Late Left Ventricular Function After Surgery for Children With Chronic Symptomatic Mitral Regurgitation
Background The use of quantitative echocardiography has been emphasized in optimizing timing of surgery in adult patients with mitral regurgitation to avoid irreversible left ventricular dysfunction. In contrast, surgery for infants and children is often delayed until the appearance of severe symptoms because of the patient’s size and anticoagulation requirements and the possible need for early reoperation. The purpose of this study was to determine long-term ventricular function after mitral valve surgery in symptomatic children and to analyze risk factors for adverse outcome.
Methods and Results Thirty-three patients (0.5 to 19 years old) operated on for mitral regurgitation as a single hemodynamically significant lesion were studied. All but 3 had medically refractory symptoms. One patient died during surgery, and 32 were followed for 0.3 to 17.1 years (mean, 4.5 years). The mean preoperative left ventricular shortening fraction was 0.38±0.09. Successful mitral valvuloplasty or replacement was documented by long-term normalization of end-diastolic dimensions. Early postoperative shortening fraction was significantly reduced (0.28±0.1, P<.01), but it improved to 0.40±0.07 (P<.01) on late follow-up, at which time only 1 patient had ventricular dysfunction. Preoperative shortening fractions did not correlate well with early or late postoperative values (r=.18 and r=.31, respectively). Seven of 32 surviving patients had preoperative shortening fractions <0.33 (mean, 0.26±0.05) and 25 >0.33 (mean, 0.39±0.08). Analysis of these subgroups showed no significant differences between the groups in early or late postoperative function. Duration of mitral insufficiency appeared to be associated with the development of atrial arrhythmias.
Conclusions Late left ventricular function normalizes in children after surgical correction of mitral insufficiency. In contrast to adults, delay of surgery in children with significant mitral regurgitation until the onset of severe symptoms does not increase the risk for long-term ventricular dysfunction, although late atrial arrhythmias are more likely to be encountered.
Early surgical intervention to correct mitral insufficiency before the onset of significant symptoms of cardiac failure has been recommended for adults to avoid irreversible left ventricular dysfunction.1–3 To this end, quantitative echocardiographic assessment of left ventricular dimensions and function has been recommended to optimize the timing for mitral valve surgery.4–6 A preoperative shortening fraction (SF) <0.33 (see References 7 and 87 8 ) or end-systolic dimension measurement >26 mm/m2 (see Reference 99 ) have been suggested as specific values indicating the need for surgical intervention.
In contrast, guidelines for treatment of children with chronic mitral regurgitation are less well defined.10 Medical and surgical management related to body size problems have influenced decisions to delay mitral valve surgery in infants and children as long as possible.11–13 Severe symptoms are often the indication for operation,10 and less emphasis is placed on echocardiographic criteria for intervention. However, late ventricular function has not been reported, nor have potential risk factors for poor outcome been analyzed in children after mitral valve repair or replacement.
We reviewed preoperative and early and late postoperative clinical, echocardiographic, and hemodynamic data in children undergoing mitral valve repair or replacement for chronic symptomatic mitral regurgitation to assess outcome.
Preoperative and postoperative echocardiographic studies were analyzed on 33 patients (6 months to 19 years old; mean age, 6.7 years) with severe mitral regurgitation as a single hemodynamically significant lesion who underwent mitral valve repair or replacement at Babies and Children’s Hospital between July 1976 and September 1996. Echocardiographic quantification of left ventricular dimensions and function was first used clinically at this institution in 1976 and was therefore chosen as the initial year of the study. Patients with significant shunts or involvement of other valves and those with left ventricular dysfunction due to causes other than mitral regurgitation were excluded from the study. Twenty-four patients had previous repair of endocardial cushion defects and subsequent mitral regurgitation requiring further surgery after a mean of 3.0±3.3 years from the original open-heart procedure (range, 3 months to 14.3 years). One patient had an initial diagnosis of unbalanced atrioventricular canal with a pulmonary artery band placed in early infancy and was later considered to have an adequate left ventricle for four-chamber repair at 3 years of age. He was operated on for mitral insufficiency 4 years later (patient 12, Table 1⇓). Five patients had rheumatic mitral regurgitation. One patient had isolated mitral regurgitation secondary to excision of mitral valve accessory tissue that had caused subaortic obstruction. One patient had congenital mitral regurgitation, and two children had mitral insufficiency of unconfirmed cause (congenital versus degenerative). The timing of surgical intervention for each patient was at the discretion of the referring cardiologist.
Nineteen patients had mitral valve repair and 14 mitral valve replacement. One patient died during surgery, and 32 were followed for 4 months to 17 years (mean, 4.5±4.8 years). The clinical, hemodynamic, and echocardiographic data from all patients are shown in Table 1⇑.
Patient records were reviewed for preoperative clinical and hemodynamic status. Thirty patients were severely symptomatic. Four patients had refractory congestive heart failure requiring inotropic support and/or intubation and were in an intensive care unit before surgery, and 26 were ambulatory and were referred for surgical repair because of refractory symptoms despite maximal medical therapy. Three patients with severe mitral regurgitation and left atrial and/or ventricular enlargement by echocardiography had mild symptoms controlled by digoxin and/or diuretics.
Duration of mitral regurgitation was calculated in the atrioventricular canal patients, assuming the day of initial atrioventricular canal surgery as the onset of isolated mitral regurgitation. The presence of at least mild to moderate mitral regurgitation in the immediate postoperative period was confirmed by color or pulse Doppler echocardiography in 21 of 24 patients. For the remaining patients in the series, audible mitral regurgitation, usually with echocardiographic confirmation, was considered to be the onset of regurgitation for purposes of analysis.
Preoperative cardiac catheterization had been performed in 23 patients. Parameters analyzed included cardiac index, pulmonary artery pressures, and pulmonary capillary wedge pressures.
On the basis of the change in cardioplegia solution from crystalloid to blood prime, the surgical era was divided into two periods (1976 to 1989 and 1990 to 1996). Nine patients had surgery in the earlier era and 24 in the later. Total cardiopulmonary bypass time and cross-clamp times were obtained for each operation.
Follow-up data were obtained by contacting the family and the physician caring for the patient, as well as from hospital charts. End points for follow-up were cardiac-related death, diagnosis of prosthetic valve stenosis, or the most recent follow-up date.
Echocardiographic Function Analysis
Three echocardiographic studies were analyzed for each patient. These included the latest preoperative study, a median of 8.5 days before surgery (range, 1 day to 5 months); the prehospital discharge study after surgery (median, 7 days before surgery; range, 5 to 21 days); and the most recent available echocardiogram in the survivors, used as the late follow-up study. Echocardiograms were performed on an ATL Mark 500 A, GE Pass II, HP Sonos 500, 1000, or 1500, Accuson 128, or SP 10 echocardiography machine. Two-dimensional and color Doppler images in the subxyphoid, apical, and parasternal views were used to assess the degree of regurgitation and to rule out other significant lesions. Left ventricular function was studied by use of M-mode measurements of the internal dimensions in diastole and systole, measured at the level of the papillary muscles in the parasternal short-axis view.14 Fractional shortening was computed from the left ventricular dimensions. Z scores were used to standardize dimension measurements to patient size by the following formula: Z score=observed−expected mean dimension/SD of mean. Expected mean values for body surface area were obtained from previously published values.14 The degree of mitral regurgitation was qualitatively assessed as mild, moderate, or severe according to the width and area of the regurgitant jet on color Doppler interrogation.15
Patients were considered to have a good outcome if they were asymptomatic, were taking no medication, and had good left ventricular function on follow-up (group 1). Suboptimal outcome was defined as death, persistent left ventricular dysfunction, or presence of atrial arrhythmias (group 2). The potential risk factors analyzed are listed in Table 2⇓.
Group statistics were expressed as mean±SD. Comparisons of indices of systolic function between preoperative and early and late postoperative time periods for all patients were done by ANOVA. Post hoc comparisons were made by Tukey’s procedure. Comparisons between individual subgroups were performed with unpaired t tests. Correlation coefficients were calculated for specific variables by standard methods.
Univariate and multivariate logistic regression analyses were performed on clinical, surgical, and echocardiographic data to determine the potential risk factors for suboptimal outcome.
Residual mitral regurgitation after successful mitral valve surgery on follow-up echocardiograms was graded qualitatively as absent to mild in 29 patients and moderate in 3 patients. To objectively demonstrate the absence of significant residual mitral insufficiency, end-diastolic dimension Z scores were calculated before surgery and at late follow-up (Fig 1⇓). The mean Z score was significantly reduced at late follow-up before surgery, 3.3±2.9 before surgery to 0.7±1.4 (P<.001) after surgery. One patient (patient 12, Table 1⇑) had an initial left ventricular end-diastolic dimension 2 SD below normal despite significant clinical and echocardiographic mitral regurgitation, which returned to normal range after surgical repair. None of the 3 patients with qualitative moderate mitral insufficiency at late follow-up had elevated end-diastolic dimensions.
Late Left Ventricular Function
The mean preoperative SF for the group of 32 surviving patients was 0.38±0.09 (original group, n=33). The early postoperative SF was significantly lower (0.28±0.1, P<.01), and the mean SF at late follow-up was 0.40±0.07, which was significantly higher than the early postoperative study (P<.01) (Fig 2⇓).
Preoperative SF for the group of 32 surviving patients did not predict early or late postoperative values (r=.18 and r=.31). Seven of the 32 patients had preoperative SF <0.33 (mean, 0.26±0.05) versus 25 patients who were >0.33 (mean, 0.37±0.07; P<.001). Follow-up analysis of these subgroups showed no significant differences on early (0.20±0.07 versus 0.27±0.10, P=.07) or late postoperative quantitative assessment (0.36±0.1 versus 0.39±0.08, P=.17) (Fig 3⇓). Regardless of preoperative or early postoperative ventricular performance, there was complete recovery of function on late follow-up in all but 1 of the 32 patients.
Preoperative end-systolic dimension Z scores did not correlate strongly with early or late postoperative left ventricular function (r=.31 and r=.40). Sixteen of 32 survivors had preoperative end-systolic dimension Z scores >2 (mean, 3.3±1.5), and 16 were <2 (mean, 0.2±1.0). Analysis of the subgroups with preoperative Z scores >2 and <2 showed a significant difference in early postoperative SF (0.24±0.09 versus 0.31±0.09, P=.03). Nevertheless, all but 1 patient showed normalization of function on late follow-up, albeit with a statistical difference in final values between the two groups (0.37±0.1 versus 0.43±0.06, P=.03, Fig 4⇓).
The patients were separated into two groups for risk analysis. Twenty-seven patients (82%) were asymptomatic with good left ventricular function and were taking no medication (group 1). Six patients had suboptimal outcomes (group 2). One patient died during surgery, and another had persistent left ventricular dysfunction and atrial fibrillation. Four children had preserved ventricular function but developed atrial arrhythmias requiring long-term antiarrhythmic therapy. The two groups were compared for possible risk factors predicting poor outcome. On multivariate analysis, only duration of mitral regurgitation (P=.01) was found to be a risk factor for poor outcome (Table 2⇑).
Delay of surgery for mitral insufficiency often results in irreversible left ventricular dysfunction in adults. This has prompted a trend toward earlier surgery in adult patients with chronic mitral regurgitation based on echocardiographic assessment.7 In contrast, many pediatric centers continue to delay surgery for children until severe symptoms are present. This approach is a result of the limitations of patient size for valve replacement, the need for anticoagulation management, and the expectation for early reoperation with patient growth.10–13 Although successful repair generally results in significant clinical improvement in children, the concern remains that as with adults, underlying myocardial dysfunction due to prolonged volume overload may be unmasked by the postoperative increase in left ventricular afterload.1,2,4–6 However, the findings of this study indicate that although early dysfunction after surgery is usual after correction of chronic symptomatic mitral regurgitation in children, late ventricular function normalizes. Of the patients in this study, 20% had borderline preoperative left ventricular function and 68% had below-normal function in the early postoperative period. Nevertheless, 31 of 32 patients (97%) returned to normal function at late follow-up. Children in the present study with preoperative end-systolic dimension Z score >2 did have slightly lower SFs on late evaluation by statistical analysis (Fig 4⇑), but there appears to be no clinical impact of this finding, because all but one of the patients were well within the normal range. The findings here are in contrast to studies in adults.1 Schuler et al1 showed persistent decreased SF in adult patients after mitral valve surgery. Fig 5⇓ shows serial SF data for children in this study and the adult data displayed graphically.
The differences between children and adults in late left ventricular performance after mitral surgery may be partially explained by the relatively shorter overall duration of mitral insufficiency, limiting the time that ventricular volume overload must be endured. Recent onset of regurgitation as an independent favorable predictor of recovery of ejection fraction before surgical correction has been well documented in adults.5 Interestingly, pediatric patients with mitral regurgitation appear to become symptomatic earlier than adults; this shortens the time from diagnosis to surgery and results in the likelihood of normalization of ventricular function. Almost all of the patients in this study were severely symptomatic at the time of surgery, in contrast to adult series, in which detection of ventricular dysfunction by echocardiography in minimally symptomatic or asymptomatic patients is more often the indication for surgery.7,8 In a large natural history study of adult patients with chronic mitral regurgitation, 77% of the patients were noted to be in NYHA class III or IV by 4 to 5 years after the onset of regurgitation.16 Of the patients in this study, 31 of 32 (94%) developed severe symptoms even at rest within 3 years after the initial diagnosis of mitral regurgitation. It would seem that mitral regurgitation that develops more insidiously, as seen in older patients, allows for greater accommodation by a compliant left atrium, and symptoms of pulmonary congestion are delayed.17
The greater resilience in left ventricular function in children than in older patients has been noted previously in other contexts and may reflect a greater potential for recovery in younger myocardium.18–22 This is perhaps best illustrated in infancy by the well-documented recovery of left ventricular function from marked dysfunction after successful repair of anomalous left coronary artery from the pulmonary artery.18,19 In addition, animal studies have demonstrated that when myocardial damage is assessed by mechanical parameters such as ejection fraction, younger myocardium shows significantly less vulnerability to ischemia-reperfusion injury, as may be seen during and before cardiac bypass procedures.20–22
Preoperative Risk Analysis
In adults, long-standing mitral regurgitation with progressive left atrial enlargement has been shown to increase the risk for the development of atrial arrhythmias.8,23,24 Persistent or recurrent atrial fibrillation before surgical correction necessitates chronic anticoagulation before valve repair, even without mechanical valve replacement. The association between prolonged duration of mitral regurgitation resulting in significant atrial enlargement and the ultimate development or persistence of atrial arrhythmia before surgery is documented by several studies.23,24 These reports led Gaasch and Aurigemma to conclude in a recent review8 that early surgical correction may be appropriate before development of symptoms, left ventricular functional changes, or substantial left atrial enlargement so that atrial fibrillation and its complications might be avoided. Late atrial arrhythmias were the primary reason for suboptimal outcome classification in 4 of 6 children in the present study. On the basis of the emergence of the duration of regurgitation as the single predictive variable for suboptimal outcome, it is reasonable also to attribute atrial arrhythmias in children to increasing atrial enlargement over time before mitral valve surgery. This creates a potential clinical paradox in that prolonged duration of mitral regurgitation before surgery does not affect late postoperative left ventricular function, but it may place the patient at a higher risk for developing rhythm disturbances. In view of the well-known risks in operating on the mitral valve in children,11–13 the risk-benefit decisions for early versus delayed surgery in pediatric patients may be difficult.
Mitral regurgitation as an isolated hemodynamic lesion is uncommon in children. This most likely accounts for the scarcity of pediatric studies in this area, and although this study provides the only data specifically addressing this issue in children, the sample is relatively small.
Furthermore, to maximize patient numbers, the definition of hemodynamically “isolated” mitral regurgitation was broadened to include many patients with previously repaired atrioventricular canal defects; cases with nonmitral residual hemodynamic abnormalities were excluded. The importance of this limitation is diminished by the fact that ventricular function was preserved during the time when isolated mitral insufficiency was present, before mitral valvuloplasty or replacement. This indicates that no other residual hemodynamic abnormality was present. Also, the late recovery after mitral valve surgery negates the potential importance of residual myocardial dysfunction secondary to the original atrioventricular defect hemodynamics.
Complex load-independent measures of left ventricular function were not included in this retrospective format. The goal of this study was to analyze preoperative risk factors regularly assessed at routine patient evaluations that could be analyzed to predict late ventricular function and outcome after surgical intervention. It is possible that more sophisticated techniques for quantification of ventricular function, such as load-independent methods, might unmask subclinical degrees of myocardial dysfunction beyond what can be recognized by routine qualitative and quantitative echocardiographic methods.
Long-term postoperative left ventricular function normalized after successful surgery for mitral insufficiency in symptomatic children. These findings are in contrast to adult data and suggest that delay of surgery for mitral regurgitation until the onset of severe symptoms does not significantly increase the risk for late postoperative ventricular dysfunction in pediatric patients. There were suboptimal outcomes in a small number of patients, manifested mainly by the development of atrial arrhythmia after prolonged duration of mitral regurgitation before surgery, and this finding potentially could speak for earlier intervention.
Decision making regarding the timing of mitral valve surgery in children continues to be difficult. Delaying intervention, even in symptomatic patients, has the advantage of being able to achieve a successful valvuloplasty or use a larger prosthetic valve at a later time at lower risk, with the expectation of good recovery of ventricular function. However, the relationship of prolonged duration of mitral regurgitation to the risk of atrial arrhythmias must also be a consideration.
- Received May 30, 1997.
- Revision received August 12, 1997.
- Accepted September 7, 1997.
- Copyright © 1997 by American Heart Association
Schuler G, Peterson KL, Johnson A, Francis G, Dennish G, Utley J, Daily PO, Ashburn W, Ross J Jr. Temporal response of left ventricular performance to mitral valve surgery. Circulation. 1979;59:1218–1231.
Hochreiter C, Niles N, Devereux RB, Kligfield P, Borer JS. Mitral regurgitation: relationship of noninvasive descriptors of right and left ventricular performance to clinical and hemodynamic findings and to prognosis in medically and surgically treated patients. Circulation. 1986;73:900–912.
Ross J Jr. Assessment of cardiac function and myocardial contractility. In: Schlant RC, Alexander RW, eds. The Heart: Arteries and Veins. 8th ed. New York, NY: McGraw-Hill; 1994;1:487–502.
Starling MR. Effects of valve surgery on left ventricular contractile function in patients with long-term mitral regurgitation. Circulation. 1995;92:811–818.
Gaasch WH, Aurigemma GP. Is corrective surgery ever indicated in the symptomatic patient with mitral regurgitation? Cardiol Rev. 1994;2:138–144.
Sanyal SK. Long term sequelae of the first attack of acute rheumatic fever in childhood. In: Emmanouilides GC, Riemenschneider TA, Allen HD, Gutgesell HP, eds. Heart Disease in Infants, Children and Adolescents. Baltimore, Md: Williams & Wilkins; 1995:1416–1440.
Borkon AM, Soule L, Reitz BA, Gott VL, Gardner TJ. Five-year follow-up after valve replacement with the St. Jude medical valve in infants and children. Circulation. 1986;74(suppl I):I-110–I-115.
Henry WL, Ware J, Gardin JM, Hepner SI, McKay J, Weiner M. Echocardiographic measurements in normal subjects. Circulation. 1978;57:278–285.
Helmcke F, Nanda NC, Hsiung MC, Soto B, Adey CK, Goyal RG, Gatewood RP Jr. Color Doppler assessment of mitral regurgitation with orthogonal planes. Circulation. 1987;75:175–183.
Rein AJJT, Colan SD, Parness IA, Sanders SP. Regional and global left ventricular function in infants with anomalous origin of the left coronary artery from the pulmonary trunk: preoperative and postoperative assessment. Circulation. 1987;75:115–123.
Baker EJ, Boerboom LE, Olinger GN, Baker JE. Tolerance of the developing heart to ischemia: impact of hypoxemia from birth. Am J Physiol. 1995;268:H1165–H1173.
Reed D, Abbott RD, Smucker ML, Kaul S. Prediction of outcome after mitral valve replacement in patients with symptomatic chronic mitral regurgitation: the importance of left atrial size. Circulation. 1991;84:23–34.