 6.15% (range, 23% to 50%). Left ventricular shortening fraction was reduced significantly at discharge with a mean value of 25% 9.3% (range, 10% to 50%; p 0.001) and then increased to a mean value of 30.5% 6.8% (range, 15% to 50%) at 3 months (Fig 5).

6 760 KALANGOS ET AL Ann Thorac Surg ANTERIOR MITRAL LEAFLET PROLAPSE 2000;69: Fig 5. Shortening fraction (SF) preoperatively (Preop) and postoperatively on the 1st week and the 3rd month. *p vs before operations **p vs before operation and vs 1 week. Comment It is known that rheumatic carditis frequently results in isolated MR. Recent echocardiographic and surgical data have demonstrated a high incidence of MV prolapse accounting for 30% to 97% of cases of rheumatic MR, which suggests that rheumatic fever is an evident cause of MV prolapse [2 4, 7 11]. Selection bias inherent in surgical series, evaluation of patients with active or chronic rheumatic carditis, differences in patients mean age between series, and differing criteria for defining rheumatic carditis are all possible explanations for the wide variation in the incidence of rheumatic MV prolapse [1, 12]. In addition, we believe that differing criteria for anterior leaflet prolapse may lead to its overdiagnosis by considering pseudoprolapse as a veritable one on echocardiographic examination or at operation. Our lower rate of MV prolapse (37%) compared with the other surgical series (82% to 97%) may be because all cases of anterior leaflet pseudoprolapse (32% in our entire surgical series) are meticulously distinguished from true prolapse. Contrary to true anterior leaflet prolapse repair in which chordal shortening or transposition is mandatory, MR due to anterior leaflet pseudoprolapse is easily corrected by the mobilization of the posterior leaflet by resecting its retracted secondary chordae tendineae, shaving its thickened tissue, and reducing the size of the annulus using a prosthetic ring. Chordal elongation seems to be the much more frequent mechanism of anterior leaflet prolapse in recurrent and chronic rheumatic carditis than in the initial episode [1, 8]. Vasan and colleagues [1], in an echocardiographic evaluation of patients with a first attack of rheumatic carditis, reported that the most common causes of MR were left ventricular with annular dilatation and restricted leaflet mobility. This observation is valuable because chordal lengthening and prolapse of the anterior leaflet usually supervene to the posterior annular dilatation caused by active rheumatic carditis. Although chordal elongation detected in the active phase may relate to the rheumatic inflammatory process, the predominant factor over time is certainly the exposure of the chordae to enhanced tensile stress during ventricular systole. Persistent posterior annular dilatation and restriction in motion of the posterior leaflet, more affected by fibrosis than the anterior one, as confirmed in our cases, reduce the coaptation area of leaflets, and hence diminish the keystone effect with a resultant increase in chordal tension. The chordae most exposed to tension are the anterior paramedian primary chordae supporting the median segment; 15 of the 25 cases (60%) in group A had pure anterior median segment prolapse due to the elongation of paramedian chordae. We also observed that in the other cases of chordal elongation concerning the anterior or posterior paracommissural segment of the anterior leaflet, alone or in association with the median segment, the mitral annulus was not only dilated but also deformed, thereby exacerbating the lack of apposition between the segments of the corresponding anterior and posterior leaflets and increasing the chordal tension in favor of the uncoaptating segments. In one case in group A which had extended prolapse of all three segments of the anterior leaflet, the mitral annulus was extremely dilated, thereby equally exposing all anterior primary chordae to an increase in tension. Rupture of the chordae tendineae may occur either as a complication of active rheumatic carditis, which weakens chordae by inflammation, or as a consequence of the increasing tension exerted on elongated chordae over time [13]. Anterior leaflet chordal rupture was detected in between 9% and 23% of cases subjected to operation [2, 3, 11]. Our incidence of 17% of chordal rupture is consistent with these series which also contain cases operated on in the presence of active rheumatic carditis. In our series, patients in group C had a time interval between the diagnosis of acute rheumatic fever and surgical procedure longer than that of the other two groups. In group C, the severity of valve regurgitation was grade IV for each patient, and all patients were in NYHA functional class IV. Scarring of valve leaflets was more important in group C than in the other groups, possibly owing to repeated attacks of valvulitis culminating in greater valvular damage and to the self-perpetuating nature of regurgitant jets. In our series, prolapse of the anterior median segment in groups A and B was easily corrected by transposition of one or two secondary chordae onto the free edge of this segment. In group A, the technique of chordal shortening consisted of burying the extra length of the elongated chordae within a trench created in the corresponding papillary muscle or splitting the tip of the papillary muscle supporting the elongated chordae and resetting it at a lower level. The latter is considered more appropriate to correct chordal elongation of the anterior or posterior paracommissural segment of the anterior leaflet. Chordal shortening at the cusp level, as described previously by Frater [6], was performed in cases that had mild to moderate chordal elongation. According to this technique, the greater the chordal elongation, the closer

7 Ann Thorac Surg KALANGOS ET AL 2000;69: ANTERIOR MITRAL LEAFLET PROLAPSE 761 the stitch through the chorda should be to the papillary muscle. This can, however, weaken the chorda since its point of insertion on the papillary muscle is usually thinner than its portion closer to the free edge of the leaflet. In group C, resection of the retracted anterior secondary chordae tendineae (especially those in the paramedian position) allows billowing of the central part of the anterior leaflet during systole and hence results in better apposition of both mitral leaflets. Moreover, the coaptation surface between both leaflets could also be improved by shaving the prominent hypertrophied fibrotic tissue located along the free edge of the anterior leaflet so that anterior primary chordae might regain their appropriate position. The reoperation rate of 5.5% in our series concerning only rheumatic MR due to anterior leaflet prolapse is similar to those reported in other series concerning different pathologic aspects of rheumatic MV involvement in older patients [2, 3, 11]. The causes of failure of MV repair in 2 patients in group C were severe mitral stenosis due to recurrence in 1 patient and infective endocarditis in the other. In addition, aggravation of MR from grade I to grade II was detected in another patient in group C, who remained asymptomatic and was not considered to require reoperation. All these events give us the impression that patients in group C are more prone to require reoperation with increasing age, because the rheumatic valvular involvement at valvular and subvalvular levels is more important than in the other two types of lesions. In conclusion, repair of anterior leaflet prolapse in pure rheumatic MR, when feasible, is a stable and safe procedure with a low prevalence of reoperation in groups A and B at midterm. Care must be taken to distinguish true rheumatic anterior leaflet prolapse from pseudoprolapse, which does not require chordal repair techniques. Prolapse due to the retraction of anterior secondary chordae tendineae seems to have more complications during follow-up than the other two mechanisms because of the greater evolutive rheumatic valvular involvement. References 1. Vasan RS, Shrivastava S, Vijayakumar M, Narang R, Bradford CL, Narula J. Echocardiographic evaluation of patients with acute rheumatic fever and rheumatic carditis. Circulation 1996;94: Antunes MJ, Magalhaes MP, Colsen PR, Kinsley RH. Valvuloplasty for rheumatic mitral valve disease. J Thorac Cardiovasc Surg 1987;94: Duran CMG, Gometza B, De Vol EB. Valve repair in rheumatic mitral disease. Circulation 1991;84(Suppl III): Skoularigis J, Sinovich V, Joubert G, Sareli P. Evaluation of the long-term results of mitral valve repair in 254 young patients with rheumatic mitral regurgitation. Circulation 1994;90(Part II): Committee on Rheumatic Fever and Bacterial Endocarditis of the American Heart Association. Jones criteria (revised) for guidance in the diagnosis of rheumatic fever. Circulation 1984;69:203A 8A. 6. Frater RWM. Mitral valvuloplasty. In: Roberts AJ, Conti CR, eds. Current surgery of the heart. Philadelphia: Lippincott, 1987: Lembo NJ, Dell Italia LJ, Crawford MH, Miller JF, Richards KL, O Rourke RA. Mitral valve prolapse in patients with prior rheumatic fever. Circulation 1988;77: Marcus RH, Sareli P, Pocock WA, et al. Functional anatomy of severe mitral regurgitation in active rheumatic carditis. Am J Cardiol 1989;63: Zhou L, Lu K. Inflammatory valvular prolapse produced by acute rheumatic carditis: echocardiographic analysis of 66 cases of acute rheumatic carditis. Int J Cardiol 1997;58: Wu M, Lue H, Wang J, Wu J. Implications of mitral valve prolapse in children with rheumatic mitral regurgitation. J Am Coll Cardiol 1994;23: Chauvaud S, Perier P, Touati G, et al. Long-term results of valve repair in children with acquired mitral valve incompetence. Circulation 1986;74(Suppl I): Feinstein AR. On blind men, elephants, spectrums and controversies: lessons from rheumatic fever revisited. J Chron Dis 1986;39: De Moor MMA, Lachman PI, Human DG. Rupture of tendinous chords during acute rheumatic carditis in young children. Int J Cardiol 1986;12:353 7.