Use of Stentless Xenografts in the Aortic Position: Determinants of Early and Late Outcome

Transcription

1 Use of Stentless Xenografts in the Aortic Position: Determinants of Early and Late Outcome A. Ruchan Akar, MD, Adam Szafranek, MD, Christos Alexiou, FRCS, Robert Janas, MD, Marek J. Jasinski, MD, Justiaan Swanevelder, FRCA, and Andrzej W. Sosnowski, FRCS Departments of Cardiothoracic Surgery and Anesthesiology, University Hospitals of Leicester, Glenfield Hospital, Leicester, United Kingdom Background. Whether to perform a stentless aortic valve replacement (AVR) is not well established. Our aim was to determine the outcome after AVR with stentless xenograft valves. Methods. Between 1996 and 2001, a total of 404 patients (mean age 70.4 years) underwent a stentless AVR by one surgeon in our unit. Concomitant procedures were performed in 132 patients (33%). Twenty patients (6.4%) had undergone previous AVR. Eleven types of stentless xenograft valves were implanted: Medtronic Freestyle in 221 patients (55%), Shelhigh in 55 (14%), Shelhigh composite conduit in 33 (8%), Sorin in 26 (6%), Cryolife O Brien in 25 (6%), Aortech-Elan in 17 (4%), Edwards Prima in 14 (4%), Toronto SPV in 7 (2%), and other valves in 6 (1%). A subcoronary implantation technique was used in 302 cases (76%), complete root replacement in 62 (15%), and a modified Bentall-De Bono procedure in 33 (8%). Mean follow-up was 19.4 months (range, 1.2 to 60.6 months). Results. Overall hospital mortality was 4.2%. This was 2.4% for isolated AVR, 3.6% for AVR and coronary artery bypass grafting, 5.5% for replacement of two or more valves, and 12% for the modified Bentall procedure. On multiple logistic regression redo cardiac operation (p ), cardiogenic shock (p 0.001), left ventricular ejection fraction less than 0.30 (p 0.01), modified Bentall procedure (p 0.03), and endocarditis (p 0.04) were predictors of in-hospital death. Five-year freedom from thromboembolism, hemorrhage, prosthetic endocarditis, structural valve deterioration, and reoperation was 97%, 99%, 99%, 98%, and 96%, respectively. Kaplan- Meier survival at 5 years was 88%. On Cox regression, cardiogenic shock (p 0.001) and older age (p 0.03) were adverse predictors of survival. At echocardiographic examination within 6 months from the operation, mean aortic valve gradients were 15 6 mm Hg, mm Hg, mm Hg, mm Hg, mm Hg, and mm Hg for valve sizes of 19, 21, 23, 25, 27, and 29 mm, respectively. Conclusions. The availability of several stentless valve designs facilitates the surgical treatment of diverse aortic valve or root diseases with encouraging early and midterm results. Patients requiring concomitant procedures may also benefit from the excellent hemodynamic characteristics of a stentless valve. We consider stentless AVR the treatment of choice for patients older than 60 years and those having small aortic roots. (Ann Thorac Surg 2002;74:1450 8) 2002 by The Society of Thoracic Surgeons Surgical treatment of symptomatic aortic valve disease improves symptoms and life expectancy [1]. The severity of aortic disease and left ventricular hypertrophy, defined by different indexes for left ventricular mass, are important determinants of survival after aortic valve replacement (AVR) [2]. Left ventricular hypertrophy, in particular, confers as much as a threefold greater risk of mortality in patients with or without coronary artery disease [3]. Residual transvalvular gradients and the effective orifice area of the aortic valve substitute affect the degree of regression of the left ventricular mass and geometry after AVR [4] and may, therefore, impact on the outcome. The first stentless aortic valve to be used for valve replacement was the allograft, first implanted in 1962 [5]. Since the introduction of stentless bioprosthesis in 1987 Accepted for publication June 5, Address reprint requests to Dr. Sosnowski, Department of Cardiothoracic Surgery, Glenfield Hospital, Groby Rd, Leicester, LE3 9QP, UK; kardio@stayfree.co.uk. by David and colleagues [6], several series were reported worldwide with stentless porcine valves in the aortic position [7, 8]. Stentless design demonstrates superior hemodynamics compared with stented valves, with normal or near-normal hemodynamics at rest [9, 10] and during exercise [11]. Their insertion, however, may be technically more demanding and requires longer ischemic time than stented valves. Since 1996 we have favored the use of a stentless bioprosthesis in patients older than 65 years of age requiring isolated AVR. Encouraged by our early experience we have performed a stentless AVR also in those requiring additional cardiac or aortic procedures. Initially one type of valve was used, but we have subsequently selected a stentless valve design according to the patients aortic valve or root disease. The purpose of this study was to describe our experience with various stentless bioprosthetic valves in the aortic position and to identify determinants of operative mortality and survival by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)

2 Ann Thorac Surg AKAR ET AL 2002;74: STENTLESS XENOGRAFTS IN AORTIC POSITION 1451 Table 1. Types and Sizes of Stentless Valves Implanted Type of Stentless Valves Number % Fig 1. Number of stentless aortic valve replacements by year. Patients and Methods Between 1996 and 2001, a total of 404 patients (201 men and 203 women, mean age, years; range, 25 to 94 years) underwent AVR by one surgeon (A.W.S.) with stentless valves in our unit. During the study period, 43 patients taking part in a prospective randomized study had an AVR with a stented bioprosthesis by the same surgeon. Also, 31 patients received a mechanical valve, and 4 had AVR with a homograft. Patient Selection A stentless valve was implanted in all patients older than 65 years and in younger patients if they specifically requested a bioprosthesis or anticoagulation therapy was contraindicated. Patients undergoing redo AVR for patient prosthesis size mismatch also received a stentless valve. Preoperative Clinical Features Indication for AVR was aortic stenosis in 250 patients (62%), aortic regurgitation in 52 (13%), and mixed aortic valve disease in 102 (25%). Twenty-six patients (6.4%) had undergone previous cardiac operations. These were AVR (n 18), AVR and coronary artery bypass grafting (n 1), coronary artery bypass grafting (n 5), AVR and mitral valve replacement (n 1), and aortic arch operation (n 1). The numbers of stentless AVR performed during the study period are shown in Figure 1, and the Fig 2. Age distribution of patients undergoing aortic valve replacement with stentless valves. Medtronic Freestyle Shelhigh Shelhigh C Sorin stentless Cryolife O Brien Aortech Elan Edwards Prima Toronto SPV Tissuemed P Conforma Tissuemed Size of Stentless Valves Number % 19 mm mm mm mm mm mm 17 4 age distribution of patients is shown in Figure 2. Fiftyfour patients (13%) were younger than 60 years. Two hundred forty-three patients (60%) were in New York Heart Association functional classes III and IV. The left ventricular function was poor (ejection fraction 30%) in 50 patients and moderately impaired (ejection fraction 30% to 50%) in 156 (39%). Eleven patients (2.7%) were in cardiogenic shock. Sinus rhythm was present in 322 (80%) and atrial fibrillation in 76 (19%). Five patients (1.2%) had a permanent pacemaker, and 1 patient had an automatic implantable cardioverter defibrillator. Nineteen patients (4.7%) experienced transient ischemic attacks more than 6 months, and 10 (2.5%) within 6 months, before the operation. Prosthetic valve infection was present in 7 patients (1.7%) and native valve endocarditis in 4 (1%). Eight patients underwent an AVR for patient prosthesis mismatch. Other preoperative risk factors were diabetes mellitus in 47 patients (12%), hypertension in 156 (39%), myocardial infarction in 31 (8%), hypercholesterolemia in 99 (25%), chronic obstructive pulmonary disease in 62 (15%), renal failure in 15 (4%), peripheral vascular disease in 24 (6%), and obesity (body mass index 35) in 17 (4%). Surgical Procedures The operations were elective in 266 patients (66%), urgent in 109 (27%), and emergency or salvage in 29 cases (7%). Table 1 shows the type and size of the stentless valves implanted. The types of aortic valve and root disease encountered are shown in Table 2. Eleven different models of stentless valves were used (Table 1). The mean valve size was mm (range, 19 to 29 mm), with 106 patients (26%) receiving 19-mm or 21-mm valve size. Three implantation techniques were used: (1) subcoro-

3 1452 AKAR ET AL Ann Thorac Surg STENTLESS XENOGRAFTS IN AORTIC POSITION 2002;74: Table 2. Aortic Valve and Aortic Root Disease a Type of Disease Number % Degenerative calcification Congenital 35 9 Rheumatic disease 8 2 Annulo-aortic ectasia Myxomatous degeneration Active endocarditis Prosthetic valve infection Paraprosthetic leak Prosthetic valve thrombosis Patient prosthesis size mismatch Aortic dissection a Total of 404 patients. nary valve replacement in 309 patients (76%), (2) complete root replacement in 62 patients (15%), and (3) modified Bentall-De Bono procedure in 33 patients (8%). The type of the implantation technique was tailored to the underlying disease (Table 2). A complete aortic root replacement was preferred if there was evidence of aortic root disease, mainly calcification, and in cases in which valve distortion was likely. If the ascending aorta was involved (aneurysm or dissection) a modified Bentall- DeBono procedure was used. A subcoronary implantation was performed in patients with isolated aortic valve disease. Myocardial protection was provided with intermittent cold crystalloid cardioplegia and moderate hypothermia (28 to 32 C). A transverse aortotomy, 5 mm above the sinotubular junction, was performed after aortic crossclamping. After valve excision and extensive decalcification of the annulus, the sizes of the annulus and the sinotubular junction were measured. The decision to perform a root reconstruction or ascending aorta replacement was made intraoperatively under the guidance of transesophageal echocardiography. During the last 2 years, intraoperative transesophageal echocardiography was used for all stentless AVR procedures. During stentless valve implantation care was taken to achieve optimal orientation of the commissures and coronary ostia. Twelve AVR procedures (3%) were performed by trainees under consultant supervision. One hundred nine patients (27%) had concomitant myocardial revascularization (average, 1.7 grafts per patient; median, 2; range, 1 to 4). Other concomitant procedures were carotid endarterectomy (n 16), left ventricular aneurysmectomy (n 1), mitral valve repair or replacement (n 13), tricuspid valve repair (n 5), repair of atrial septal defect (n 6), aortic root enlargement (n 2), and septectomy (n 3). Study Valves The types and the stentless valves implanted and the corresponding manufacturers were Medtronic Freestyle Stentless Xenograft (Medtronic, Minneapolis, MN) [12], Shelhigh No-React Stentless Bioprosthesis NR 2000 and No-React Composite Valve Conduits (Shelhigh, Inc, Millburn, NJ) [13], Sorin Pericarbon Stentless valve (Sorin Biomedica, Saluggia, Italy) [14], Cryolife-O Brien Composite Aortic Stentless Xenograft Model 300 (Cryolife International, Atlanta, GA) [15], AorTech Freesewn Porcine Elan (Tech Europe Ltd, Swillington, Leeds, UK), Edwards-Prima Plus Stentless bioprosthesis Model 2500P (Baxter Healthcare Corporation, Edwards CVS Division, Irwin, CA) [16], Toronto SPV (St. Jude Medical, St. Paul, MN) [17], Conforma Stentless Model 2000 (developed by WD Hancock and manufactured by Heartline Medical, Leeds, UK), and Tissuemed Stentless pulmonary valve and aortic/pulmonary roots (Tissuemed Ltd, Swillington). Follow-Up Clinical, operative, and follow-up data were recorded prospectively in a computerized database. Additional information was obtained by reviewing the medical records and by contacting the patients general practitioners or families as required. All survivors were seen in the outpatient clinic 3 and 6 months postoperatively and annually thereafter. A Doppler echocardiography was performed before discharge from the hospital and during each outpatient visit. Aortic valve performance was assessed for competence and mean and peak pressure valve gradients. To assess the severity of aortic regurgitation, the ratio of regurgitant of jet width (height; Perry index) to left ventricular outflow tract was used. Also, the length and the area of the regurgitant jet in relation to the left ventricular cavity were studied. The aortic regurgitation was graded as mild, moderate, and severe when the Perry index was 0.1 to 0.3, 0.3 to 0.6, and more than 0.6, respectively. A trace of regurgitation was considered to be present when this index was less than 0.1. Postoperatively patients were given antiplatelet therapy. Warfarin was administered to those who were in atrial fibrillation or who underwent concomitant valve procedures as indicated. The mean follow-up was 20.2 months (range, 1.2 to 60.6 months), and it was 100% complete. Statistical Analysis Valve-related complications are reported according to the Guidelines of the Ad Hoc Liaison Committee of The Society of Thoracic Surgeons and The American Association of Thoracic Surgery [18]. Categorical variables are presented as percentages and continuous variables as means ( SD). The significance of 38 variables on operative mortality and survival was tested with univariate and multivariate analysis (Appendix). UNIVARIATE ANALYSIS. Categorical variables were compared with 2 test or Fisher s exact test as appropriate, and means were compared with unpaired Student s t test. Freedom from valve-related events and survival were calculated with Kaplan-Meier ( SEM) product limit method, and the resulting curves were compared with log rank test. Because there were not significant differences in the outcome for different types of valves, the constructed Kaplan- Meier curves represent all study patients.

4 Ann Thorac Surg AKAR ET AL 2002;74: STENTLESS XENOGRAFTS IN AORTIC POSITION 1453 Table 3. Mean Systolic Gradients of Commonly Implanted Stentless Valves Within 6 Months of Operation a Valve Size (mm) Freestyle No. Shelhigh Suprastentless No. Shelhigh Composite No. Sorin No. O Brien No. Elan No. Edwards Prima No. Toronto SPV No (15) (2) (51) (12) (4) 13 (1) (4) (8) 15 (1) 15 (1) (64) (11) (4) (12) 10 1 (2) (4) (4) (59) (18) (10) (9) (10) (4) (2) 8.5 (1) (24) (11) (10) 9.7 (1) (6) 8 (1) (3) (3) (5) (3) (3) (1) (2) a Results for gradients are mean standard deviation, reported in millimeters of mercury. MULTIVARIATE ANALYSIS. The variables that attained a probability value of less than 0.1 on univariate analysis were entered into multiple logistic regression and Cox proportional hazards models. A probability value less than 0.05 was considered significant in both univariate and multivariate analysis. Data were subjected to quantitative and qualitative analysis using the biostatistical capabilities of the Patient Analysis and Tracking Systems (PATS; Axis Clinical Software, Inc, Portland, OR). The statistical software SPSS PC (version 8.0, SPSS Inc, Chicago, IL) was used for final data analysis. Results Early Outcome Overall operative mortality was 4.2% (17 patients). Causes of death were cardiac failure (n 8), ventricular arrhythmias (n 2), thromboembolism (n 2), multisystem organ failure (n 2), respiratory failure caused by fibrosing alveolitis (n 1), acute pancreatitis (n 1), and adult respiratory distress syndrome (n 1). Significant univariate factors for operative mortality were New York Heart Association functional classes III and IV (p 0.04), nonelective operation (p 0.008), poor left ventricular function (p 0.004), modified Bentall procedure (p ), endocarditis (p ), cardiogenic shock (p ), and redo AVR (p ). On multiple logistic regression analysis, redo AVR (p 0.004), cardiogenic shock (p 0.008), ejection fraction less than 30% (p 0.02), and endocarditis (p 0.05) were the significant risk factors for operative death. Fifty-two patients (12.9%) required inotropic support, 81 (20%) were ventilated for more than 24 hours postoperatively, 19 (4.7%) had renal replacement therapy, 10 (2.5%) needed intraaortic balloon support, 12 (3%) were reopened for bleeding, 4 required rewiring of sternum for mediastinitis, and 21 (5%) had a permanent pacemaker inserted before discharge from the hospital. The mean length of stay in the intensive care unit was days and mean hospital stay was days. Hospital survivors had improved New York Heart Association functional class status 6 weeks after their operation. Compared with their preoperative status, New York Heart Association functional class was improved in 347 patients (85.8%), remained the same in 38 (9.4%), and worsened in 2 patients (0.5%). Doppler Echocardiography Assessment Postoperative transthoracic echocardiography was performed at a mean of days after AVR. At echocardiographic examination within 6 months of the operation, the mean aortic valve gradients were 15 6mmHg, mm Hg, mm Hg, mm Hg, mm Hg, and mm Hg for 19, 21, 23, 25, 27, and 29 mm valve sizes, respectively. Mean systolic gradients of commonly implanted stentless valves are shown in Table 3. Grade 1 (trace) aortic regurgitation was present in 37 patients (9%), mild aortic regurgitation (grade 2) in 20 (4.9%), and moderate aortic regurgitation (grade 3) in 3 patients (0.7%), who developed perivalvular leak. Six patients had significantly high peak transvalvular gradients ( 50 mm Hg) early postoperatively, which had regressed by the end of first year in all but 1 patient. One of these patients required redo AVR with aortic root enlargement and homograft replacement owing to patient prosthesis mismatch associated with endocarditis. Late Outcome There have been 15 late deaths. A prosthesis-related death occurred in 5 patients as a result of structural valve deterioration (n 2), endocarditis (n 1), intracranial hemorrhage (n 1), and upper gastrointestinal bleeding (n 1). Six patients died of cardiac failure or ischemia; 4 of them had coronary artery bypass grafting at the time of the stentless AVR, and 1 had nongraftable diseased coronary arteries. There were four nonvalve-related deaths, caused by lung cancer, trauma, viral pneumonia, and emphysema. Kaplan-Meier overall survival at 5 years was 88% 2% (Fig 3). Valve-Related Complications At the most recent follow-up, 372 patients were alive, with 370 having their original stentless valves. HEMORRHAGE. Two patients experienced gastrointestinal bleeding, and 1 of them died. Another patient had a fatal intracranial hemorrhage. All 3 patients involved were taking warfarin at the time of the bleeding event. Kaplan- Meier 5-year freedom from hemorrhage was 99% 1% (Fig 4).

5 1454 AKAR ET AL Ann Thorac Surg STENTLESS XENOGRAFTS IN AORTIC POSITION 2002;74: Fig 3. Kaplan-Meier overall survival at 5 years (88% 2%). THROMBOEMBOLISM. There were four thromboembolic episodes. Two were major with permanent deficit, described under operative mortality, and two were minor with good functional recovery. Kaplan-Meier 5-year freedom from thromboembolism was 97% 1% (Fig 5). ENDOCARDITIS. Two patients who had received a 19-mm and 23-mm Freestyle valve developed endocarditis with Staphylococcus aureus and Enterococci 2 and 24 months after their operation, respectively. The patient who exhibited early infection of his aortic prosthesis died after a redo AVR with a homograft and aortic root enlargement. The patient who acquired late prosthetic endocarditis also had an invasive aortic root abscess and died after a second repeat valve and root replacement. Kaplan-Meier 5-year freedom from endocarditis was 99% 1% (Fig 6). Fig 5. Kaplan-Meier freedom from thromboembolic events at 5 years (97% 1%). STRUCTURAL VALVE DETERIORATION. Primary structural failure was diagnosed at autopsy 12 and 18 months after AVR in 2 patients who received O Brien valves. The first patient became unwell 10 months postoperatively, exhibiting symptoms consistent with endocarditis. An echocardiogram showed aortic valve vegetations, but the blood cultures gave negative results. Culture-negative prosthetic valve endocarditis was suspected, and after favorable initial response to empiric antibiotic treatment the patient was discharged from the hospital. He died suddenly 2 months later at home, and at postmortem examination the aortic valve was severely calcified. The second patient had chronic renal failure and was renal dialysis dependent and scheduled for a renal transplant operation soon after her AVR, but for reasons unclear to us a renal transplant procedure was not performed. A magnetic resonance image at 6 months and an echocardiogram at 12 months postoperatively demonstrated excellent hemodynamic function of the prosthesis with no sign of calcification. However, 18 months after her AVR she was readmitted in a peripheral hospital in cardiogenic shock owing to critical aortic stenosis and died during transfer to our unit. At postmortem examination there was extensive calcification of the entire aortic prosthetic device (leaflets and wall). Actuarial freedom from structural valve deterioration was 98% 1% at 5 years (Fig 7). NONSTRUCTURAL VALVE DETERIORATION AND REOPERATION. There has been no significant hemolytic event in this series. A perivalvular leak was diagnosed in 3 patients 5 Fig 4. Kaplan-Meier freedom from hemorrhage at 5 years (99% 1%). Fig 6. Kaplan-Meier freedom from endocarditis at 5 years (99% 1%).

6 Ann Thorac Surg AKAR ET AL 2002;74: STENTLESS XENOGRAFTS IN AORTIC POSITION 1455 Fig 7. Kaplan-Meier freedom from structural valve deterioration at 5 years (98% 1%). and 8 months and 2 years postoperatively. Implanted bioprostheses originally were Toronto SPV, O Brien stentless, and Shelhigh stentless valves. Two of these patients received new stentless valves, and the third patient underwent repair of his paravalvular leak. There was no evidence of structural valve deterioration in any of these valves. A total of 7 patients underwent reoperations as already described. Kaplan-Meier 5-year freedom from reoperation after AVR was 96% 1% at 5 years (Fig 8). Comment Aortic valve replacement has been considered a form of palliative therapy in that a prosthetic valve with its attendant complications is substituted for a diseased native valve [19]. Age-corrected survival after AVR is excellent for patients older than 65 years and is similar to that of the normal population of that age [20]. The indications for the use of stentless valve are, in principle, the same as those for stented bioprosthesis. Stentless bioprosthesis offered improved hemodynamic performance [21] and survival [22], and it is well documented that elimination of the stent and sewing ring Fig 8. Kaplan-Meier freedom from reoperation at 5 years (96% 1%). allows implantation of a larger bioprosthesis into an individual annulus [23]. Implantation of a stentless valve, however, is technically more demanding, and there are concerns that the longer ischemic times required might increase the operative risk [23]. It is also known that the function and durability of a stentless valve can be operator dependent as even minor errors in valve sizing and orientation can distort leaflet coaptation with adverse effects. The mortality rates observed in the present study are acceptable, being comparable to those reported recently by The Society of Thoracic Surgeons National Cardiac Surgery Database [24]. Ischemic times did not appear to increase hospital mortality in our series, even among patients undergoing complex additional procedures. It is likely that good hemodynamic performance of the stentless valve compensates for longer ischemic times with inherent risk of left ventricular impairment in the early postoperative period. We found that repeat cardiac operation, cardiogenic shock, poor left ventricular function, modified Bentall procedure, and endocarditis increase the risk of hospital mortality. This suggests that the early outcome after stentless AVR could further improve if the patients were referred for operation before cardiac failure or cardiogenic shock occurs. The availability of different stentless valve substitutes offers the surgeon flexibility in managing diverse aortic root diseases. Every valve has design-specific advantages and drawbacks for surgical implantation, and the valve substitute should be individualized to the patient profile. Throughout the study period we used 11 different types of stentless aortic valves. We prefer to use the full root Freestyle or Edwards Prima valves especially for patients with an asymmetric aortic root, a heavily calcified ascending aorta, a dilated sinotubular junction, or after septectomy in the presence of calcification extending toward the septum or septal hypertrophy. The possibility of scalloping these valves, after completion of the inflow suture line, helps to achieve optimal leaflet coaptation, avoiding distortion and regurgitation. The Sorin Pericarbon Stentless valve, which is the only stentless pericardial valve in this report, can be inserted with the inversion technique. The Cryolife-O Brien Stentless valve has the advantage of supraannular implantation using a single suture line, leaving the left ventricular outflow tract, including the aortic annulus, free of any foreign material. We prefer to use subcoronary implantation to root replacement technique in younger patients to avoid technical difficulties of the redo aortic root replacement procedures. The Shelhigh Stentless Composite Valve is the only complete biologic conduit available in the market for replacement of the entire ascending aorta, aortic root, and aortic valve. Implantation of the Shelhigh Stentless Composite Valve is surgeon-friendly, improves hemostatic suture line, and offers the hemodynamic advantages of a stentless valve including the avoidance of long-term anticoagulation. Freedom from valve-related complications and survival at 5 years were satisfactory in this series and compare favorably with those quoted in earlier studies

7 1456 AKAR ET AL Ann Thorac Surg STENTLESS XENOGRAFTS IN AORTIC POSITION 2002;74: [25]. Hvass and O Brien [26] reported no structural valve deterioration with CryoLife-O Brien porcine valves in 366 patients at 5 years. We encountered fatal structural valve deterioration in 2 patients 12 and 18 months after AVR, even though both patients postoperative echocardiographic studies showed no evidence of aortic regurgitation and low transvalvular gradients. Epidemiological evidence suggests that treatment of aortic valve disease should include reversal of left ventricular hypertrophy [27, 28]. The regression of myocardial hypertrophy is a process that occurs for many years after correction of the primary hemodynamic stress. Del Rizzo and associates [4] demonstrated that effective orifice area and residual gradients improve with time, with most of the change occurring in the first 6 months as a result of ventricular remodeling with the Toronto SPV valve. This improvement occurs even though the prosthetic valve itself constitutes a relative and persistent, although not progressive, stenotic obstruction to left ventricular ejection. In a subgroup of 20 patients in each arm (stented versus stentless valves), we confirmed the early regression of left ventricular hypertrophy using echocardiography and magnetic resonance imaging in a randomized study (unpublished data). The interpretation of the unique echocardiographic appearance of each stentless valve (11 in this series) can be challenging [29]. We routinely perform intraoperative transesophageal echocardiography and consider normal leaflet mobility, unrestricted leaflet motion during systole, and freedom from significant aortic regurgitation as important determinants for stentless valve function. Intraoperative valve gradients are measured for research purposes only and do not influence our operative decisions. This is because early postpump gradient measurements tend to overestimate real transvalvular gradients owing to the variability of cardiac output, inotropic support, myocardial stunning, and left ventricular hypertrophy. During the first postoperative week a transthoracic echocardiogram is performed to measure transvalvular gradients, evaluate left ventricular function, and exclude the presence of pericardial effusion before hospital discharge. In agreement with previous reports, all stentless valves used in this study exhibited low transvalvular gradients on echocardiography. Some differences in the mean aortic valve gradients were observed (Table 3), but these might have been caused by chance alone (p 0.05). Considerable evidence supports the use of stentless xenografts in elderly patients for AVR especially if they have small aortic roots. Indications for stentless AVR in our practice include age more than 60 years and patients with reduced life expectancy owing to severely compromised ventricular function, coronary artery disease, or other systemic disease. Stentless valves can also provide a good solution in cases in which use of anticoagulation and mechanical prosthesis is contraindicated. There is no ideal valve substitute for AVR. However, the availability of several stentless valve designs facilitates the management of diverse aortic valve or root diseases with encouraging early and midterm results. Patients requiring concomitant procedures should not be denied the recognized hemodynamic benefits of a stentless valve. The choice of the most appropriate stentless prosthesis depends on the aortic disease, the clinical condition of the patient, and the preference of the surgeon. In our view, the Freestyle and Edwards Prima are the most versatile valves; they can be used in all types of aortic root disease and can be inserted with various techniques such as total root replacement, subcoronary implantation or modified subcoronary implantation. We also find that the Shelhigh valve is the easiest to implant from a technical viewpoint, and its use may be advantageous in cases in which a reduced aortic cross-clamp time is highly desirable. Improved symptomatic status, low incidence of valverelated complications, and survival data strongly support the use of stentless xenografts in elderly patients. Longer follow-up studies will show whether an extension of the indications for the use of stentless xenografts in younger age groups is warranted. We thank Sue Holleworth for her work, without which this article would not have been possible. We also thank Lorraine Ricketts and the other staff of the Department of Echocardiography at Glenfield Hospital for expert assistance. References 1. Bonow RO, Carabello B, de Leon AC Jr, et al. Guidelines for the management of patients with valvular heart disease: executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Patients with Valvular Heart Disease). Circulation 1998;98: Levy D. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham heart study. N Engl J Med 1990;322: McGiffinD,O Brien M, Galbraith A, et al. An analysis of risk factors for death and mode-specific death after aortic valve replacement with allograft, xenograft, and mechanical valves. J Thorac Cardiovasc Surg 1993;106: Del Rizzo DF, Goldman BS, Christakis GT, David TE. Hemodynamic benefits of the Toronto Stentless Valve. J Thorac Cardiovasc Surg 1996;112: Ross DN. Homograft replacement of the aortic valve. Lancet 1962;2: David TE, Ropchan GC, Butany JW. Aortic valve replacement with stentless porcine bioprostheses. J Card Surg 1988; 3: Goldman BS, David TE, Wood JR, et al. Clinical outcomes after aortic valve replacement with the Toronto stentless porcine valve. Ann Thorac Surg 2001;71(Suppl):S Luciani GB, Bertolini P, Vecchi B, Barozzi L, Casali G, Mazzucco A. Stentless aortic xenograft valve replacement: lessons learned after 300 implants. Semin Thorac Cardiovasc Surg 1999;11: Bach DS, David T, Yacoub M, et al. Hemodynamics and left ventricular mass regression following implantation of the Toronto SPV stentless porcine valve. Am J Cardiol 1998;82: Dumesnil JG, Leblanc MH, Cartier PC, et al. Hemodynamic features of the freestyle aortic bioprosthesis compared with stented bioprosthesis. Ann Thorac Surg 1998;66(Suppl): S Eriksson MJ, Rosfors S, Radegran K, Brodin LA. Effects of exercise on Doppler-derived pressure difference, valve re-

8 Ann Thorac Surg AKAR ET AL 2002;74: STENTLESS XENOGRAFTS IN AORTIC POSITION 1457 sistance, and effective orifice area in different aortic valve prostheses of similar size [see comments]. Am J Cardiol 1999;83:619 22, A Sintek CF, Pfeffer TA, Kochamba GS, Yun KL, Fletcher AD, Khonsari S. Freestyle valve experience: technical considerations and mid-term results. J Card Surg 1998;13: Abolhoda A, Yu S, Oyarzun R, McCormick JR, Bogden JD, Gabbay S. Calcification of bovine pericardium: glutaraldehyde versus No-React biomodification. Ann Thorac Surg 1996;62: Rubay JE, El Khoury G, Buche M. Aortic valve replacement with the stentless bovine pericardial valve: early experience with the Sorin valve. In: Huysmans HA, David TE, Westaby S, eds. Stentless bioprostheses, 2nd ed. Oxford: Isis Medical Media, Herndon, VA, 1999: O Brien MF. The Cryolife-O Brien composite aortic stentless xenograft: surgical technique of implantation. Ann Thorac Surg 1995;60(Suppl):S Konertz W, Weyand M, Sidiropoulos A, Schwammenthal E, Breithardt G, Scheld HH. Technique of aortic valve replacement with the Edwards stentless aortic bioprosthesis Eur J Cardiothorac Surg 1992;6: David TE, Bos J, Rakowski H. Aortic valve replacement with the Toronto SPV bioprosthesis. J Heart Valve Dis 1992;1: Edmunds LHJ, Grunkemeier GL, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg 1996;62: Hammermeister KE, Sethi GK, Henderson WG, Oprian C, Kim T, Rahimtoola S. A comparison of outcomes in men 11 years after heart-valve replacement with a mechanical valve or bioprosthesis: Veterans Affairs Cooperative Study on Valvular Heart Disease. N Engl J Med 1993;328: Lindblom D, Lindblom U, Qvist J, Lundstrom H. Long-term relative survival rates after heart valve replacement. J Am Coll Cardiol 1990;15: Fries R, Wendler O, Schieffer H, Schafers HJ. Comparative rest and exercise hemodynamics of 23-mm stentless versus 23-mm stented aortic bioprostheses. Ann Thorac Surg 2000; 69: Westaby S, Horton M, Jin XY, et al. Survival advantage of stentless aortic bioprostheses. Ann Thorac Surg 2000;70: Berg GA, McLaughlin KE, Akar R, Pathi V, MacArthur KJ. A three year experience with the Toronto stentless porcine valve. Ann Thorac Cardiovasc Surg 1998;4: Jamieson WR, Edwards FH, Schwartz M, Bero JW, Clark RE, Grover FL. Risk stratification for cardiac valve replacement. National Cardiac Surgery Database. Ann Thorac Surg 1999; 67: Doty DB, Cafferty A, Cartier P, et al. Aortic valve replacement with Medtronic Freestyle bioprosthesis: 5-year results. Semin Thorac Cardiovasc Surg 1999;11: Hvass U, O Brien MF. The stentless Cryolife-O Brien aortic porcine xenograft: a five-year follow-up. Ann Thorac Surg 1998;66(Suppl):S Gelsomino S, Frassani R, Morocutti G, et al. Left ventricular mass regression after aortic valve replacement with CryoLife-O Brien stentless aortic bioprosthesis. J Heart Valve Dis 2001;10: Gelsomino S, Frassani R, Morocutti G, et al. Time course of left ventricular remodeling after stentless aortic valve replacement. Am Heart J 2001;142: Bach DS. Echocardiographic assessment of stentless aortic bioprosthetic valves. J Am Soc Echocardiogr 2000;13: Appendix The following variables were examined in multivariable analyses. Demography: Age (years), sex, body surface area (m 2 ), and body mass index (kg/m 2 ). Preoperative status: New York Heart Association functional class, severity of angina, nonelective operation, redo cardiac operation, left ventricular function, pulmonary artery pressures (systolic 60 mm Hg), history of myocardial infarction, cardiogenic shock at the time of operation, recent congestive cardiac failure, extent of coronary artery disease, presence of preoperative atrial fibrillation, and presence of complete heart block. Noncardiac comorbidity: Smoking, hypercholesterolemia, diabetes mellitus, hypertension, peripheral vascular disease, previous stroke, chronic obstructive pulmonary disease, renal disease, requirement for renal replacement therapy, history of gastrointestinal bleeding or duodenal ulcer, and hepatic disease. Operation-related variables: Type of stentless valve, type of operation (subcoronary, root, modified Bentall), prosthesis size (mm), indexed prosthesis size (mm/ m 2 ), concomitant coronary artery bypass grafting, cross-clamp time, and cardiopulmonary bypass time. Late postoperative variables: Postoperative New York Heart Association functional class at 6 weeks, postoperative mean transvalvular gradients, residual aortic regurgitation, and years since 1996 when operation took place. INVITED COMMENTARY Dr Akar and colleagues have obviously acquired appreciable expertise in the field of stentless valves. Their series, with results that clearly parallel those of other authors, is one of the largest and probably also the most original because of the variety of stentless valves used. I am also very keen on learning future results from their nascent randomized study. Although only 40 patients were included, follow-up studies, if correctly conducted and expanded, should yield important information differentiating stented and stentless devices. However, despite my appreciation for their standards and direction, I clearly reject the teaching that a large menu of available devices permits optimal matching of a specific valve design and patient s particulars (pathology and precise root anatomy). To young surgeons, this policy may send a misleading and discouraging message that use of stentless devices is highly selective and beyond their ability. My experience is diametrically opposite; the credo is one stentless valve for all, excluding only the few patients who really need aortic root reconstruction (homografts for destructive active endocarditis and a valve by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)