|
Supplement to
THE ANNALS OF THORACIC SURGERY Official Journal of The Society of Thoracic Surgeons and the Southern Thoracic Surgical Association
Proceedings of the VII International Symposium on Cardiac Bioprostheses, Volume 71 • Number 5 (Suppl.) • May 2001 • ISSN 0003-4975 Highlights of a Ten-Year Experience With the Ross Procedure
John D. Oswalt, MD, Stephen J. Dewan, MD, Michael C. Mueller, MD, and Suzi Nelson, RN Background. A review of a 10-year experience with the Ross procedure as a root replacement by a single group of surgeons featuring specific highlights is presented. Highlights include our results from a subset of patients with endocarditis and their management and a comparison of outcomes in patients with aortic insufficiency based on technical changes made after 5 years’ experience. Methods. The total patient group was 191, with 148 male and 43 female participants with an age range from 1 day to 69 years. Five of the patients in the 0-to-20 age group were newborns. Fifty-three of the adults were operated on for infectious endocarditis. In the total patient group 43% had aortic insufficiency, 28% had aortic stenosis, and 29% had mixed disease. Results. Operative mortality was 5.2% with a late mortality of 2.6%. The actuarial survival was 90.2% at 10 years. Freedom from autograft explantation was 93.2% and freedom from homograft replacement was 98.4%. The endocarditis patients had an operative mortality of 3.8% with 100% cure of the infection. Freedom from reinfection on the autograft was 98.1%, and freedom from infection of the pulmonary homograft was 98.1%. The actuarial survival was 86.3%. Conclusions. A specific review of the patients with aortic insufficiency resulted in a failure of the autograft in 7 patients among a cohort of 41 during the first 5 years of the study. After a change in technique in which the aortic annulus is narrowed and fixed to a measured size to match the body surface area, we have had no failures in the autograft. Although these results are early, we believe that these data support the use of the autograft as an excellent choice for replacement of the aortic valve in infective endocarditis. Finally, the use of the autograft for aortic insufficiency is reasonable with fixation of the aortic annulus so that subsequent dilation does not occur. (Ann Thorac Surg 2001;71:S332-5)© 2001 by The Society of Thoracic Surgeons 0ver the last 10 years, this single group of surgeons has embraced the Ross procedure as an excellent replacement for the aortic valve, not only in the pediatric age group but also in active adults. Although there are several 10-year series on the Ross procedure, we have found that our review yielded some particular highlights. These highlights include our outcomes and management in a subset of patients with aortic endocarditis as well as a comparison of outcomes in patients with aortic insufficiency based on technical changes made after 5 years’ experience. Material and Methods From November 1, 1990, to November 1, 2000, 191 patients (148 males and 43 females) had a Ross operation by total root replacement. The age range for our series was wide, as the Ross is performed in pediatric and adult patients (Fig 1). Five patients in the pediatric age group were high-risk newborns, with 1 in particular being a 1-day-old diagnosed with critical aortic stenosis. This patient was diagnosed in utero with aortic stenosis and the replacement was planned to follow the delivery. Fifty-three of the adult patients were operated on for aortic endocarditis. Included in this group were 2 extremely ill patients: 1 with cardiogenic shock secondary to aortoatrial fistula and 1 with an acute ventricular septal defect (VSD) and mitral regurgitation secondary to disruption of the ventricular septum and mitral annulus with abscess. Other severe cases have previously been reported [1]. Our series included 83 patients with aortic insufficiency (43%), 53 patients with aortic stenosis (28%), and 55 patients with mixed disease (29%). Concomitant operations included mitral valve repair in 4 patients, tailoring ascending aortoplasty in 17 patients, aortic annuloplasty in 45 patients, coronary bypass grafting in 7 patients, VSD repair in 1 patient, atrial septal defect repair in 1 patient, repair of aortoatrial fistula in 1 patient, ascending aortic replacement in 1 patient, patent ductus arteriosus ligation in 1 patient, and Ross-Konno procedures in 4 patients. Previous cardiac operations included: freehand aortic homograft in 1 patient, VSD repair in 1 patient, balloon valvuloplasty in 4 patients, coarctation repair in 4 patients, atrial septal defect repair in 1 patient, subaortic stenosis in 1 patient, porcine valve replacement in 1 patient, mechanical valve replacement in 1 patient, and aortic valvotomy in 4 patients. All patients are registered in the Ross Procedure Registry and follow-up echocardiography has followed those guidelines (79% follow-up). The Kaplan-Meier method was used for actuarial survival and X’ for freedom from reoperations.
The root replacement technique has been described previously [2]. Initially, a running technique with three triangulating sutures was used for the proximal suture line of the autograph. We then transitioned to a single running 4-0 Prolene (Ethicon, Somerville, NJ) for the proximal anastomosis within the first year of performing the operation. In the initial 5 years, the autograph was implanted into the native aortic annulus with no attention to the size of the annulus. Patients with annuli greater than 27 mm were excluded in the first 5 years. Annular buttressing with autologous pericardium was generally used. In the second 5 years, patients with aortic annuli greater than that consistent with their body surface areas (BSA) had an annular narrowing and fixation by 2-0 Prolene placed subannularly in a double-row pursestring. The annulus was narrowed. to a size consistent with the BSA derived from a chart of standard human heart valve diameters [3]. No exclusion by annular size was made in this group of patients. Concomitantly, if a narrowing annuloplasty was performed, then the ascending aorta was tailored so the sinotubular junction of the autograft was matched to the annulus, to within 2 mm. The end result achieved an annulus consistent with BSA and a sinotubular junction of similar size. All annuli were reinforced with autologous pericardium. If the ascending aorta appeared thin or aneurysmal, a Teflon felt was used to reinforce the distal suture line. Results Overall operative mortality was 5.2% with a late mortality of 2.6%. There have been no valve-related late deaths. Early autograft function by intraoperative transesophageal echocardiography was zero to trace aortic insufficiency. A single endocarditis patient in the early 5-year group had 2+ aortic insufficiency at the end of the operation. Due to the diagnosis, the 2+ insufficiency was acceptable and the patient was cured of endocarditis. Later the patient had to have his autograft explanted secondary to 3 to 4+ aortic insufficiency. Twelve patients (6.3%) have had their autograft replaced because of aortic insufficiency. Eleven of these cases occurred in the first 5 years of the study. The single replacement in the second 5 years occurred at 2 months secondary to a leaflet tear. Ten of the remaining replacements were for annular or sinotubular dilation. Homograft replacement occurred in only 3 patients. One of the replacements occurred in the 1-day-old patient and was replaced for growth reasons. He is now 6 years of age and doing well. The other 2 patients had pulmonary stenosis. We believe the cause of stenosis was likely related to a rejection phenomenon. Actuarial survival was 90.2% at 10 years. Ninety-eight percent of survivors are NYHA class I. Among the 53 patients (28%) with endocarditis, we found an operative mortality of 3.8%, with 100% cure of the infection. Freedom from reinfection on the autograft is 98.1%. One patient who abused intravenous cocaine both before his first endocarditis and then after the Ross procedure accounted for the single reinfection in the fifth year after the original Ross operation. The pulmonary homograft had a freedom from infection of 98.1%. The reinfection was cured with medical management. There have been 5.8% late deaths, none valve-related. Actuarial survival at 10 years was 86.3°/a. We also specifically reviewed the 83 patients (42.4%) with a diagnosis of aortic insufficiency. These patients were separated into two groups: group A (n = 41), who underwent operations during the first 5 years or our experience and group B (n = 42), who underwent operations during the second 5 years of our experience. This grouping is consistent with our change in technique of annuloplasty and fixation at the 5-year mark. Mean follow-up for group A was 7.7 years. Mean follow-up for group B was 2.2 years. The results show an early mortality rate of 2.4% in group A and 2.5% in group B. There has been a late mortality of 9.8% in Group A and no late deaths in Group B (p = NS). Seven autografts were explanted for progressive aortic insufficiency in group A and no autografts explanted in group B. When the first 5 years follow-up of group A (mean 2.9 years) was compared with the first 5 years follow-up of group B (mean 2.2 years) the results were statistically significant (p = < 0.02). Echocardiographic data of survivors is shown in Figure 2. Comment In the United States following the leadership of Elkins [4], more surgeons are adding the Ross procedure to their skills. Whereas it is an operation that requires much attention to technical detail, we believe the procedure also requires study and review to determine what changes should be made to improve it. In the 30-plus years since Ross’s first description, the procedure has been modified several times [2]. Most of these modifications have occurred in the last 10 years. There are more than 200 surgeons listing patients in the Ross Procedure Registry. It is possible that there are 200 modifications to the procedure. Certainly far more variables exist than that seen with replacement of a mechanical aortic valve and yet the data from the Registry suggest that the autograft performs as well in durability, lack of thromboembolism, and survivability as any valve available [5]. Zellner and coworkers [6] reported on a 17-year series of patients (mean age 54.8 ± 15.1 years) each with a St. Jude mechanical aortic valve; their actuarial survival was 58% ± 3.2% and the thromboembolism rate was 2.3% per year. A shorter follow-up reported by Arom and coworkers [7] had a 6-year survival rate with the St. Jude as an aortic replacement of 78.4% ± 5%. In his discussion of Zellner and colleagues’ study [6], Arom and coworkers revealed a 20-year survival rate of 54%. In a prospective randomized trial between autografts and homografts, Aklog and coworkers [8] found a 4-year actuarial survival of 87.7% in the homografts while autografts had a 4-year actuarial survival of 95.3%. Although not statistically significant at 4 years, their conclusions suggested the superiority of the autograft. Our actuarial survival of 90.2% is similar to other 10-year Ross series [4, 9]. The freedom from autograft replacement of 93.4% in the entire group of Ross patients remains acceptable. 
Close follow-up and critique of our surgical technique lead to our modification of annular narrowing and fixation. Clearly, most failures occurred in the patients with aortic insufficiency; we believe secondary to dilation of the annulus or the sinotubular junction. The annular dilatation may have occurred due to lack of pericardial support. Unfortunately, our documentation cannot support this conclusion. Although early but statistically significant, our follow-up of group B patients has yielded promising results. The concerns discussed by David and colleagues [10] regarding the use of the autograft in patients with aortic insufficiency has been noted. We remain watchful as these early results suggest that annular fixation and sinotubular matching will prevent late dilation and aortic insufficiency. A limitation in this study is the lack of annular and sinus size comparisons over the years. What remains clinically pertinent are our echo data on aortic insufficiency; 98% of the total population was NYHA class I. Our technique leaves only the sinuses unbuttressed, but sinus dilation without sinotubular or annular dilation does not promote aortic insufficiency. Currently we are offering the Ross procedure to patients with annuli greater than 27 mm. This approach is also supported by Elkins and colleagues [11]. Some surgeons, however, would disagree with this approach and recommend subcoronary implantation in patients with bicuspid valves and aortic insufficiency [12]. Our use of the Ross procedure for infective aortic endocarditis began in 1990 with our first report published in 1993 [1]. Subsequent papers have concurred with this practice [13]. A loose comparison with other endocarditis series that used mechanical or homograft valves as replacements has not matched the results of the autograft in operative mortality, reinfection, thrornboembolism, or survivability [14]. A comparative prospective trial is needed; however, in a clinical practice, we find it difficult to replace a 20-year-old patient’s valve with a homograft. The endocarditis may be cured but the patient will need a reoperation in 10 to 15 years because of homograft failure. The autograft, with its flexibility, viability, and availability, has made our approach to infective endocarditis the key to success. We believe foremost that aortic endocarditis is a surgical disease. Once the function of the aortic valve is impaired, that valve will need replacement. The use of the autograft with a 100% cure rate has allowed for early surgical replacement. Awaiting irradication of the infection is not necessary and in fact increases the risks of emboli and continued infection with abscess formation. We now operate within 24 to 48 hours after the diagnosis of aortic endocarditis if the valve is insufficient. If the valve has only vegetations and no aortic insufficiency, then the patient continues on medical management. Should the patient continue to have fever, emboli, or conduction defects, the medical management transitions to surgical treatment. This aggressive approach has allowed us to operate on patients who have fewer complications before the operation. With fewer annular abscesses, we have easier operations; with fewer emboli and no conduction defects the patients have less morbidity and better outcomes. Also significant in our approach to the endocarditis patients is replacement by the root operation. This method allows for total extirpation of the infected material, and reconstruction is made easier by the flexibility of the valve. Ralph-Edwards and colleagues [15] have shown good results in patients with prosthetic endocarditis who have valved-conduit replacements. Their essential technique for success in curing the infection was complete debridement, which requires removing much of the annulus and possibly the aortic root. The autograft conforms to the debrided annulus better than a rigid mechanical sewing ring, yielding less postoperative bleeding or paravalvular leaks. We have found the pulmonary homograft is readily available, therefore there is no delay in the surgical timing. In summary, our 10-year review shows that the Ross procedure is a highly successful operation for a wide range of patients. Our modifications with the annular narrowing and fixation have yielded early-improved results in the development of late aortic insufficiency. Finally, the autograft, as part of our overall management of aortic endocarditis, has proved to be an excellent choice for valve replacement in critically ill patients. References
|
