- Email: [email protected]
Aortic Valve Replacement With the Pulmonary Autograft: Mid-Term Results
CARDIOVASCULAR
Aortic Valve Replacement With the Pulmonary Autograft: Mid-Term Results Arkalgud Sampath Kumar, MCh, Sachin Talwar, MCh, Raghunath Mohapatra, MS, Anita Saxena, DM, and Rajvir Singh, MS, PhD Cardiothoracic Centre and Department of Biostatistics, All India Institute of Medical Sciences, New Delhi, India
Background. The purpose of this study is to assess the mid-term results of aortic valve replacement with the pulmonary autograft. Methods. From October 1993 through September 2003, 153 patients with aortic valve disease (81 rheumatic and 72 non-rheumatic), with a mean age of 28 ⴞ 14.2 years underwent the Ross procedure with root replacement technique and right ventricular outflow tract reconstruction using a homograft. Associated procedures included mitral valve repair (n ⴝ 19), open mitral commissurotomy (n ⴝ 15), tricuspid valve repair (n ⴝ 2), homograft mitral valve replacement (n ⴝ 2), and subaortic membrane resection (n ⴝ 1). Results. Early mortality was 6.5% (10 patients). Mean follow-up was 77 ⴞ 42 months (range, 7 to 132 months; median, 90 months). One hundred, twenty-one survivors (84.6%) had no significant aortic regurgitation. Reoperation was required in 10 patients for autograft dysfunction alone (n ⴝ 3), infective endocarditis (n ⴝ 2), autograft
dysfunction with failed mitral valve repair (n ⴝ 3), and failed mitral valve repair alone (n ⴝ 2). No reoperations were required for the pulmonary homograft. There were 8 late deaths. Actuarial and reoperation-free survival at 90 months were 91.% ⴞ 3.5%, 95.3% ⴞ 2.7%, in nonrheumatics and 86.1 ⴞ 3.9%, 90.5 ⴞ 3.7% in rheumatics, respectively. Freedom from significant aortic stenosis or regurgitation was 91.5 ⴞ 2.8% in non-rheumatics and 80.6 ⴞ 4.8% in rheumatics. Event-free survival was 86.2 ⴞ 4.9% in non-rheumatics and only 68.9 ⴞ 5.3% in rheumatics. Conclusions. The Ross procedure is not recommended for young patients (< 30 years) with rheumatic heart disease. It provides satisfactory hemodynamic and clinical results in properly selected patients. Important autograft dilatation was not observed in our patients.
D
aortoarteritis. A total of 28 patients (18 non-rheumatics and 10 rheumatics) were 15 years of age or less. In the rheumatic group, 40 patients were 30 years of age or younger (young rheumatics). One hundred, twenty-two patients (79.7%) had dyspnea as the predominant symptom. One hundred fourteen patients (74.5%) were in New York Heart Association functional class III and IV. Preoperative transthoracic echocardiography was performed in all patients. Cardiac catheterization and cine angiography was performed if there was suspicion of associated mitral valve or coronary artery disease. Informed consent was obtained from all patients. Intraoperative transesophageal echocardiography was performed in all patients. If the aortic root diameter was 30 mm or more, aortic valve replacement with a mechanical prosthesis was preferred. Ross procedure was not performed in patients requiring emergency surgery, significant left ventricular dysfunction, mitral valve disease requiring prosthetic valve replacement, significant mismatch between the aortic root diameter and the pulmonary valve, and in patients with bicuspid pulmonary valve or pulmonary regurgitation. In 2 patients with bicuspid aortic valves, the Ross procedure was abandoned after the pulmonary valve was found to be bicuspid at operation. Since 1998, we abandoned this procedure in young rheumatics (ie, those under 30 years of age).
espite the increasing trend toward valve repair, valve replacement with a mechanical valve or bioprosthesis is the most commonly performed operation for patients with aortic valve disease [1]. The pulmonary autograft, first used by Ross [2] in 1967 to replace the diseased aortic valve, has been shown to provide good haemodynamics, ability to remodel, and growth potential [3–9]. We first performed this procedure in October 1993 and reported our early experience in young patients with rheumatic heart disease [10]. In this study, we present detailed results in 153 consecutive patients.
Patients and Methods From October 1993 through September 2003, 153 patients (Table 1) underwent aortic valve replacement with the pulmonary autograft (Ross procedure). The cause of the aortic valve lesion was rheumatic in 81 patients (53%) and non-rheumatic in 72 patients (47%). The latter group included patients with bicuspid aortic valve (n ⫽ 64); degenerative aortic stenosis (n ⫽ 7); and 1 patient with Accepted for publication March 4, 2005. Address reprint requests to Dr Kumar, Department of Cardiothoracic and Vascular Surgery, Cardiothoracic Centre, All India Institute of Medical Sciences, Ansari Nagar, New Delhi, 110 029 India; e-mail: asampath_ [email protected].
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2005;80:488 –94) © 2005 by The Society of Thoracic Surgeons
0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2005.03.023
KUMAR ET AL ROSS PROCEDURE
Table 1. Profile of Patients Undergoing the Ross Procedure Cause Rheumatic Non rheumatic Male:female Age Symptoms Dyspnea Palpitations Chest pain Congestive heart failure New York Heart Association functional class I II III IV Clinical diagnosis Aortic stenosis Aortic regurgitation Aortic stenosis plus aortic regurgitation Associated mitral valve disease Associated tricuspid valve disease
Number
Percentage
81 72 103:50 8 mo to 65 y (mean, 28 ⫾ 14.2 y) 122 61 65 48
79.7 39.9 42.5 31.4
5 34 96 18
3.3 22.2 62.7 11.8
40 66 47
26 43 31
36
25
2
1.3
The surgical procedure has been described in detail earlier by us [5, 10, 11]. Root replacement technique with coronary artery implantation was used in all patients. The left coronary artery was not implanted as a button. Instead, the left coronary ostium was left attached to a tongue of the distal aortic wall. Right ventricle– pulmonary artery continuity was restored using an antibiotic preserved (n ⫽ 47) or cryopreserved homograft (n ⫽ 106) obtained from our own tissue valve bank. A pulmonary homograft was used in 151 patients and aortic homograft in 2 patients. Associated procedures included mitral valve repair (n ⫽ 19), open mitral commissurotomy (n ⫽ 15), tricuspid valve repair (n ⫽ 2), homograft mitral valve replacement (n ⫽ 2), and subaortic membrane resection (n ⫽ 1) using techniques described earlier by us [12]. For the isolated Ross procedure, the mean aortic cross-clamp and cardiopulmonary bypass times were 121.4 ⫾ 19.4 minutes (range, 81 to 197 min) and 153 ⫾ 24.7 minutes (range, 121 to 226 min), respectively. When the mitral valve procedure was done concomitantly, these increased to 135 ⫾ 31 minutes (109 to 230 min) and 168 ⫾ 32 minutes (140 to 280 min), respectively. After weaning from cardiopulmonary bypass, transesophageal echocardiography was performed in all patients to confirm normal autograft function and to assess associated procedures. Prior to discharge from the hospital, transthoracic echocardiography was performed in all patients, and this was repeated every 6 months. Aortic regurgitation (AR) was assessed on a scale of ⫹1 to ⫹4
489
according to published criteria [13]. Aortic regurgitation with a grade of ⫹1 was considered mild. Peak gradients less than 25 mm Hg across the aortic valve were considered as mild aortic stenosis and ⱖ50 mm Hg was considered significant aortic stenosis. No anticoagulants or anti-platelet drugs were prescribed. Long-acting benzathine penicillin 3 weekly was prescribed to all patients less than 45 years of age with rheumatic heart disease. All patients received itraconazole for 6 weeks after surgery as prophylaxis against fungal endocarditis.
Statistical Analysis Statistical analysis was performed using SPSS for Windows 7.5 software package (SPSS Inc, Chicago, IL). The rheumatic and non-rheumatic groups of patients were analyzed separately. Descriptive statistics (ie, mean and standard deviation) were calculated for continuous variables. Freedom from valve-related events and actuarial estimates were calculated using the Kaplan-Meier analysis. The estimates were compared with the Mantel-Cox (log-rank) test. A valve-related event was defined as any episode of thromboembolism, hemorrhage, congestive heart failure, infective endocarditis, structural deterioration, or significant gradients as per published criteria [14]. Linearized rates for these events are reported. Cox’s proportional hazard model was used to analyze factors associated with a higher early and late mortality and the development of significant aortic stenosis or AR.
Results Hospital Mortality All patients survived the operation. Thirty-day mortality was 6.5% (n ⫽ 10) and was due to bleeding (n ⫽ 3), arrhythmias (n ⫽ 3), persistent low cardiac output (n ⫽ 2), chest infection (n ⫽ 1), and fungal endocarditis (n ⫽ 1). All 3 deaths due to bleeding and its sequelae were in the first 12 patients operated on. Subsequently we modified our operative technique to eliminate this problem [15]. There were only 3 early deaths in the last 100 consecutive patients.
Early Reoperation Six patients required reoperation for excessive mediastinal bleeding. Four of these were in the first 12 patients in which the bleeding was from the raw surface of the posterior wall of the right ventricular outflow tract. In the other 2 patients, the source of bleeding was not related to the suture lines. One patient required reoperation within a month for fungal endocarditis of the autograft, and she died after surgery from multiorgan dysfunction.
Early Autograft Function In all patients, intraoperative transesophageal echocardiography revealed trivial or no aortic regurgitation. Three patients had mild mitral regurgitation. Transthoracic echocardiography prior to discharge from the hospital showed mild AR in 1 patient.
CARDIOVASCULAR
Ann Thorac Surg 2005;80:488 –94
490
KUMAR ET AL ROSS PROCEDURE
Ann Thorac Surg 2005;80:488 –94
Table 2. Reoperations After the Ross Procedure
CARDIOVASCULAR
SN
Age/Sex
Cause
Interval
Cause of Reoperation (mo)
Operation
Fate
Cause of Death
1 2 3 4 5 6 7 8 9 10
24/M 22/F 17/M 13/M 20/M 21/M 15/F 30/M 36/F 11/M
RHD BAV RHD RHD RHD RHD RHD BAV RHD BAV
108 3 48 12 48 56 26 4 31 17
RHD, AR, dilated autograft Fungal IE AR, MR AR, MR, TR AR, MR, dilated autograft AR MR Fungal IE * MR Cusp perforation
AVR HAVR DVR DVR plus TV repair DVR AVR HMVR HAVR MVR AVR
Survived Died Survived Survived Survived Survived Died after 3 months Died Survived Survived
NA IE, MSOF NA NA NA NA CHF IE, MSOF NA NA
AR ⫽ aortic regurgitation; AS ⫽ aortic stenosis; AVR ⫽ aortic valve replacement; BAV ⫽ bicuspid aortic valve; CHF ⫽ congestive heart failure; DVR ⫽ double valve replacement; F ⫽ female; HAVR ⫽ homograft aortic valve replacement; HMVR ⫽ homograft mitral valve replacement; IE ⫽ infective endocarditis; M ⫽ male; MR ⫽ mitral regurgitation; MSOF ⫽ multisystem organ failure; MVR ⫽ mitral valve replacement; NA ⫽ not applicable; RHD ⫽ rheumatic heart disease; SN ⫽ serial number; TR ⫽ tricuspid regurgitation.
Follow-Up All patients were seen in the outpatient clinic at 6-month intervals and underwent clinical examination and echocardiography. Between January 2004 and September 2004, the records of 140 of the 143 survivors were obtained, and their last follow-up during this period was taken to report the results. The follow-up data (97.9% complete) ranged from 7 to 132 months (mean, 77 ⫾ 42 months; median, 90 months) and totaled 917.6 patientyears. Among survivors, 20 patients (14%) were followed-up for 10 or more years, 62 (43.4%) were followed-up for 7 or more years, 12 (8.4%) for more than 5 years, 24 (16.8%) for more than 3 years, and 6 (4.2%) for more than 2 years.
Thromboembolism There were no thromboembolic complications in the survivors.
Hemolysis Four patients had significant hemolysis. In the first 3 patients this was associated with excessive mediastinal bleeding, disseminated intravascular coagulation, and early death. In the fourth patient, this was associated with mild mitral regurgitation and subsided gradually. The incidence of this complication in survivors was 0.1 event per 100 patient-years of follow-up.
Infective Endocarditis Five patients had infective endocarditis (0.4 events per 100 patient-years) develop. This was fungal in 3 patients and bacterial in 2. One patient with fungal endocarditis died within 1 month of surgery, the other 2 underwent homograft aortic valve replacement 3 and 4 months after the initial operation and died. The 2 patients with bacterial endocarditis died 9 and 13 months after operation.
Late Autograft Function At last follow-up between January 2004 and September 2004, transthoracic echocardiography showed no aortic regurgitation in 121 patients (84.6%). Nineteen patients
had significant aortic regurgitation (2 events per 100 patient-years). Moderate (n ⫽ 14) to severe (n ⫽ 5) aortic regurgitation was noted in 19 patients (13.3%) of the 143 survivors at a median follow-up of 90 months after operation. Fifteen of these patients had rheumatic heart disease, were young (⬍ 30 years of age), and had significant AR develop 57.6 ⫾ 36.2 months after the operation; five of these patients had evidence of recurrent attacks of rheumatic fever and had significant AR develop within 2 years of operation. Two of these (patients 1 and 5, Table 2) had stopped penicillin prophylaxis and required hospitalization for recurrent attacks of acute rheumatic fever. In the non-rheumatic group, AR had developed 81.9 ⫾ 42.8 months after the operation (log rank test, 1.97; p ⫽ 0.16). On multivariate analysis, the probability of developing of AR was higher in young rheumatics (⬍ 30 years) with an odds ratio of 4.2, and a 95% confidence interval (CI) of 1.6 to 10.4 (p ⫽ 0.0033). All 5 patients with severe AR underwent reoperation, and 2 of these required double valve replacement for failure of associated mitral valve repair. In all these patients, mechanical valves were implanted. One patient with moderate AR and failed mitral valve repair underwent double valve replacement with a mechanical valve 48 months after the initial operation. Patients 1 and 5 (Table 2) were found to have significant autograft dilatation due to recurrent attacks of rheumatic activity. Two patients died from congestive heart failure 24 and 5 months after the initial operation. One patient with moderate AR died of bacterial pneumonia 2 months after operation. The remaining 10 patients with moderate AR continue to be in New York Heart Association functional class I and are being closely followed-up. Two patients with a preoperative diagnosis of bicuspid aortic valve have a gradient of 25 mm Hg and 40 mm Hg 5 years and 7 years, respectively, after the initial operation, but they are asymptomatic. Autograft dilatation was not seen in any other patient. At 90 months, freedom from autograft dysfunction was 84.3 ⫾ 3.6 % (95% CI 77.3 to 91.3) for the entire group; in rheumatics it was 80.6 ⫾ 4.8% (95% CI 71.2 to 90) as compared with 91.5 ⫾ 2.8% (95% CI 86 to 97) in non-
KUMAR ET AL ROSS PROCEDURE
491
CARDIOVASCULAR
Ann Thorac Surg 2005;80:488 –94
Fig 1. Freedom from aortic regurgitation (AR) (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ non-rheumatic.
rheumatic group (log rank test, 1.9; p ⫽ 0.16). At 10 years, it was 79.2 ⫾ 4.9 % (95% CI 69.8 to 88.6) for the entire group; in rheumatics it was 74.7 ⫾ 6.1% (95% CI 62.7 to 86.7) as compared with 91.5 ⫾ 2.8% (95% CI 86 to 97) in the non-rheumatic group (Fig 1).
Fate of Right Ventricular Outflow Tract Four patients had pulmonary homograft dysfunction (0.4 events per 100 patient-years). Three patients had a gradient of 40 mm Hg or more across the right ventricular outflow tract at 4, 7, and 9 years of follow-up. In the first 2 patients the gradient was at the distal pulmonary anastomosis, and in the third patient it was at the level of the pulmonary valve. The fourth patient had moderate pulmonary regurgitation. However, these patients are asymptomatic and have not required reoperation.
Late Reoperation Late reoperation (Table 2) was required in 10 patients (1.1 events per 100 patient-years) after a mean period of 34.5 ⫾ 31.9 months (range, 3 to 108 months). Two patients required operation for fungal endocarditis 3 and 4 months after surgery. In both these patients the autograft was explanted and replaced with a cryopreserved aortic homograft. Both these patients died. Two patients required reoperation for failed mitral valve repair with normal neoaortic valve function. The first of these 2 patients underwent homograft mitral valve replacement 26 months after the initial operation; however she died 3 months later. The other patient underwent mechanical mitral valve replacement 31 months after initial operation. Three patients required mechanical double valve replacement for autograft failure and failure of associated mitral valve repair 12 and 48 months (2 patients) after initial operation. Two patients with severe aortic regurgitation in the rheumatic group underwent aortic valve replacement with a mechanical prosthesis 56 and 108 months after initial operation. The last patient, an 11year-old boy required mechanical aortic valve replacement for autograft failure secondary to a perforation in
Fig 2. Reoperation-free survival (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ nonrheumatic.
the autograft cusps, which was probably due to endocarditis. In this patient, repair of the autograft and homograft aortic valve replacement was not performed at the patient’s request. In patients undergoing reoperation other than for endocarditis, the geometry of the aortic sinuses was maintained. The autograft cusps showed thickening, retraction, and failure of coaptation without any prolapse or commissural fusion. Besides this, the aortic root was dilated at reoperation in 2 patients in the rheumatic group. In the rheumatic group, the histology of explanted autografts revealed valve thickening, fibrosis, active chronic inflammation with small vessel, and intimal proliferation. This has been described in detail in our prior publication [10]. At 90 months, freedom from reoperation was 91.8 ⫾ 2.7 % (95% CI 86.5 to 97.1) for the entire group; in rheumatics it was 90.5 ⫾ 3.7 % (95% CI 83.2 to 97.8) as compared with 95.3 ⫾ 2.7% (95% CI 90 to 100) in non-rheumatics. At 10 years, it was 86 ⫾ 6.1 % (95% CI 74 to 97.9) for the entire group; in rheumatics it was 76.24 ⫾ 12.4% (95% CI 52.7 to 99.6) compared with 95.3 ⫾ 2.7% (95% CI 90 to 100) in non-rheumatics (log rank test, 1.06; p ⫽ 0.30) (Fig 2).
Late Deaths and Survival There were 8 late deaths (0.9 events per 100 patientyears). These were due to infective endocarditis (n ⫽ 4), mitral homograft failure (n ⫽ 2), and persistent congestive heart failure due to autograft failure (n ⫽ 2). Of the 4 patients with infective endocarditis, 2 died after reoperation. One patient who had undergone mitral valve replacement with a homograft died 6 months later due to the mitral homograft failure. Another patient had a mitral homograft replacement after failed mitral valve repair. However, 2 months later, she died due to mitral homograft failure. Two patients with autograft failure died from persistent congestive heart failure 24 months and 5 months after operation. At 90 months, the actuarial survival was 88.3 ⫾ 2.7%
492
KUMAR ET AL ROSS PROCEDURE
Ann Thorac Surg 2005;80:488 –94
Table 3. Relative Risks in Rheumatics (n ⫽ 81) Compared With the Non-Rheumatics (n ⫽ 72)
CARDIOVASCULAR
Outcome Aortic regurgitations Early deaths Late deaths Reoperations
Rheumatics ⬍ 30 years (n ⫽ 40)
Rheumatics (all ages) (n ⫽ 81)
6.2 (1.9–20)* 3.1 (0.6–15.9) 2.63 (0.5–13.5) 2.1 (0.6–8.2)
3.6 (1.1–11.5) 1.3 (0.3–4.7) 2.5 (0.5–13.5) 1.99 (0.5–8.5)
* ⫽ Only significant observation (p ⫽ 0.0023), meaning that rheumatics ⬍ 30 years of age are at high risk compared to others. The reference category is non-rheumatics (relative risk ⫽ 1). Figures in parentheses indicate 95% confidence limits.
Fig 3. Actuarial survival (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ non-rheumatic. (Cum ⫽ cumulative.)
(95% CI 83 to 93.6) for the entire group; in rheumatics it was 86.1 ⫾ 3.9% (95% CI 78.5 to 93.7) compared with 91 ⫾ 3.5% (95% CI 84.1 to 97.9) in non-rheumatics. At 10 years, it was 86.9 ⫾ 2.9% (95% CI 81.2 to 92.6) for the entire group; in rheumatics it was 84.5 ⫾ 4.1% (95% CI 76.5 to 92.5) compared with 91 ⫾ 3.5% (95% CI 84.1 to 97.9) in non-rheumatics (log rank test, 0.89; p ⫽ 0.34) (Fig 3). Event-free survival at 90 months was 75 ⫾ 3.9% (95% CI 67.4 to 82.6) for the entire group; in rheumatics it was 68.9 ⫾ 5.3% (95% CI 58.6 to 99.2) compared with 86.2 ⫾ 4.9% (95% CI 76.6 to 95.8) in non-rheumatics. At 10 years, it was 68.3 ⫾ 5.3% (95% CI 58 to78.6) for the entire group; in rheumatics it was 61.2 ⫾ 6.4% (95% CI 51.6 to 70.8) compared with 86.2 ⫾ 4.9% (95% CI 76.6 to 95.8) in non-rheumatics (log rank test, 4.23; p ⫽ 0.039) (Fig 4). Multivariate analysis for identifying risk factors revealed that the risk of autograft failure was higher in young rheumatics (⬍ 30 years of age) compared with older patients with non-rheumatic aortic valve disease
Fig 4. Event-free survival (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ nonrheumatic.
(odds ratio, 4.2; 95% CI 1.6 to10.4; p ⫽ 0.0033). Relative risks for aortic regurgitation, early and late deaths, and reoperation are summarized in Table 3. There were no specific predictors of early or late deaths.
Comment After the pulmonary autograft was first implanted by Ross, the procedure has gradually gained widespread acceptance as a good valve substitute for patients requiring surgery for aortic valve disease [3–9]. However, some problems remain. There has been no consensus on the best technique of autograft implantation. Ross used the scalloped subcoronary technique with good results [16], whereas Elkins and colleagues [17] used the aortic root inclusion and root replacement technique. The subcoronary technique is now rarely used as it is more complex, especially in the asymmetrical annulus, and there is a higher incidence of autograft insufficiency and reoperation [5, 18]. The cylinder inclusion technique carries a higher incidence of autograft obstruction and distortion of coronary anastomosis because of a blood-filling space between the autograft and the native aortic wall [18, 19]. We believe that when the autograft is fixed to the relatively dense collagen structure of the aortic annulus with the root replacement technique, the distortion of the commissures is avoided, which leads to better autograft function in the mid-term. Also the asymmetry of the native aorta in bicuspid aortic valves can be ignored with this technique. There have been concerns with the use of root replacement technique in patients with bicuspid aortic valve disease as it predisposes to autograft dilatation leading to progressive aortic regurgitation [20]. Some have reported that this autograft dilatation is likely to occur if the autograft is implanted in a dilated aortic annulus [20 –22]. We have not observed important autograft dilatation in patients with a bicuspid aortic valve and also in the majority of the rheumatics. We do not perform this operation if the aortic annulus measures 30 mm or more on transesophageal echocardiography. In addition, a strip of pericardium was used in all adults to buttress the proximal suture line, and the autograft annulus was seated in the exact position of the native aortic valve annulus. However, autograft dilatation was observed in 2
young patients with rheumatic aortic valve disease, which was due to recurrent attacks of rheumatic activity and is consistent with our earlier observation that the pulmonary autograft is best avoided in this subset of patients [10]. Reoperation for autograft dysfunction has been a major concern [4 –9]. This may be due to technical factors, especially in the learning curve, progressive AR due to non-coaptation of the leaflets, or pulmonary autograft to aortic annulus mismatch. The major reason for reoperation in this group of patients was our enthusiastic use of the Ross procedure in the young rheumatic hoping for a cure. The observation of rheumatic valvulitis of the autograft on histology [10] in addition to the failed mitral valve procedures were the main reasons for abandoning this operation in young rheumatics and also in patients with associated mitral valve disease. Since 1998, we offer the Ross procedure for older patients (⬎ 30 years) with rheumatic heart disease only if they have isolated aortic valve disease. Freedom from reoperation in our series was more than 80% in the entire group and 95% in non-rheumatics. These results compare favorably with Ross’s own results in which 85% and 61% of the hospital survivors were free of reoperation 10 and 20 years, respectively, after the initial operation [6]. The freedom from reoperation on the pulmonary homograft was 100% at 10 years in our patients and has ranged from 80% to 86% in various series [4 –9]. We believe that the cryopreserved pulmonary homograft offers the best freedom from reoperation. We have not used other commercially available substitutes. The higher age of our patients, along with less number of children, may also be a possible reason for none of our patients requiring reoperation for pulmonary homograft dysfunction. Our early mortality of 6.5% is higher than the early mortality of 3.3% from the International Registry for the Ross procedure [23]. However, the early deaths in our series were predominantly in the early part of our experience. In the last 100 consecutive patients we have had only 3 early deaths. Therefore, although the Ross procedure may carry a higher mortality than a standard prosthetic aortic valve replacement, the results become fairly standardized after the learning curve is over. We did have a higher incidence of endocarditis in our patients, which was probably related to environmental factors due to unsatisfactory air conditioning. This problem has now been eliminated. One of the limitations of this study is that the actuarial probabilities are reported at either a median follow-up of 90 months or at 10 years using the Kaplan-Meier analysis. Calculating freedom from valve-related events in this manner may overestimate the truly occurring probabilities. However these are overcome to an extent by providing linearized rates. Also we had 10 patients with moderate aortic regurgitation and 4 with pulmonary homograft dysfunction who have not undergone reoperation yet because of having no change in the functional class. Although these patients have been included for the
KUMAR ET AL ROSS PROCEDURE
493
purpose of estimating the actuarial freedom from autograft dysfunction, they have not been included for calculating the reoperation-free survival. These patients are being closely followed. We now do not perform the Ross procedure in young rheumatics (⬍ 30 years) and we do offer it to older patients (⬎ 30 years) with rheumatic heart disease only if they have isolated aortic valve disease. In patients with non-rheumatic causes, the procedure is only performed if the aortic root size is 30 mm or less. It is not performed in patients requiring emergency surgery, significant left ventricular dysfunction, mitral valve disease requiring prosthetic valve replacement, and patients who have bicuspid pulmonary valve or pulmonary regurgitation. Proper patient selection has improved our results and helped us achieve better survival and autograft function.
Conclusion The Ross procedure is not suitable for young patients (⬍ 30 years) with rheumatic heart disease. It provides satisfactory hemodynamic and clinical results in properly selected patients with a low reoperation rate for the autograft and pulmonary homograft. Close surveillance of these patients is mandatory as many are likely to require a later reoperation. We have not observed important autograft dilatation in these patients in contrast to reports by others with the use of the root replacement technique.
References 1. Baudet EM, Puel V, McBride JT, et al. Long-term results of valve replacement with the St Jude medical prosthesis. J Thorac Cardiovasc Surg 1995;109:858 –70. 2. Ross DN. Replacement of aortic and mitral valves with a pulmonary autograft. Lancet 1967;2:956 – 8. 3. Ross D, Jackson M, Davies J. The pulmonary autograft: a permanent aortic valve. Eur J Cardiothorac Surg 1992;6: 113– 6. 4. Gatzoulis MA. Ross procedure: the treatment of choice for aortic valve disease? Int J Cardiol 1999;71:205– 6. 5. Choudhary SK, Govil A, Kumar AS. Ross procedure: aortic valve replacement with pulmonary autograft. Indian J Thorac Cardiovasc Surg 2001;17:243–57. 6. Chambers JC, Somerville J, Stone S, Ross DN. Pulmonary autograft procedure for aortic valve disease: long-term results of the pioneer series. Circulation 1997;96:2206 –14. 7. Elkins RC. The Ross operation: a 12-year experience. Ann Thorac Surg 1999; 68(3 Suppl):S14 – 8. 8. Oswalt JD, Dewan SJ, Mueller MC, Nelson S. Highlights of a ten-year experience with the Ross procedure. Ann Thorac Surg 2001;71(5 Suppl):S332–5. 9. Kouchoukos NT, Masetti P, Nickerson NJ, Castner CF, Shannon WD, Davila-Roman VG. The Ross procedure: longterm clinical and echocardiographic follow-up. Ann Thorac Surg 2004;78:773– 81. 10. Choudhary SK, Mathur A, Sharma R, et al. Pulmonary autograft: should it be used in young patients with rheumatic disease? J Thorac Cardiovasc Surg 1999;118:483–90. 11. Kumar AS, Rao PN, Dharmapuram AK, Chander H, Trehan H. Pulmonary autograft aortic valve replacement. Early experience with the Ross procedure. Tex Heart Inst J 1995; 22:177–9. 12. Choudhary SK, Talwar S, Dubey B, Chopra A, Saxena A, Kumar AS. Mitral valve repair in a predominantly rheumatic population: long-term results. Tex Heart Inst J 2001;28:8 –15.
CARDIOVASCULAR
Ann Thorac Surg 2005;80:488 –94
494
KUMAR ET AL ROSS PROCEDURE
CARDIOVASCULAR
13. Perry GJ, Hemeke F, Nanda MC, Byard C, Soto B. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol 1987;9:952–9. 14. Edmunds LH, Clark Re, Cohn LH, Grunkemeier GL, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. J Thorac Cardiovasc Surg 1996;112:708 –11. 15. Kumar AS, Rao PN, Trehan H. A technique to prevent bleeding after Ross procedure. J Heart Valve Dis 1995;4:405– 6. 16. Ross DN. The sub coronary technique versus the root technique for autograft surgery. J Heart Valve Dis 2003;12:553– 8. 17. Elkins RC, Santangelo K, Stelzer P, Randolph JD, KnottCraig CJ. Pulmonary autograft replacement of the aortic valve: an evolution of technique. J Card Surg 1992;7:108 –16. 18. O’Brien MF. Aortic valve implantation techniques: should they be any different for the pulmonary autograft and the aortic homograft? J Heart Valve Dis 1993;2:385–7.
Ann Thorac Surg 2005;80:488 –94
19. Elkins RC, Lane MM, McCue C. Pulmonary autograft reoperation: incidence and management. Ann Thorac Surg 1996; 62:450 –5. 20. David TE, Omran A, Ivanov J, et al. Dilation of the pulmonary autograft after the Ross procedure. J Thorac Cardiovasc Surg 2000;119:210 –20. 21. David TE, Omran A, Webb G, Rakowski H, Armstrong S, Sun Z. Geometric mismatch of the aortic and pulmonary roots causes aortic insufficiency after the Ross procedure. J Thorac Cardiovasc Surg 1996;112:1231–7. 22. Schmidtke C, Bechtel J, Hueppe M, Noetzold A, Sievers HH. Size and distensibility of the aortic root and aortic valve function after different techniques of the Ross procedure. J Thorac Cardiovasc Surg 2000;119:990 –7. 23. Ross registry international summary reports and graphs. The Ross Procedure International Registry, Mount Rushmore Heart Institute, Rapidy City, SD, 2004.
INVITED COMMENTARY The experience of Kumar and colleagues offers an extensive and detailed 10-year appraisal of the Ross procedure. However it is characterized by a relatively small number of patients in the growing age and a complete absence of any major pediatric cardiac operation (Ross Konno), which probably accounts for an in-hospital mortality lower than what has been reported by others [1]. In addition, 3 out of 4 cases were primarily operated on for bicuspid aortic valves. Thus it can be speculated that the frequent occurrence of autograft dilatation in this series (only 43% freedom from dilatation at 10 years) was related to dystrophy of the arterial wall known to be frequently combined with the bicuspid aortic valve. Aortic valve repair would have been another therapeutic option; unfortunately the lack of resistance of the cusp tissue can be responsible for suture dehiscence [2]. When compared with other biological devices, this series of Ross procedures appears to offer freedom from reoperation comparable with porcine bioprostheses [3] or aortic homografts [4] in a comparable population aged 20 to 40 years (approximately 80% at 10 years). On the other side, a mechanical valve would have offered a superior durability with an additional risk of bleeding, which along with a bileaflet valve in the aortic position and a low target international normalized ratio of 1.5 to 2.5, can be considered minimal in knowing the patient’s age [5]. When dilated, replacement of the supracoronary portion of the ascending aorta with a Dacron tube could have been achieved with virtually no additional risk. The author’s policy is to propose a Ross procedure to
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
every single patient whose life expectancy exceeds 20 years. Halfway to the end of the study, his data does not entirely support this strategy. We believe that the last statement of the present article, should be more carefully restricted to the group of patients less than 20 years of age and in whom a biological subsitute is absolutely necessary. Christophe Acar, MD Cardiovascular Surgery Hôpital de la Salpétrière 50-52 Bd Vincent Auriol, Paris 75013 France e-mail: [email protected]
References 1. Elkins RC, Knott-Craig CJ, Ward KE, Lane MM. The Ross operation in children: ten-year experience. Ann Thorac Surg 1998;65:496 –502. 2. Cosgrove DM, Rosenkranz ER, Hendren WG, Bartlett JC, Stewart WJ. Valvuloplasty for aortic insuffisance. J Thorac Cardiovasc Surg 1991;102:571–7. 3. Jamieson WRE, Ling H, Burr LH, et al. Carpentier Edwards supraannular porcine bioprosthesis evaluation over 15 years. Ann Thorac Surg 1998,66:S49 –52. 4. O’Brien MF, Harrocks S, Stafford EG, et al. The homograft aortic valve: a 29 year, 99.3% follow-up of 1022 valve replacements. J Heart Valve Dis 2001;10:334 –5. 5. Akins CW. Results with mechanical cardiac valvular prostheses. Ann Thorac Surg 1995,60:1836 – 44.
0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2005.05.001
Aortic Valve Replacement With the Pulmonary Autograft: Mid-Term Results Arkalgud Sampath Kumar, MCh, Sachin Talwar, MCh, Raghunath Mohapatra, MS, Anita Saxena, DM, and Rajvir Singh, MS, PhD Cardiothoracic Centre and Department of Biostatistics, All India Institute of Medical Sciences, New Delhi, India
Background. The purpose of this study is to assess the mid-term results of aortic valve replacement with the pulmonary autograft. Methods. From October 1993 through September 2003, 153 patients with aortic valve disease (81 rheumatic and 72 non-rheumatic), with a mean age of 28 ⴞ 14.2 years underwent the Ross procedure with root replacement technique and right ventricular outflow tract reconstruction using a homograft. Associated procedures included mitral valve repair (n ⴝ 19), open mitral commissurotomy (n ⴝ 15), tricuspid valve repair (n ⴝ 2), homograft mitral valve replacement (n ⴝ 2), and subaortic membrane resection (n ⴝ 1). Results. Early mortality was 6.5% (10 patients). Mean follow-up was 77 ⴞ 42 months (range, 7 to 132 months; median, 90 months). One hundred, twenty-one survivors (84.6%) had no significant aortic regurgitation. Reoperation was required in 10 patients for autograft dysfunction alone (n ⴝ 3), infective endocarditis (n ⴝ 2), autograft
dysfunction with failed mitral valve repair (n ⴝ 3), and failed mitral valve repair alone (n ⴝ 2). No reoperations were required for the pulmonary homograft. There were 8 late deaths. Actuarial and reoperation-free survival at 90 months were 91.% ⴞ 3.5%, 95.3% ⴞ 2.7%, in nonrheumatics and 86.1 ⴞ 3.9%, 90.5 ⴞ 3.7% in rheumatics, respectively. Freedom from significant aortic stenosis or regurgitation was 91.5 ⴞ 2.8% in non-rheumatics and 80.6 ⴞ 4.8% in rheumatics. Event-free survival was 86.2 ⴞ 4.9% in non-rheumatics and only 68.9 ⴞ 5.3% in rheumatics. Conclusions. The Ross procedure is not recommended for young patients (< 30 years) with rheumatic heart disease. It provides satisfactory hemodynamic and clinical results in properly selected patients. Important autograft dilatation was not observed in our patients.
D
aortoarteritis. A total of 28 patients (18 non-rheumatics and 10 rheumatics) were 15 years of age or less. In the rheumatic group, 40 patients were 30 years of age or younger (young rheumatics). One hundred, twenty-two patients (79.7%) had dyspnea as the predominant symptom. One hundred fourteen patients (74.5%) were in New York Heart Association functional class III and IV. Preoperative transthoracic echocardiography was performed in all patients. Cardiac catheterization and cine angiography was performed if there was suspicion of associated mitral valve or coronary artery disease. Informed consent was obtained from all patients. Intraoperative transesophageal echocardiography was performed in all patients. If the aortic root diameter was 30 mm or more, aortic valve replacement with a mechanical prosthesis was preferred. Ross procedure was not performed in patients requiring emergency surgery, significant left ventricular dysfunction, mitral valve disease requiring prosthetic valve replacement, significant mismatch between the aortic root diameter and the pulmonary valve, and in patients with bicuspid pulmonary valve or pulmonary regurgitation. In 2 patients with bicuspid aortic valves, the Ross procedure was abandoned after the pulmonary valve was found to be bicuspid at operation. Since 1998, we abandoned this procedure in young rheumatics (ie, those under 30 years of age).
espite the increasing trend toward valve repair, valve replacement with a mechanical valve or bioprosthesis is the most commonly performed operation for patients with aortic valve disease [1]. The pulmonary autograft, first used by Ross [2] in 1967 to replace the diseased aortic valve, has been shown to provide good haemodynamics, ability to remodel, and growth potential [3–9]. We first performed this procedure in October 1993 and reported our early experience in young patients with rheumatic heart disease [10]. In this study, we present detailed results in 153 consecutive patients.
Patients and Methods From October 1993 through September 2003, 153 patients (Table 1) underwent aortic valve replacement with the pulmonary autograft (Ross procedure). The cause of the aortic valve lesion was rheumatic in 81 patients (53%) and non-rheumatic in 72 patients (47%). The latter group included patients with bicuspid aortic valve (n ⫽ 64); degenerative aortic stenosis (n ⫽ 7); and 1 patient with Accepted for publication March 4, 2005. Address reprint requests to Dr Kumar, Department of Cardiothoracic and Vascular Surgery, Cardiothoracic Centre, All India Institute of Medical Sciences, Ansari Nagar, New Delhi, 110 029 India; e-mail: asampath_ [email protected].
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2005;80:488 –94) © 2005 by The Society of Thoracic Surgeons
0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2005.03.023
KUMAR ET AL ROSS PROCEDURE
Table 1. Profile of Patients Undergoing the Ross Procedure Cause Rheumatic Non rheumatic Male:female Age Symptoms Dyspnea Palpitations Chest pain Congestive heart failure New York Heart Association functional class I II III IV Clinical diagnosis Aortic stenosis Aortic regurgitation Aortic stenosis plus aortic regurgitation Associated mitral valve disease Associated tricuspid valve disease
Number
Percentage
81 72 103:50 8 mo to 65 y (mean, 28 ⫾ 14.2 y) 122 61 65 48
79.7 39.9 42.5 31.4
5 34 96 18
3.3 22.2 62.7 11.8
40 66 47
26 43 31
36
25
2
1.3
The surgical procedure has been described in detail earlier by us [5, 10, 11]. Root replacement technique with coronary artery implantation was used in all patients. The left coronary artery was not implanted as a button. Instead, the left coronary ostium was left attached to a tongue of the distal aortic wall. Right ventricle– pulmonary artery continuity was restored using an antibiotic preserved (n ⫽ 47) or cryopreserved homograft (n ⫽ 106) obtained from our own tissue valve bank. A pulmonary homograft was used in 151 patients and aortic homograft in 2 patients. Associated procedures included mitral valve repair (n ⫽ 19), open mitral commissurotomy (n ⫽ 15), tricuspid valve repair (n ⫽ 2), homograft mitral valve replacement (n ⫽ 2), and subaortic membrane resection (n ⫽ 1) using techniques described earlier by us [12]. For the isolated Ross procedure, the mean aortic cross-clamp and cardiopulmonary bypass times were 121.4 ⫾ 19.4 minutes (range, 81 to 197 min) and 153 ⫾ 24.7 minutes (range, 121 to 226 min), respectively. When the mitral valve procedure was done concomitantly, these increased to 135 ⫾ 31 minutes (109 to 230 min) and 168 ⫾ 32 minutes (140 to 280 min), respectively. After weaning from cardiopulmonary bypass, transesophageal echocardiography was performed in all patients to confirm normal autograft function and to assess associated procedures. Prior to discharge from the hospital, transthoracic echocardiography was performed in all patients, and this was repeated every 6 months. Aortic regurgitation (AR) was assessed on a scale of ⫹1 to ⫹4
489
according to published criteria [13]. Aortic regurgitation with a grade of ⫹1 was considered mild. Peak gradients less than 25 mm Hg across the aortic valve were considered as mild aortic stenosis and ⱖ50 mm Hg was considered significant aortic stenosis. No anticoagulants or anti-platelet drugs were prescribed. Long-acting benzathine penicillin 3 weekly was prescribed to all patients less than 45 years of age with rheumatic heart disease. All patients received itraconazole for 6 weeks after surgery as prophylaxis against fungal endocarditis.
Statistical Analysis Statistical analysis was performed using SPSS for Windows 7.5 software package (SPSS Inc, Chicago, IL). The rheumatic and non-rheumatic groups of patients were analyzed separately. Descriptive statistics (ie, mean and standard deviation) were calculated for continuous variables. Freedom from valve-related events and actuarial estimates were calculated using the Kaplan-Meier analysis. The estimates were compared with the Mantel-Cox (log-rank) test. A valve-related event was defined as any episode of thromboembolism, hemorrhage, congestive heart failure, infective endocarditis, structural deterioration, or significant gradients as per published criteria [14]. Linearized rates for these events are reported. Cox’s proportional hazard model was used to analyze factors associated with a higher early and late mortality and the development of significant aortic stenosis or AR.
Results Hospital Mortality All patients survived the operation. Thirty-day mortality was 6.5% (n ⫽ 10) and was due to bleeding (n ⫽ 3), arrhythmias (n ⫽ 3), persistent low cardiac output (n ⫽ 2), chest infection (n ⫽ 1), and fungal endocarditis (n ⫽ 1). All 3 deaths due to bleeding and its sequelae were in the first 12 patients operated on. Subsequently we modified our operative technique to eliminate this problem [15]. There were only 3 early deaths in the last 100 consecutive patients.
Early Reoperation Six patients required reoperation for excessive mediastinal bleeding. Four of these were in the first 12 patients in which the bleeding was from the raw surface of the posterior wall of the right ventricular outflow tract. In the other 2 patients, the source of bleeding was not related to the suture lines. One patient required reoperation within a month for fungal endocarditis of the autograft, and she died after surgery from multiorgan dysfunction.
Early Autograft Function In all patients, intraoperative transesophageal echocardiography revealed trivial or no aortic regurgitation. Three patients had mild mitral regurgitation. Transthoracic echocardiography prior to discharge from the hospital showed mild AR in 1 patient.
CARDIOVASCULAR
Ann Thorac Surg 2005;80:488 –94
490
KUMAR ET AL ROSS PROCEDURE
Ann Thorac Surg 2005;80:488 –94
Table 2. Reoperations After the Ross Procedure
CARDIOVASCULAR
SN
Age/Sex
Cause
Interval
Cause of Reoperation (mo)
Operation
Fate
Cause of Death
1 2 3 4 5 6 7 8 9 10
24/M 22/F 17/M 13/M 20/M 21/M 15/F 30/M 36/F 11/M
RHD BAV RHD RHD RHD RHD RHD BAV RHD BAV
108 3 48 12 48 56 26 4 31 17
RHD, AR, dilated autograft Fungal IE AR, MR AR, MR, TR AR, MR, dilated autograft AR MR Fungal IE * MR Cusp perforation
AVR HAVR DVR DVR plus TV repair DVR AVR HMVR HAVR MVR AVR
Survived Died Survived Survived Survived Survived Died after 3 months Died Survived Survived
NA IE, MSOF NA NA NA NA CHF IE, MSOF NA NA
AR ⫽ aortic regurgitation; AS ⫽ aortic stenosis; AVR ⫽ aortic valve replacement; BAV ⫽ bicuspid aortic valve; CHF ⫽ congestive heart failure; DVR ⫽ double valve replacement; F ⫽ female; HAVR ⫽ homograft aortic valve replacement; HMVR ⫽ homograft mitral valve replacement; IE ⫽ infective endocarditis; M ⫽ male; MR ⫽ mitral regurgitation; MSOF ⫽ multisystem organ failure; MVR ⫽ mitral valve replacement; NA ⫽ not applicable; RHD ⫽ rheumatic heart disease; SN ⫽ serial number; TR ⫽ tricuspid regurgitation.
Follow-Up All patients were seen in the outpatient clinic at 6-month intervals and underwent clinical examination and echocardiography. Between January 2004 and September 2004, the records of 140 of the 143 survivors were obtained, and their last follow-up during this period was taken to report the results. The follow-up data (97.9% complete) ranged from 7 to 132 months (mean, 77 ⫾ 42 months; median, 90 months) and totaled 917.6 patientyears. Among survivors, 20 patients (14%) were followed-up for 10 or more years, 62 (43.4%) were followed-up for 7 or more years, 12 (8.4%) for more than 5 years, 24 (16.8%) for more than 3 years, and 6 (4.2%) for more than 2 years.
Thromboembolism There were no thromboembolic complications in the survivors.
Hemolysis Four patients had significant hemolysis. In the first 3 patients this was associated with excessive mediastinal bleeding, disseminated intravascular coagulation, and early death. In the fourth patient, this was associated with mild mitral regurgitation and subsided gradually. The incidence of this complication in survivors was 0.1 event per 100 patient-years of follow-up.
Infective Endocarditis Five patients had infective endocarditis (0.4 events per 100 patient-years) develop. This was fungal in 3 patients and bacterial in 2. One patient with fungal endocarditis died within 1 month of surgery, the other 2 underwent homograft aortic valve replacement 3 and 4 months after the initial operation and died. The 2 patients with bacterial endocarditis died 9 and 13 months after operation.
Late Autograft Function At last follow-up between January 2004 and September 2004, transthoracic echocardiography showed no aortic regurgitation in 121 patients (84.6%). Nineteen patients
had significant aortic regurgitation (2 events per 100 patient-years). Moderate (n ⫽ 14) to severe (n ⫽ 5) aortic regurgitation was noted in 19 patients (13.3%) of the 143 survivors at a median follow-up of 90 months after operation. Fifteen of these patients had rheumatic heart disease, were young (⬍ 30 years of age), and had significant AR develop 57.6 ⫾ 36.2 months after the operation; five of these patients had evidence of recurrent attacks of rheumatic fever and had significant AR develop within 2 years of operation. Two of these (patients 1 and 5, Table 2) had stopped penicillin prophylaxis and required hospitalization for recurrent attacks of acute rheumatic fever. In the non-rheumatic group, AR had developed 81.9 ⫾ 42.8 months after the operation (log rank test, 1.97; p ⫽ 0.16). On multivariate analysis, the probability of developing of AR was higher in young rheumatics (⬍ 30 years) with an odds ratio of 4.2, and a 95% confidence interval (CI) of 1.6 to 10.4 (p ⫽ 0.0033). All 5 patients with severe AR underwent reoperation, and 2 of these required double valve replacement for failure of associated mitral valve repair. In all these patients, mechanical valves were implanted. One patient with moderate AR and failed mitral valve repair underwent double valve replacement with a mechanical valve 48 months after the initial operation. Patients 1 and 5 (Table 2) were found to have significant autograft dilatation due to recurrent attacks of rheumatic activity. Two patients died from congestive heart failure 24 and 5 months after the initial operation. One patient with moderate AR died of bacterial pneumonia 2 months after operation. The remaining 10 patients with moderate AR continue to be in New York Heart Association functional class I and are being closely followed-up. Two patients with a preoperative diagnosis of bicuspid aortic valve have a gradient of 25 mm Hg and 40 mm Hg 5 years and 7 years, respectively, after the initial operation, but they are asymptomatic. Autograft dilatation was not seen in any other patient. At 90 months, freedom from autograft dysfunction was 84.3 ⫾ 3.6 % (95% CI 77.3 to 91.3) for the entire group; in rheumatics it was 80.6 ⫾ 4.8% (95% CI 71.2 to 90) as compared with 91.5 ⫾ 2.8% (95% CI 86 to 97) in non-
KUMAR ET AL ROSS PROCEDURE
491
CARDIOVASCULAR
Ann Thorac Surg 2005;80:488 –94
Fig 1. Freedom from aortic regurgitation (AR) (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ non-rheumatic.
rheumatic group (log rank test, 1.9; p ⫽ 0.16). At 10 years, it was 79.2 ⫾ 4.9 % (95% CI 69.8 to 88.6) for the entire group; in rheumatics it was 74.7 ⫾ 6.1% (95% CI 62.7 to 86.7) as compared with 91.5 ⫾ 2.8% (95% CI 86 to 97) in the non-rheumatic group (Fig 1).
Fate of Right Ventricular Outflow Tract Four patients had pulmonary homograft dysfunction (0.4 events per 100 patient-years). Three patients had a gradient of 40 mm Hg or more across the right ventricular outflow tract at 4, 7, and 9 years of follow-up. In the first 2 patients the gradient was at the distal pulmonary anastomosis, and in the third patient it was at the level of the pulmonary valve. The fourth patient had moderate pulmonary regurgitation. However, these patients are asymptomatic and have not required reoperation.
Late Reoperation Late reoperation (Table 2) was required in 10 patients (1.1 events per 100 patient-years) after a mean period of 34.5 ⫾ 31.9 months (range, 3 to 108 months). Two patients required operation for fungal endocarditis 3 and 4 months after surgery. In both these patients the autograft was explanted and replaced with a cryopreserved aortic homograft. Both these patients died. Two patients required reoperation for failed mitral valve repair with normal neoaortic valve function. The first of these 2 patients underwent homograft mitral valve replacement 26 months after the initial operation; however she died 3 months later. The other patient underwent mechanical mitral valve replacement 31 months after initial operation. Three patients required mechanical double valve replacement for autograft failure and failure of associated mitral valve repair 12 and 48 months (2 patients) after initial operation. Two patients with severe aortic regurgitation in the rheumatic group underwent aortic valve replacement with a mechanical prosthesis 56 and 108 months after initial operation. The last patient, an 11year-old boy required mechanical aortic valve replacement for autograft failure secondary to a perforation in
Fig 2. Reoperation-free survival (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ nonrheumatic.
the autograft cusps, which was probably due to endocarditis. In this patient, repair of the autograft and homograft aortic valve replacement was not performed at the patient’s request. In patients undergoing reoperation other than for endocarditis, the geometry of the aortic sinuses was maintained. The autograft cusps showed thickening, retraction, and failure of coaptation without any prolapse or commissural fusion. Besides this, the aortic root was dilated at reoperation in 2 patients in the rheumatic group. In the rheumatic group, the histology of explanted autografts revealed valve thickening, fibrosis, active chronic inflammation with small vessel, and intimal proliferation. This has been described in detail in our prior publication [10]. At 90 months, freedom from reoperation was 91.8 ⫾ 2.7 % (95% CI 86.5 to 97.1) for the entire group; in rheumatics it was 90.5 ⫾ 3.7 % (95% CI 83.2 to 97.8) as compared with 95.3 ⫾ 2.7% (95% CI 90 to 100) in non-rheumatics. At 10 years, it was 86 ⫾ 6.1 % (95% CI 74 to 97.9) for the entire group; in rheumatics it was 76.24 ⫾ 12.4% (95% CI 52.7 to 99.6) compared with 95.3 ⫾ 2.7% (95% CI 90 to 100) in non-rheumatics (log rank test, 1.06; p ⫽ 0.30) (Fig 2).
Late Deaths and Survival There were 8 late deaths (0.9 events per 100 patientyears). These were due to infective endocarditis (n ⫽ 4), mitral homograft failure (n ⫽ 2), and persistent congestive heart failure due to autograft failure (n ⫽ 2). Of the 4 patients with infective endocarditis, 2 died after reoperation. One patient who had undergone mitral valve replacement with a homograft died 6 months later due to the mitral homograft failure. Another patient had a mitral homograft replacement after failed mitral valve repair. However, 2 months later, she died due to mitral homograft failure. Two patients with autograft failure died from persistent congestive heart failure 24 months and 5 months after operation. At 90 months, the actuarial survival was 88.3 ⫾ 2.7%
492
KUMAR ET AL ROSS PROCEDURE
Ann Thorac Surg 2005;80:488 –94
Table 3. Relative Risks in Rheumatics (n ⫽ 81) Compared With the Non-Rheumatics (n ⫽ 72)
CARDIOVASCULAR
Outcome Aortic regurgitations Early deaths Late deaths Reoperations
Rheumatics ⬍ 30 years (n ⫽ 40)
Rheumatics (all ages) (n ⫽ 81)
6.2 (1.9–20)* 3.1 (0.6–15.9) 2.63 (0.5–13.5) 2.1 (0.6–8.2)
3.6 (1.1–11.5) 1.3 (0.3–4.7) 2.5 (0.5–13.5) 1.99 (0.5–8.5)
* ⫽ Only significant observation (p ⫽ 0.0023), meaning that rheumatics ⬍ 30 years of age are at high risk compared to others. The reference category is non-rheumatics (relative risk ⫽ 1). Figures in parentheses indicate 95% confidence limits.
Fig 3. Actuarial survival (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ non-rheumatic. (Cum ⫽ cumulative.)
(95% CI 83 to 93.6) for the entire group; in rheumatics it was 86.1 ⫾ 3.9% (95% CI 78.5 to 93.7) compared with 91 ⫾ 3.5% (95% CI 84.1 to 97.9) in non-rheumatics. At 10 years, it was 86.9 ⫾ 2.9% (95% CI 81.2 to 92.6) for the entire group; in rheumatics it was 84.5 ⫾ 4.1% (95% CI 76.5 to 92.5) compared with 91 ⫾ 3.5% (95% CI 84.1 to 97.9) in non-rheumatics (log rank test, 0.89; p ⫽ 0.34) (Fig 3). Event-free survival at 90 months was 75 ⫾ 3.9% (95% CI 67.4 to 82.6) for the entire group; in rheumatics it was 68.9 ⫾ 5.3% (95% CI 58.6 to 99.2) compared with 86.2 ⫾ 4.9% (95% CI 76.6 to 95.8) in non-rheumatics. At 10 years, it was 68.3 ⫾ 5.3% (95% CI 58 to78.6) for the entire group; in rheumatics it was 61.2 ⫾ 6.4% (95% CI 51.6 to 70.8) compared with 86.2 ⫾ 4.9% (95% CI 76.6 to 95.8) in non-rheumatics (log rank test, 4.23; p ⫽ 0.039) (Fig 4). Multivariate analysis for identifying risk factors revealed that the risk of autograft failure was higher in young rheumatics (⬍ 30 years of age) compared with older patients with non-rheumatic aortic valve disease
Fig 4. Event-free survival (Kaplan-Meier analysis) in patients undergoing the Ross procedure. Œ ⫽ rheumatic; ‚ ⫽ nonrheumatic.
(odds ratio, 4.2; 95% CI 1.6 to10.4; p ⫽ 0.0033). Relative risks for aortic regurgitation, early and late deaths, and reoperation are summarized in Table 3. There were no specific predictors of early or late deaths.
Comment After the pulmonary autograft was first implanted by Ross, the procedure has gradually gained widespread acceptance as a good valve substitute for patients requiring surgery for aortic valve disease [3–9]. However, some problems remain. There has been no consensus on the best technique of autograft implantation. Ross used the scalloped subcoronary technique with good results [16], whereas Elkins and colleagues [17] used the aortic root inclusion and root replacement technique. The subcoronary technique is now rarely used as it is more complex, especially in the asymmetrical annulus, and there is a higher incidence of autograft insufficiency and reoperation [5, 18]. The cylinder inclusion technique carries a higher incidence of autograft obstruction and distortion of coronary anastomosis because of a blood-filling space between the autograft and the native aortic wall [18, 19]. We believe that when the autograft is fixed to the relatively dense collagen structure of the aortic annulus with the root replacement technique, the distortion of the commissures is avoided, which leads to better autograft function in the mid-term. Also the asymmetry of the native aorta in bicuspid aortic valves can be ignored with this technique. There have been concerns with the use of root replacement technique in patients with bicuspid aortic valve disease as it predisposes to autograft dilatation leading to progressive aortic regurgitation [20]. Some have reported that this autograft dilatation is likely to occur if the autograft is implanted in a dilated aortic annulus [20 –22]. We have not observed important autograft dilatation in patients with a bicuspid aortic valve and also in the majority of the rheumatics. We do not perform this operation if the aortic annulus measures 30 mm or more on transesophageal echocardiography. In addition, a strip of pericardium was used in all adults to buttress the proximal suture line, and the autograft annulus was seated in the exact position of the native aortic valve annulus. However, autograft dilatation was observed in 2
young patients with rheumatic aortic valve disease, which was due to recurrent attacks of rheumatic activity and is consistent with our earlier observation that the pulmonary autograft is best avoided in this subset of patients [10]. Reoperation for autograft dysfunction has been a major concern [4 –9]. This may be due to technical factors, especially in the learning curve, progressive AR due to non-coaptation of the leaflets, or pulmonary autograft to aortic annulus mismatch. The major reason for reoperation in this group of patients was our enthusiastic use of the Ross procedure in the young rheumatic hoping for a cure. The observation of rheumatic valvulitis of the autograft on histology [10] in addition to the failed mitral valve procedures were the main reasons for abandoning this operation in young rheumatics and also in patients with associated mitral valve disease. Since 1998, we offer the Ross procedure for older patients (⬎ 30 years) with rheumatic heart disease only if they have isolated aortic valve disease. Freedom from reoperation in our series was more than 80% in the entire group and 95% in non-rheumatics. These results compare favorably with Ross’s own results in which 85% and 61% of the hospital survivors were free of reoperation 10 and 20 years, respectively, after the initial operation [6]. The freedom from reoperation on the pulmonary homograft was 100% at 10 years in our patients and has ranged from 80% to 86% in various series [4 –9]. We believe that the cryopreserved pulmonary homograft offers the best freedom from reoperation. We have not used other commercially available substitutes. The higher age of our patients, along with less number of children, may also be a possible reason for none of our patients requiring reoperation for pulmonary homograft dysfunction. Our early mortality of 6.5% is higher than the early mortality of 3.3% from the International Registry for the Ross procedure [23]. However, the early deaths in our series were predominantly in the early part of our experience. In the last 100 consecutive patients we have had only 3 early deaths. Therefore, although the Ross procedure may carry a higher mortality than a standard prosthetic aortic valve replacement, the results become fairly standardized after the learning curve is over. We did have a higher incidence of endocarditis in our patients, which was probably related to environmental factors due to unsatisfactory air conditioning. This problem has now been eliminated. One of the limitations of this study is that the actuarial probabilities are reported at either a median follow-up of 90 months or at 10 years using the Kaplan-Meier analysis. Calculating freedom from valve-related events in this manner may overestimate the truly occurring probabilities. However these are overcome to an extent by providing linearized rates. Also we had 10 patients with moderate aortic regurgitation and 4 with pulmonary homograft dysfunction who have not undergone reoperation yet because of having no change in the functional class. Although these patients have been included for the
KUMAR ET AL ROSS PROCEDURE
493
purpose of estimating the actuarial freedom from autograft dysfunction, they have not been included for calculating the reoperation-free survival. These patients are being closely followed. We now do not perform the Ross procedure in young rheumatics (⬍ 30 years) and we do offer it to older patients (⬎ 30 years) with rheumatic heart disease only if they have isolated aortic valve disease. In patients with non-rheumatic causes, the procedure is only performed if the aortic root size is 30 mm or less. It is not performed in patients requiring emergency surgery, significant left ventricular dysfunction, mitral valve disease requiring prosthetic valve replacement, and patients who have bicuspid pulmonary valve or pulmonary regurgitation. Proper patient selection has improved our results and helped us achieve better survival and autograft function.
Conclusion The Ross procedure is not suitable for young patients (⬍ 30 years) with rheumatic heart disease. It provides satisfactory hemodynamic and clinical results in properly selected patients with a low reoperation rate for the autograft and pulmonary homograft. Close surveillance of these patients is mandatory as many are likely to require a later reoperation. We have not observed important autograft dilatation in these patients in contrast to reports by others with the use of the root replacement technique.
References 1. Baudet EM, Puel V, McBride JT, et al. Long-term results of valve replacement with the St Jude medical prosthesis. J Thorac Cardiovasc Surg 1995;109:858 –70. 2. Ross DN. Replacement of aortic and mitral valves with a pulmonary autograft. Lancet 1967;2:956 – 8. 3. Ross D, Jackson M, Davies J. The pulmonary autograft: a permanent aortic valve. Eur J Cardiothorac Surg 1992;6: 113– 6. 4. Gatzoulis MA. Ross procedure: the treatment of choice for aortic valve disease? Int J Cardiol 1999;71:205– 6. 5. Choudhary SK, Govil A, Kumar AS. Ross procedure: aortic valve replacement with pulmonary autograft. Indian J Thorac Cardiovasc Surg 2001;17:243–57. 6. Chambers JC, Somerville J, Stone S, Ross DN. Pulmonary autograft procedure for aortic valve disease: long-term results of the pioneer series. Circulation 1997;96:2206 –14. 7. Elkins RC. The Ross operation: a 12-year experience. Ann Thorac Surg 1999; 68(3 Suppl):S14 – 8. 8. Oswalt JD, Dewan SJ, Mueller MC, Nelson S. Highlights of a ten-year experience with the Ross procedure. Ann Thorac Surg 2001;71(5 Suppl):S332–5. 9. Kouchoukos NT, Masetti P, Nickerson NJ, Castner CF, Shannon WD, Davila-Roman VG. The Ross procedure: longterm clinical and echocardiographic follow-up. Ann Thorac Surg 2004;78:773– 81. 10. Choudhary SK, Mathur A, Sharma R, et al. Pulmonary autograft: should it be used in young patients with rheumatic disease? J Thorac Cardiovasc Surg 1999;118:483–90. 11. Kumar AS, Rao PN, Dharmapuram AK, Chander H, Trehan H. Pulmonary autograft aortic valve replacement. Early experience with the Ross procedure. Tex Heart Inst J 1995; 22:177–9. 12. Choudhary SK, Talwar S, Dubey B, Chopra A, Saxena A, Kumar AS. Mitral valve repair in a predominantly rheumatic population: long-term results. Tex Heart Inst J 2001;28:8 –15.
CARDIOVASCULAR
Ann Thorac Surg 2005;80:488 –94
494
KUMAR ET AL ROSS PROCEDURE
CARDIOVASCULAR
13. Perry GJ, Hemeke F, Nanda MC, Byard C, Soto B. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol 1987;9:952–9. 14. Edmunds LH, Clark Re, Cohn LH, Grunkemeier GL, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. J Thorac Cardiovasc Surg 1996;112:708 –11. 15. Kumar AS, Rao PN, Trehan H. A technique to prevent bleeding after Ross procedure. J Heart Valve Dis 1995;4:405– 6. 16. Ross DN. The sub coronary technique versus the root technique for autograft surgery. J Heart Valve Dis 2003;12:553– 8. 17. Elkins RC, Santangelo K, Stelzer P, Randolph JD, KnottCraig CJ. Pulmonary autograft replacement of the aortic valve: an evolution of technique. J Card Surg 1992;7:108 –16. 18. O’Brien MF. Aortic valve implantation techniques: should they be any different for the pulmonary autograft and the aortic homograft? J Heart Valve Dis 1993;2:385–7.
Ann Thorac Surg 2005;80:488 –94
19. Elkins RC, Lane MM, McCue C. Pulmonary autograft reoperation: incidence and management. Ann Thorac Surg 1996; 62:450 –5. 20. David TE, Omran A, Ivanov J, et al. Dilation of the pulmonary autograft after the Ross procedure. J Thorac Cardiovasc Surg 2000;119:210 –20. 21. David TE, Omran A, Webb G, Rakowski H, Armstrong S, Sun Z. Geometric mismatch of the aortic and pulmonary roots causes aortic insufficiency after the Ross procedure. J Thorac Cardiovasc Surg 1996;112:1231–7. 22. Schmidtke C, Bechtel J, Hueppe M, Noetzold A, Sievers HH. Size and distensibility of the aortic root and aortic valve function after different techniques of the Ross procedure. J Thorac Cardiovasc Surg 2000;119:990 –7. 23. Ross registry international summary reports and graphs. The Ross Procedure International Registry, Mount Rushmore Heart Institute, Rapidy City, SD, 2004.
INVITED COMMENTARY The experience of Kumar and colleagues offers an extensive and detailed 10-year appraisal of the Ross procedure. However it is characterized by a relatively small number of patients in the growing age and a complete absence of any major pediatric cardiac operation (Ross Konno), which probably accounts for an in-hospital mortality lower than what has been reported by others [1]. In addition, 3 out of 4 cases were primarily operated on for bicuspid aortic valves. Thus it can be speculated that the frequent occurrence of autograft dilatation in this series (only 43% freedom from dilatation at 10 years) was related to dystrophy of the arterial wall known to be frequently combined with the bicuspid aortic valve. Aortic valve repair would have been another therapeutic option; unfortunately the lack of resistance of the cusp tissue can be responsible for suture dehiscence [2]. When compared with other biological devices, this series of Ross procedures appears to offer freedom from reoperation comparable with porcine bioprostheses [3] or aortic homografts [4] in a comparable population aged 20 to 40 years (approximately 80% at 10 years). On the other side, a mechanical valve would have offered a superior durability with an additional risk of bleeding, which along with a bileaflet valve in the aortic position and a low target international normalized ratio of 1.5 to 2.5, can be considered minimal in knowing the patient’s age [5]. When dilated, replacement of the supracoronary portion of the ascending aorta with a Dacron tube could have been achieved with virtually no additional risk. The author’s policy is to propose a Ross procedure to
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
every single patient whose life expectancy exceeds 20 years. Halfway to the end of the study, his data does not entirely support this strategy. We believe that the last statement of the present article, should be more carefully restricted to the group of patients less than 20 years of age and in whom a biological subsitute is absolutely necessary. Christophe Acar, MD Cardiovascular Surgery Hôpital de la Salpétrière 50-52 Bd Vincent Auriol, Paris 75013 France e-mail: [email protected]
References 1. Elkins RC, Knott-Craig CJ, Ward KE, Lane MM. The Ross operation in children: ten-year experience. Ann Thorac Surg 1998;65:496 –502. 2. Cosgrove DM, Rosenkranz ER, Hendren WG, Bartlett JC, Stewart WJ. Valvuloplasty for aortic insuffisance. J Thorac Cardiovasc Surg 1991;102:571–7. 3. Jamieson WRE, Ling H, Burr LH, et al. Carpentier Edwards supraannular porcine bioprosthesis evaluation over 15 years. Ann Thorac Surg 1998,66:S49 –52. 4. O’Brien MF, Harrocks S, Stafford EG, et al. The homograft aortic valve: a 29 year, 99.3% follow-up of 1022 valve replacements. J Heart Valve Dis 2001;10:334 –5. 5. Akins CW. Results with mechanical cardiac valvular prostheses. Ann Thorac Surg 1995,60:1836 – 44.
0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2005.05.001