The risk-benefit ratio of mitral valve operation is

Transcription

1 Degenerative Mitral Regurgitation: When Should We Operate? Malcolm J. R. Dalrymple-Hay, PhD, Mark Bryant, Richard A. Jones, MRCP, Stephen M. Langley, FRCS, Steven A. Livesey, FRCS, and James L. Monro, FRCS Wessex Cardiothoracic Centre, Southampton General Hospital, Southampton, United Kingdom Background. Left untreated, severe mitral regurgitation in asymptomatic patients can lead to irreversible cardiac damage, which can develop with little warning. Over the period of this study, we have tended to operate earlier in the disease process and on less symptomatic patients. We report here our experience. Methods. Between January 1985 and June 1996, 710 patients with mitral regurgitation underwent operations. Three hundred twenty-nine (213 male and 116 female with a mean age of 65.5 years) had degenerative mitral valve disease and of this group, 169 patients underwent repair and 160, replacement. Results. The overall operative mortality was 4 patients (1.2%). There were no operative deaths among patients having isolated mitral valve repair. Survival at 1 year, 5 years, and 10 years was 94% 1.4% ( the standard error of the mean), 77% 2.9%, and 41% 5.8%, respectively. Survival was significantly better in the group having repair (p < 0.05). Ten patients (6%) in the repair group and 13 (8%) in the replacement group required reoperation. Increased age, worse left ventricular function, type of operation (replacement worse than repair), and increased left ventricular size were significantly associated with poorer survival. Conclusions. These data confirm the superior results achieved with mitral valve repair and support early mitral valve repair before functional deterioration. (Ann Thorac Surg 1998;66: ) 1998 by The Society of Thoracic Surgeons The risk-benefit ratio of mitral valve operation is continually changing. Several series [1 7] have indicated a number of factors that increase operative risk. These include older age, higher preoperative New York Heart Association (NYHA) functional class, associated coronary artery disease, impaired left ventricular (LV) function, and elevated pulmonary artery pressure. There is now general agreement that surgical intervention should be performed before LV dysfunction occurs, but reliable detection of early LV dysfunction is difficult. New York Heart Association status, ejection fraction, LV enddiastolic and end-systolic dimensions, rate of rise of LV pressure, left atrial size, and pulmonary artery pressure are all potential indicators, but with the afterload reduction of the regurgitant mitral valve, it is possible to have LV dysfunction in the presence of normal LV dimensions and minimal symptoms. Mitral valve repair has been shown to confer considerable advantages improved operative mortality, cost of in-hospital stay, survival, and morbidity over mitral valve replacement [3, 8 10]. To compare mitral valve repair with replacement and to assess the trend toward early operation, we reviewed our experience over the last decade. Presented at the Thirty-fourth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 26 28, Address reprint requests to Mr Dalrymple-Hay, Department of Cardiothoracic Surgery, Southampton General Hospital, Tremona Rd, Southampton SO166YD, UK ( MDH@btInternet.com). Material and Methods Between January 1985 and June 1996, 710 consecutive patients underwent mitral valve operation for mitral regurgitation (MR) in our institution. Three hundred twenty-nine of them had degenerative MR and form the basis of this report. There were 213 male and 116 female patients with a mean age of years ( standard error of the mean) (range, 16 to 84 years). Information was collected from clinical notes, operative findings, and preoperative and postoperative studies including echocardiographic and right and left heart catherization data. Preoperatively, 28 patients were in NYHA functional class I, 118 in class II, 131 in class III, and 56 in class IV. On echocardiographic criteria, 297 had severe MR and 32, moderate MR, and we were unable to determine the condition of the remainder. Operative Procedure The patients were operated on by five different consultant surgeons. Of the 329 patients, 169 underwent repair and 160, replacement; 75 had concomitant cardiac procedures (Table 1). During the study, there were changes in surgical management. The trend to valve repair increased as the efficacy of the procedure became apparent (Fig 1). Repair techniques also changed. In the early part of the series, the most common technique was quadrangular resection with suture annuloplasty alone. In the latter part of the experience, an annuloplasty ring was used routinely in the repair as the treatment of choice. Sliding leaflet plasty 1998 by The Society of Thoracic Surgeons /98/$19.00 Published by Elsevier Science Inc PII S (98)

2 1580 DALRYMPLE-HAY ET AL Ann Thorac Surg DEGENERATIVE MITRAL REGURGITATION 1998;66: Table 1. Concomitant Procedures Table 2. Procedure and Operative Mortality Concomitant Procedure Repair Principal Procedure Concomitant Procedure No. of Patients Operative Mortality CABG Tricuspid annuloplasty 4 3 Closure PFO or ASD 2 2 AVR 1 2 AVR CABG 1 1 Aortic valvuloplasty 2 0 Closure PFO or ASD tricuspid 1 1 annuloplasty Ascending aorta replacement 2 0 Aortic root CABG 1 0 ASD atrial septal defect; AVR aortic valve replacement; CABG coronary artery bypass grafting; PFO patent foramen ovale. is now used in the majority of posterior leaflet reconstructions. Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) chordae and chordal transposition have been used in anterior leaflet reconstruction. When replacement was undertaken, a mechanical prosthesis was implanted unless a specific contraindication was present. Operative Details The morphology of the abnormal valves is shown in Table 3. Of the 169 patients undergoing repair, 108 had leaflet repair with suture annuloplasty alone, and 61 had injection of an annuloplasty ring. Of the 160 patients having replacement, 127 received a mechanical prosthesis (63 Björk-Shiley, 53 CarboMedics, 6 Starr-Edwards, and 5 St. Jude Medical) and 33, a porcine prosthesis (6 Carpentier-Edwards, 1 Medtronic Intact, and 26 Wessex). Mean aortic cross-clamp and cardiopulmonary bypass times were minutes ( standard error of the mean) and minutes, respectively. Repair No Repair Yes 44 2 No Yes 34 1 Total (1.2%) Statistical Analysis Statistical analysis was performed according to standard statistical protocols incorporated in the SAS statistical package JMP (SAS Institute, Cary, NC). Risk factors for operative mortality were analyzed using logistic regression. Operative death was defined as death within 30 days or before hospital discharge. Survival was analyzed using Kaplan-Meier curves. Survival between groups was compared using log-rank and Wilcoxon tests. Cox s proportional hazards method was used to relate variables to survival. Results There were no significant differences in preoperative cardiac and comorbid status between the two groups. Operative Mortality The overall operative mortality was 4 patients (1.2%). There were no operative deaths among patients having isolated mitral valve repair (Table 2). The average age of the 4 patients who died was 77.8 years. A 78-year-old man underwent mitral valve repair and tricuspid annuloplasty, sustained a perioperative cerebrovascular accident, and eventually died of bronchopneumonia on the ward. A 79-year-old woman with preoperative renal impairment underwent mitral valve repair and closure of an atrial septal defect. After the procedure, she had persistent low cardiac output and died of multiorgan failure 24 days after the procedure. The atrioventricular groove was disrupted in 1 patient who underwent replacement. Despite attempts at repair, it was not possible to discontinue cardiopulmonary bypass. An 84-year-old man underwent emergency operation. He had a history of pulmonary emboli and had severe cardiac failure; he died of multiorgan failure after the procedure. Because of the low number of operative deaths, no Table 3. Morphology of Abnormal Valves at Repair and at Leaflet Prolapse Repair Fig 1. Number of patients having repair or replacement in past 12 years. Posterior Anterior Anterior posterior 20 32

3 Ann Thorac Surg DALRYMPLE-HAY ET AL 1998;66: DEGENERATIVE MITRAL REGURGITATION 1581 Table 4. Valve-Related Morbidity Complication Repair CVA 6 4 TIA 5 5 Endocarditis 3 4 Anticoagulation-related hemorrhage 5 6 Peripheral embolism 0 1 Peripheral embolism TIA 1 0 Anticoagulation-related hemorrhage 1 0 TIA Hemolysis 2 0 Paravalvular leak 0 12 CVA cerebrovascular accident; TIA transient ischemic attack. preoperative or operative variables were significantly associated with an increased risk of operative mortality. Valve-Related Complications and Reoperation-Free Survival Valve-related complications are summarized in Table 4. Ten patients (5.9%) who initially underwent repair and 13 patients (8.1%) who underwent primary replacement had a second mitral valve procedure. The reasons for the second procedure in the repair group included breakdown of the repair in 2 patients, endocarditis in 3, new area of valve prolapse in 4, and hemolysis caused by a regurgitant jet [11] in 1 patient (Table 5). The reasons for Table 5. Initial Operation and Reason for Reoperation in Repair Initial Operation Quad. Res. P2 SA and annuloplasty ring Gore-Tex chords to anterior leaflet and annuloplasty ring chordal shortening to anterior leaflet Reason for Reoperation Hemolysis secondary to regurgitant jet hitting Teflon pledget Prolapse further section posterior leaflet; repair intact New anterior leaflet prolapse Endocarditis; repair intact Prolapse further section posterior leaflet, new chordae ruptured; repair was intact Regurgitation through hole in posterior leaflet adjacent to repair after endocarditis; repair intact Posterior leaflet prolapse; Gore- Tex repair intact Repair sutures torn out of leaflet Endocarditis-damaged valve; repair intact Anterior repair failed; posterior repair intact PA posterior annuloplasty; P2 central component of posterior leaflet; Quad. Res. quadrangular resection; SA sliding leaflet plasty. Fig 2. Freedom from reoperation after surgical intervention for degenerative mitral regurgitation. the second procedure in the replacement group were paravalvular leak in 7 patients (five prostheses were inserted with a continuous suture), prosthetic valve endocarditis in 4, and xenograft failure in 2. Late valve-related mortality occurred in 4 patients in the replacement group and 2 in the repair group. In the former group, 2 died as a result of a cerebrovascular accident, 1 died secondary to a pulmonary embolism, and 1 with prosthetic valve endocarditis died. In the repair group, 1 patient died after a cerebrovascular accident, and 1 died after reoperation and replacement necessitated by failure of the primary repair as a result of hemolysis. Freedom from reoperation was 95.8% 1.1% ( standard error of the mean) (n 314) at 1 year, 92.6% 1.6% (n 156) at 5 years, and 91.9% 1.7% (n 44) at 10 years postoperatively (Fig 2). There was no difference in the reoperation rate between the two groups (log-rank test, p 0.82, and Wilcoxon test, p 0.91). Survival Survival was significantly better in patients who underwent repair compared with those who had replacement (log-rank test, p 0.005, and Wilcoxon test, p 0.003) (Fig 3). Five-year survival was 84.4% 0.63% ( standard error of the mean) for patients in NYHA functional classes I and II who underwent repair (Fig 4A). There was no significant benefit for repair as opposed to replacement for patients in functional class I or II (log-rank test, p 0.21, and Wilcoxon test, p 0.30), but the difference was significant for patients in NYHA functional class III or IV (log-rank test, p 0.01, and Wilcoxon test, p 0.01) (Fig 4B). For all patients, factors significantly associated with worse survival (Cox s proportional hazards method) included increased age ( p 0.003), poorer preoperative

4 1582 DALRYMPLE-HAY ET AL Ann Thorac Surg DEGENERATIVE MITRAL REGURGITATION 1998;66: Fig 3. Survival after operation for degenerative mitral regurgitation. LV function ( p ), type of operation ( p ) (replacement worse than repair), and increased preoperative LV size ( p 0.04). they undergo exercise tests; they have merely adapted to their limitation. Fourth, improvements in medical treatment continually change the natural history of the treated disorder. The data avialble suggest that the annual mortality rate for patients in NYHA functional class I or II with degenerative MR is 4% and that 30% of these deaths are not preceded by any class III or IV symptoms. The 10-year survival rate is only 69% for patients in NYHA functional class I with an ejection fraction greater than 0.60 [17]. This suggests that early surgical intervention has a role and indeed results in lower cardiovascular morbidity, heart failure, and atrial fibrillation and in increased survival. The benefit of early operation remained when patients from only functional classes I and II were compared [13]. Our experience with patients in NYHA functional classes I and II is short. However, to date, results support the position that early surgical intervention should be undertaken in the patient with degenerative mitral valve disease when the valve is amenable to repair. Comment Accepting the difficulties and pitfalls of a retrospective analysis, we found our series clearly demonstrates that mitral valve repair can be performed with a similar (if not lower) operative mortality to mitral valve replacement and that repair confers a significant survival advantage. Several other centers [12, 13] reported similar operative mortalities. Our freedom from reoperation is similar to that in other series and shows that mitral valve repair is a reproducible and durable operation. The incidence of valve-related complications after repair is low. These promising results with mitral valve repair have led us to operate on patients with few symptoms but moderate or severe MR. However, is this approach justifiable? The appropriateness of early surgical intervention depends on a number of factors, including the ability to predict which valves are suitable for repair and the natural history of medically treated MR caused by degenerative mitral valve disease. The challenge of predicting which valves are repairable has been met by transesophageal echocardiography. In experienced hands, it is excellent at defining mitral valve morphology [14] and is the gold standard in the preoperative identification of the mitral valve abnormality. The natural history of MR treated medically is incomplete, and data are now historical [15, 16]. Estimates of survival rates range from 95% at 20 years to 27% at 5 years. The discrepancies are due to several reasons. First, we rarely know the time the valve becomes regurgitant. The starting point in any series is the first time the patient is seen, and this is rarely the onset of MR. Second, series are small and exhibit selection bias and variable degrees of regurgitation. Third, patients who claim to be asymptomatic are frequently functionally limited when Fig 4. (A) Survival of patients in New York Heart Association functional classes I and II after repair versus replacement for degenerative mitral regurgitation. (B) Survival of patients in New York Heart Association functional classes III and IV after repair versus replacement for degenerative mitral regurgitation.

5 Ann Thorac Surg DALRYMPLE-HAY ET AL 1998;66: DEGENERATIVE MITRAL REGURGITATION 1583 In addition, there is an association between mitral valve prolapse and ventricular arrhythmias. It is controversial whether the presence of ventricular arrhythmias is a predictor of sudden death in these patients [18] or whether surgical correction of the flail leaflet decreases this risk [19]. Where doubt exists, it is probably safer to correct the regurgitation. The pathophysiology of the degenerative process should be viewed as a continuum. Initially a limited area of the valve chordae is abnormal; this alters the loadbearing mechanism on the valve and leads to further chordal damage. Thus, a continual destructive cycle is initiated. The earlier this cycle is interrupted, the greater the longevity of the repair. The data on which these conclusions are drawn include our early experience in mitral valve repair. The fact that surgical results have improved in recent years would, if anything, increase the advantage conferred by early operation. In contrast, there is little evidence that medical treatment has changed the natural history of this condition. In conclusion, Carpentier stated at the annual meeting of The American Association for Thoracic Surgery in 1996 that two conditions had to be met before early surgical intervention could be supported: we must be able to guarantee that we will repair the valve rather than replace it and we must have a near zero operative mortality [12]. A recent editorial in The New England Journal of Medicine stated: when there is a record of low operative mortality by a surgical team experienced in valve repair then surgical treatment at the time of diagnosis in patients in NYHA functional class I and II could be appropriate [20]. Several series [8, 12, 21] from a number of centers report experiences similar to ours. We believe that evidence is now accumulating to support operation in patients in NYHA functional class I or II with moderate or severe MR caused by degenerative mitral valve disease. References 1. Borow K, Green L, Mann T. End-systolic volume as a predictor of postoperative left ventricular performance in volume overload from valvular regurgitation. Am J Med 1980;68: Crawford M, Souchek J, Oprian C. Determinants of survival and left ventricular performance after mitral valve replacement. Circulation 1990;81: Galloway AC, Colvin SB, Baumann FG, et al. A comparison of mitral valve reconstruction with mitral valve replacement: intermediate-term results. Ann Thorac Surg 1989;47: Michel P, Iung B, Blanchard B, et al. Long-term results of mitral valve repair for non-ischaemic mitral regurgitation. Eur Heart J 1991;12(Suppl B): Van Herwerden L, Tjan D, Tjissen J, et al. Determinants of survival after surgery for mitral valve regurgitation in patients with and without coronary artery disease. Eur J Cardiothorac Surg 1990;4: Salomon NW, Stinson EB, Griepp RB, Shumway NE. Patient-related risk factors as predictors of results following isolated mitral valve replacement. Ann Thorac Surg 1977;24: Starling M. Effect of valve surgery on left ventricular contractile function in patients with long-term mitral regurgitation. Circulation 1995;92: Enriquez-Sarano M, Schaff H, Orszulak T, et al. Valve repair improves the outcome of surgery for mitral regurgitation. Circulation 1995;91: Cohn LH, Kowalker W, Bhatia S, et al. Comparative morbidity of mitral valve repair versus replacement for mitral regurgitation with and without coronary artery disease. Ann Thorac Surg 1988;45: Perier P, Deloche A, Chauvaud S, et al. Comparative evaluation of mitral valve repair and replacement with Starr, Bjork, and porcine valve prostheses. Circulation 1984;70 (Suppl 1): Dilip K, Vachaspathi P, Clarke P, et al. Haemolysis following mitral valve repair. J Cardiovasc Surg 1992;33: Sousa Uva M, Dreyfus G, Rescigno G, et al. Surgical treatment of asymptomatic and mildly symptomatic mitral regurgitation. J Thorac Cardiovasc Surg 1996;112: Ling L, Enriquez-Sarano M, Seward J, et al. Early surgery in patients with mitral regurgitation due to flail leaflets. Circulation 1997;96: Kay G, Aoki A, Zubiate P, et al. Probability of valve repair for pure mitral regurgitation. J Thorac Cardiovasc Surg 1994; 108: Delahaye J, Gare J, Viguier E, et al. Natural history of severe mitral regurgitation. Eur Heart J 1991;12(Suppl B): Hammermeister K, Fisher L, Kennedy W, et al. Prediction of late survival in patients with mitral valve disease from clinical, hemodynamic and quantative angiographic variables. Circulation 1978;57: Ling L, Enriquez-Sarano M, Seward J, et al. Clinical outcome of mitral regurgitation due to a flail leaflet. N Engl J Med 1996;335: Farb A, Tang A, Atkinson J, et al. Comparison of findings in patients with mitral valve prolapse who die suddenly to those who have congestive heart failure from mitral regurgitation and to those with fatal noncardiac conditions. Am J Cardiol 1992;70: Vohra J, Sathe S, Warren R, et al. Malignant ventricular arrhythmias in patients with mitral valve prolapse and mild mitral regurgitation. PACE 1993;16: Ross J Jr. The timing of surgery for severe mitral regurgitation [Editorial]. N Engl J Med 1996;335(19): Perier P, Stumpf J, Götz C, et al. Valve repair for mitral regurgitation caused by isolated prolapse of the posterior leaflet. Ann Thorac Surg 1997;64: DISCUSSION DR KEVIN D. ACCOLA (Orlando, FL): I enjoyed your very well thought-out paper and presentation. In our practice, Dr Meredith Scott has done most of our valve repairs. On the basis of our extensive experience, we do not think that it is controversial to proceed with valve repair in patients with severe mitral regurgitation, especially in those who have an anatomic dysfunction of the valve despite few to no symptoms and normal ventricular function. We think they do very well. We see a lot of elderly patients with insufficiency. We attempt repair in all patients who have any type of insufficiency with mitral valve disease as an isolated event. What is your philosophy in regard to valve repair for the elderly population? I noticed that two of your patients who died were elderly. Do you still attempt to repair valves