Stentless aortic xenografts were introduced a decade

Transcription

1 Survival After Stentless and Stented Xenograft Aortic Valve Replacement: A Concurrent, Controlled Trial Giovanni Battista Luciani, MD, Gianluca Casali, MD, Stefano Auriemma, MD, Francesco Santini, MD, and Alessandro Mazzucco, MD Division of Cardiac Surgery, University of Verona, Verona, Italy Background. To define the impact of stentless versus stented valve design on survival late after xenograft aortic valve replacement, a retrospective analysis of all consecutive patients operated on between January 1992 and April 2000 was undertaken. Methods. Two hundred ninety-two patients had stented (group 1) and 376 stentless (group 2) xenograft aortic valve replacements. Age was older in group 1 (75 4vs70 7 years, p 0.01), whereas male gender and aortic stenosis were equally prevalent. Advanced New York Heart Association class III-IV (85% vs 78%, p 0.03) and associated procedures (53% vs 41%, p 0.01) were more common in group 1. Aortic cross-clamp (80 28 vs minutes, p 0.01) and bypass (91 56 vs minutes, p 0.01) times were shorter in group 1. Logistic regression and Cox proportional hazard methods were used to define the role of demographic and operative variables on hospital and late survival, freedom from valve-related mortality, and reintervention. Results. Early mortality was higher in group 1 (6.2% vs 2.6%, p 0.02). Smaller aortic anulus (p 0.008), aortic cross-clamp (p 0.03), and coronary disease requiring bypass (p 0.03) were associated with hospital mortality. During follow-up (37 30 vs months, p NS), 66 late deaths were recorded (12% vs 9%, p NS). At 8 years, survival (70 5% vs 81 3%, p 0.01), freedom from cardiac- (85 1% vs 92 3%, p 0.02), and valve-related death (79 5% vs 95 2%, p 0.004) were higher in group 2. Freedom from structural deterioration was similar (92 5% vs 93 3%, p NS), but freedom from reoperation was lower in group 2 (99 1% vs 90 4%, p 0.009). Multivariate analysis showed female gender (p 0.02), age (p 0.03), and smaller valve size (p 0.05) to be associated with late mortality; age (p 0.06) and diagnosis of aortic stenosis (p 0.008) with cardiac mortality; longer intensive care unit stay (p 0.001) and stented xenografts (p 0.05) with valve-related mortality; and younger age (p 0.01) and stentless xenograft (p 0.05) with reoperation. Conclusions. Use of stentless xenografts correlates with better survival and freedom from cardiac- and valverelated mortality than stented valves. However, bias favoring stented valves in older and sicker patients exists. Selective survival advantage of stentless xenograft is confined to valve-related mortality. Stentless valves are more likely to be replaced for dysfunction. (Ann Thorac Surg 2002;74:1443 9) 2002 by The Society of Thoracic Surgeons Stentless aortic xenografts were introduced a decade ago in an attempt to overcome the major shortcomings associated with stented bioprostheses, namely limited durability and suboptimal hemodynamic behavior [1]. Comparative outcome analysis of stentless and stented xenografts has thus far been carried out almost exclusively by retrospective, nonrandom studies [2 7]. Several reasons explain the paucity of prospective, randomized trials, including: the need to master the technique of implantation, the limitations imposed by aortic root pathology (calcification, dilatation, coronary artery anomalies), the necessity to minimize grafting time in the presence of left ventricular dysfunction or associated cardiac disease, and the variety of stentless valve models released into the market. Among the methods to mitigate the bias introduced by retrospective studies, case-match analysis has recently Accepted for publication June 26, Address reprint requests to Dr Luciani, Division of Cardiac Surgery, University of Verona, O. C. M. Piazzale Stefani 1, Verona 37126, Italy; gbluciani@yahoo.com. been proposed. In a work by David and associates, stentless aortic xenografts have been associated with durability similar to stented valves, but with greater survival and freedom from adverse cardiac events [8]. To account for these observations, a correlation has been suggested between the property of stentless valves to enhance left ventricular mass regression [9] and the observed decrease in cardiac mortality [8]. In an attempt to validate these findings and to extend them to other models of stentless valves, an analysis of outcome after xenograft aortic valve replacement in 668 consecutive patients has been undertaken. Patients and Methods Patient Population All consecutive patients undergoing aortic valve replacement (AVR) with a porcine xenograft at the University of Verona, between January 1992 and April 2000, were included in the present nonrandomized, concurrent control trial. Selection criteria for a bioprosthesis during the 2002 by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)

2 1444 LUCIANI ET AL Ann Thorac Surg STENTLESS VERSUS STENTED XENOGRAFTS 2002;74: study interval were: (1) age older than 65 years; (2) contraindications to oral anticoagulant therapy; and (3) deliberate request of a biological valve by the patient. When the decision to replace the aortic valve with a xenograft was reached, the ultimate choice between a stented (group 1) or freehand stentless (group 2) valve was left to the surgeon. The only model of stented porcine bioprosthesis used during the study period was the Hancock II valve (Medtronic Inc., Minneapolis, MN). Five different models of stentless porcine bioprosthesis were used during the study interval, starting in July Three models, including the Biocor PSB valve (Biocor Industria e Pesquisa Ltda, Belo Horizonte, MG, Brazil), the Toronto SPV valve (St. Jude Medical, Inc., St. Paul, MN), and the Cryolife-O Brien valve (Cryolife, Inc., Atlanta, GA), comprised 99% of the stentless bioprostheses employed. The remaining two models (Baxter Prima Plus; Baxter Health Care Inc., Irvine, CA, and Medtronic Freestyle; Medtronic Inc., Minneapolis, MN) were utilized in a minority of patients. The choice of the type of stentless valve to implant was done by the surgeon in a nonrandom fashion. Operative Technique Porcine xenograft valve replacement was performed with the aid of moderately hypothermic cardiopulmonary bypass, during a period of cold myocardial ischemia. Before 1994, crystalloid cardioplegia was used for myocardial protection. Thereafter, blood cardioplegia was routinely employed. The technique of stentless valve implant was freehand, subcoronary grafting using inflow and outflow suturelines for all xenograft models, except for Cryolife-O Brien valves. The latter were implanted in subcoronary, supraanular position with a single sutureline. Stented xenografts were anchored to the native aortic anulus by means of interrupted pledget-reinforced sutures. Anticoagulant Therapy Oral anticoagulation was adopted for patients undergoing stented AVR during the first 3 postoperative months and discontinued thereafter. No anticoagulation was used in patients having stentless AVR. Routine oral antiplatelet therapy was adopted in patients with associated vascular (ie, coronary or carotid artery) lesions. All patients with chronic atrial fibrillation were treated with lifelong anticoagulant therapy regardless of the type of aortic valve prosthesis. Follow-Up Methods Cross-sectional patient follow-up was completed between February and April 2000 by means of in-hospital clinical assessment and telephone interview, carried out by nonblinded medical personnel. Four (0.6%) patients were lost to follow-up investigation. Statistical Analysis Continuous variables were expressed as means SD. Categorical variables were expressed as percentages. Comparison of continuous variables was performed using the two-tailed Student s t test for paired data and comparison of discrete using Pearson 2 test or Fisher s exact, as appropriate. Time-related events were described using the Kaplan-Meier estimate and compared with the log-rank analysis. Significance was inferred at a p End points of the study were survival, survival free from cardiac death, valve-related death structural valve deterioration, and reoperation on the xenograft. Linearized rates were used for repeated valve-related complications and were expressed as percent per patient-years. Because the disparity in baseline demographic and operative variables between the two patient groups may reflect in clinical outcome, multivariate analysis was performed on the overall patient population using logistic regression, to identify predictors of hospital mortality, and Cox proportional hazards method, to identify risk factors for time-related occurrence of events after AVR, including: late mortality, cardiac mortality, valverelated mortality, and reoperation. The variables entered in the analysis are listed in the Appendix. Definitions of the above events were established according to previously recommended guidelines [10]. Results Patient Population Six hundred sixty-eight consecutive patients underwent porcine xenograft AVR during the study period: 292 received stent (group 1) and 376 received stentless valves (group 2) (Table 1). Comparison of baseline demographic and operative variables in the overall population disclosed significant disparities between the two patient groups. Average age was older and prevalence of severe cardiac symptoms (New York Heart Association [NYHA] class III or IV) greater in patients receiving stented valves. In addition, prosthetic valve dysfunction as an indication to AVR was more common among patients having stented, whereas mixed aortic lesion was more prevalent among those having stentless xenografts. Lastly, duration of aortic cross-clamp and cardiopulmonary bypass was longer among patients receiving stentless xenografts, despite greater prevalence of associated disease requiring concomitant repair among patients with stented xenografts. Pathology of the aortic root and ascending aorta was primarily responsible for the higher prevalence of associated procedures. Survival There were 28 (4.2%) hospital deaths: 18 (6.2%) among patients having stented and 10 (2.7%) among those having stentless AVR (p 0.02). Cardiac death was the leading cause of hospital mortality, and significantly more common in group 1 (16 [5.5%] vs 6 [1.6%], p 0.009). Logistic regression analysis showed smaller aortic anulus, duration of aortic cross-clamp, and coronary disease requiring bypass to be predictive of greater hospital mortality (Table 2). Advanced age and diagnosis of aortic stenosis were marginally significant. Average follow-up in the two groups was comparable (37 30 vs

3 Ann Thorac Surg LUCIANI ET AL 2002;74: STENTLESS VERSUS STENTED XENOGRAFTS 1445 Table 1. Demographic and Operative Variables of Entire Population Group 1 (Stented) Group 2 (Stentless) p No Age (years) Male gender (patients) 146 (50%) 195 (52%) 0.6 NYHA class III IV (patients) 248 (85%) 294 (78%) 0.03 Aortic stenosis (patients) Aortic regurgitation (patients) Mixed lesion (patients) Prosthetic valve dysfunction (patients) Type of xenograft valve (patients) Hancock II, 292 Toronto SPV, 164 Biocor PSB, 106 Cryolife-O Brien, 101 Other, 5 Average valve size (mm) Aortic cross-clamp time (min) Cardiopulmonary bypass time (min) Associated procedure: (patients) 154 (53%) 153 (41%) 0.01 CABG (patients) MVR/repair (patients) Aortic operation (patients) Other (patients) NYHA New York Heart Association; CABG coronary artery bypass grafting; MVR mitral valve replacement months, p 0.6). There were 66 (10.3%) late deaths: 32 (11.7%) in group 1 and 34 (9.3%) in group 2 (p 0.4). Cardiac death was again the leading cause of late mortality and significantly more prevalent among patients having stented xenografts (23 [8.4%] vs 13 [4.1%], p 0.01). Actuarial survival at 8 years was higher for patients receiving stentless xenografts (70 5% vs 81 3%, p 0.01) (Fig 1). Late freedom from cardiac death was also higher in group 2 (85 1% vs 92 3%, p 0.02). Cox proportional hazards analysis demonstrated female gender, older age, and smaller valve size to be associated with greater late mortality, and older age and diagnosis of aortic stenosis with greater cardiac mortality (Table 3). Valve-Related Events Overall prevalence of valve-related complications was comparable in the two groups (23 [8.4%] vs 33 [9.0%], p 0.8), as confirmed by similar freedom from any valverelated event at 8 years (75 7% vs 78 6%, p 0.4). However, 8-year freedom from valve-related death was higher in patients having stentless valves (79 5% vs 95 2%, p 0.004) (Fig. 2). Cox analysis disclosed longer duration of intensive care to be directly and stentless xenograft to be inversely correlated with valve-related mortality (Table 3). Linearized rates of hemorrhage (0.1% vs 0.4%/patientyears, p 0.2) and endocarditis (0.3% vs 0.1%/patientyears, p 0.4) were similar; rate of embolism was significantly higher in patients with stented valves (1.6% vs 0.6%/patient-years, p 0.02). Structural valve failure was equally prevalent in the two groups (6 [2.2%] vs 10 [2.7%], p 0.6), as shown by the similar actuarial freedom from valve deterioration at 8 years (92 5% vs 93 3%, p 0.6). Reoperation was performed in 2 (structural deterioration, endocarditis) group 1 patients versus 17 (10 structural deterioration, five nonstructural deterioration, two endocarditis) group 2 patients (p 0.004). Cases of nonstructural failure were Table 2. Predictors of Hospital Mortality by Logistic Regression Analysis Variable Odds Ratio p Value Age (years) Diagnosis (aortic stenosis) Coronary artery disease Aortic anulus size (mm) Aortic cross-clamp (min) Fig 1. Actuarial survival of 668 patients undergoing xenograft aortic valve replacement between January 1992 and April 2000 at the University of Verona. Dotted line depicts stentless aortic valves; solid line depicts stented valves. Error bars represent SEM.

4 1446 LUCIANI ET AL Ann Thorac Surg STENTLESS VERSUS STENTED XENOGRAFTS 2002;74: Table 3. Predictors of Outcome by Cox Proportional Hazards Method Outcome SE t Exponent Wald Statistics p Late Mortality Gender (female) Age (years) Valve size (mm) Cardiac mortality Age (years) Diagnosis (aortic stenosis) Valve-Related Mortality Intensive care stay (days) Stentless xenograft Reoperation Age (years) Stentless xenograft Type of xenograft a a Cryolife-O Brien stentless valve. due to technical cause in 4 patients (all with the Cryolife- O Brien valve) and to pannus formation in 1. Freedom from reoperation on the xenograft was thus significantly lower in patients with stentless valves (99 1% vs 90 4%, p 0.009), due to higher prevalence of nonstructural failure. Cox analysis showed younger age, stentless xenograft, and, among the latter, Cryolife-O Brien valve, to be predictive of reoperation on the xenograft (Table 3). There were no deaths in patients having replacement of a stented valve, and three (17.6%) among patients having replacement of a stentless valve, all operated on at nearby hospitals. Mortality for emergent reoperation was significantly higher than for elective (2/3 [66.7%] vs 1/14 [7.1%], p 0.01). Functional Status Follow-up clinical conditions were equally rewarding with 224 (92.6%) of group 1 and 299 (90.6%) of group 2 patients in NYHA class I or II (p 0.2). Fig 2. Actuarial freedom from valve-related death of 668 patients undergoing xenograft aortic valve replacement between January 1992 and April 2000 at the University of Verona. Dotted line depicts stentless aortic valves; solid line depicts stented vales. Error bars represent SEM. Comment In the present nonrandomized, concurrent control trial, stentless xenografts have been associated with overall survival, and survival free from cardiac- and from valverelated death superior to stented. On the contrary, freedom from structural valve deterioration has proved comparable with the two types of bioprostheses. Multivariate analysis, used to attenuate the influence of confounding variables, has suggested the advantages of stentless AVR to be confined to freedom from valve-related mortality. Ever since it has been observed that durability of stentless valves may indeed not prove longer than stented [8], survival late after operation has become the primary focus of clinical research [3, 4, 6, 7]. A wealth of data has shown distinct hemodynamic advantages of stentless xenografts, as relates to timing and extent of regression of left ventricular hypertrophy early after AVR [9, 11, 12]. Although uncertainty remains about the possibility that such changes might be maintained over time, great effort has been devoted to the demonstration that more physiologic behavior of the valve would translate into enhanced survival [2 8]. To date, however, conclusive evidence on survival advantages of stentless xenografts over stented is not available due to absence of long-term prospective, randomized trials. Explanations for the paucity of prospective studies include the technical demands of stentless valve surgery, the restrictions imposed by pathology of the aortic root (dilatation, calcification, anomalies of coronary ostia), the concern over longer grafting time, particularly in the presence of a failing left ventricle, and the great variety of stentless valve models. Several retrospective clinical observations comparing stentless xenografts with stented have shown superior survival and freedom from valve-related adverse events with the former [2 4, 6, 7]. The results in the present work are therefore in line with previous studies. Similar to prior work [2], however, bias in the selection of

5 Ann Thorac Surg LUCIANI ET AL 2002;74: STENTLESS VERSUS STENTED XENOGRAFTS 1447 patients assigned to the two valve substitutes is evident in the different profile of the two populations. Patients receiving stented xenografts are generally older and more commonly affected by advanced cardiac failure and associated disease than those having stentless valves. To reduce the impact of such disparities on survival, David and associates have proposed a case-match study, where homogeneous patient pairs have been identified based on age, NYHA class, left ventricular ejection fraction, and presence of coronary artery disease [8]. These variables have all been recognized to affect long-term survival after AVR with stented bioprostheses [13]. In the work by David and colleagues, superior overall survival, freedom from cardiac death, and valve-related morbidity have been shown in operative survivors with the Toronto SPV stentless valve compared with the Hancock II stented valve [8]. A Cox proportional hazard model has confirmed the latter two results. Because case-match analysis is theoretically exposed to the same bias influencing simple retrospective comparison, in the current study, a choice was made to apply multivariate analysis to the entire xenograft AVR population instead. The findings herein partly agree with the observations by David, as overall survival and survival free from cardiac- and from valve-related death are greater in recipients of stentless valves, but Cox analysis suggests the benefits of stentless valves to be limited to valve-related mortality. Reasons that may account for this difference comprise the older average age (62 vs 72 years), the higher prevalence of female patients (29% vs 49% of patients) and of associated cardiac disease (32% vs 45% of patients), and the inclusion of patients with associated mitral, aortic, and carotid artery disease, in the present series. Accordingly, demographic (advanced age, female gender, aortic stenosis, smaller aortic anulus, associated disease) and operative variables (size of prosthesis, aortic cross-clamp time) emerge as predictors of hospital, overall, and cardiac-related mortality, in the current analysis. The disparity in patient age has relevant implications, as Del Rizzo and coworkers have shown that survival advantage of stentless bioprostheses is most prominent in younger patients (less than 60 years of age) and diminishes with advancing age [4]. The present concurrent, control trial corroborates the observations by David and associates on superior freedom from valve-related mortality in patients having the Toronto SPV stentless xenografts [8] and extends the same findings to older patients and to other models of stentless valves (Biocor PSB, Cryolife-O Brien). Prior work by Del Rizzo and colleagues has shown the unique hemodynamic advantages of the Toronto SPV valve [9]. Similar behavior has been documented after AVR using the Biocor PSB [6] and the Cryolife-O Brien valves [14]. Common to most stentless xenograft models is the property to enhance regression of left ventricular hypertrophy to a greater extent than stented bioprostheses, as two recent prospective, randomized trials have proved [11, 12]. It has been demonstrated that the larger effective orifice area of stentless valves, when compared with stented of the same external diameter (ie, lower incidence of prosthesis patient mismatch) [15], may account for the more thorough regression of left ventricular hypertrophy after operation [16]. As ventricular hypertrophy has since been associated with increased cardiac mortality [17], the most credited hypothesis is that the unique hemodynamic properties of stentless xenografts are responsible for the observed greater survival free from cardiac death. Additional reasons, however, explain the selective survival advantage of stentless valves in the current study. In fact, a sizable proportion of late casualties has been caused by valve-related events. In particular, rate of embolic complications, including lethal, has proved higher among recipients of stented xenografts, as previously found by others [6, 7]. Whether lower thrombogenicity of stentless valves or, simply, unsatisfactory control of blood anticoagulation in patients with stented valves is responsible for this result awaits verification. Based on the current findings, the survival advantage conferred by stentless xenograft may be due not so much to the more favorable impact of the valve on left ventricular function and remodeling, as much to the lower morbidity related to the prosthetic device per se. Contrary to the hypothesis on selective survival benefits of stentless valves, the expectation of longer durability than stented bioprostheses thus far has been unsatisfied. Similar to the observations by David and associates [8], the present control trial confirms that structural deterioration affects all xenografts, whether stentless or stented, at the same rate, albeit still low 8 years after implant. A series of explanations exists that can account for this finding, including: an incorrect hypothesis underlying the causes of degeneration of bioprostheses; the instability of the anatomic substrate, where stentless valves are grafted; the detrimental impact of technical factors connected with stentless valve surgery; and the substantial improvement in durability of secondgeneration stented bioprostheses. The original observation that freehand-sewn aortic homografts have proved less prone to structural deterioration than stent-mounted has led to the anticipation that freehand stentless xenografts would outlast stented bioprostheses [1]. The hypothesis entertained was that greater resemblance to the native aortic valve anatomy and physiology would minimize shear stress on the leaflets, a factor recognized among the primary causes of tissue valve degeneration [18]. It is possible that this hypothesis may not be correct and that other factors, such as immune response to the xenograft (therefore irrespective of valve design), play a predominant role. Moreover, the functional properties of stentless valves are strictly dependent on the integrity of the aortic valve-root complex, as the prostheses are anchored to the native aortic root. Jin and Westaby have shown how progressive dilatation of the sinotubular junction, which frequently occurs with aging, causes progressive prosthetic valve insufficiency due to reduced leaflet coaptation [19]. Dysfunction of the stentless valve may, in turn, be associated with increased shear stress and premature xenograft deterioration. Prior work from our institution has shown that, contrary to stented bioprostheses, deterioration of stentless valves manifests

6 1448 LUCIANI ET AL Ann Thorac Surg STENTLESS VERSUS STENTED XENOGRAFTS 2002;74: with progressive stiffening and rupture, often abrupt, of the leaflets at the commissures, in the absence of calcification [20]. In addition, mode and rate of degeneration are identical with the three different valve models: Toronto SPV, Biocor PSB, and Cryolife-O Brien [20]. Therefore, a unique way of failure exists, common to most stentless xenografts, that reflects prosthetic valve design, as demonstrated by the morphologic findings of degenerated xenografts. As a corollary, it follows that even minor mishaps with the implant technique, often undetected at the time of operation, may translate into relevant valve dysfunction with time [21]. This finding is again specific of stentless xenografts and will obviously not affect stent-mounted bioprostheses. Lastly, a growing body of evidence has shown that durability of secondgeneration stented xenografts has dramatically improved, possibly due to low-pressure fixation and antimineralization treatment of the leaflets [13, 22]. When considered as a whole and added to the observation that nonstructural deterioration may be more common due to greater complexity of operation, as seen in the present and in prior series [23, 24], it is not surprising that freedom from valve failure with stentless xenografts has not proved more gratifying than with stented. Survival free from reoperation on the valve is not a valid end point to compare the durability of stentless and stented xenografts. Given the slowly progressing nature of transprosthetic obstruction in deteriorated stented bioprostheses, the decision to intervene may often be deferred, particularly in the very elderly patient. On the contrary, degeneration of stentless valves is less predictable, and the volume overload consequent to prosthetic valve regurgitation is poorly tolerated by a noncompliant left ventricle, such as the one in senile aortic stenosis. Therefore, replacement of a failing stentless valve must be scheduled promptly at the time of diagnosis. Accordingly, multivariate analysis has confirmed older age to be inversely and stentless valve to be directly related with the replacement of a dysfunctional xenograft, in the present experience. The importance of elective planning is evident from the current and prior work [25], showing significantly higher mortality for emergent procedures. The influence of clinical experience is also apparent, as operative risk may be greatly contained at centers with longstanding acquaintance with aortic root surgery. Similar to the findings on freedom from structural valve failure and reoperation, functional status late after implant does not allow to distinguish between stentless and stented xenografts. Both types of prostheses are associated with satisfactory clinical conditions, even in an elderly population commonly affected by associated cardiac disease, such as the one herein. This result agrees with previous studies on stentless and stented valves [2, 3, 6, 7, 13, 22]. However, it seems at odds with the superior hemodynamic behavior demonstrated for stentless xenografts [11, 12, 15], which could in theory be associated with improved clinical status. Restrictions inherent with the NYHA classification, particularly its subjective nature, and the advanced patient age may conceivably account for this divergence. Limitations The present study presents the limitations of any retrospective analysis, primarily bias in the selection of patients. Contrary to prior works, however, it reports on a numerous and mostly older population, undergoing concurrent, rather than consecutive, surgical treatment using either stentless or stented bioprostheses. A second limitation is that three stentless valve models are simultaneously compared with one stented valve model, possibly weakening the conclusions of the study due to small sample size. However, the ability to extend the observations on the Toronto SPV valve to other prostheses may at the same time represent the original aspect of the current analysis. Based on the premise that functional behavior is similar with different stentless valve models, the present work confirms that clinical outcome may also be equally favorable. Lastly, and most notably, the study herein only partly overcomes the limitations inherent with simple retrospective comparison by using multivariate analysis methodology. In theory, only prospective, randomized long-term trials will provide rigorous information on comparative clinical outcome. In conclusion, this concurrent, control trial demonstrates that stentless aortic xenografts may confer selective advantages in terms of survival free from valverelated mortality when compared with stented. Because freedom from valve deterioration is similar, extension of stentless AVR to any patient without anatomic contraindication deserves further consideration. References 1. David TE, Pollick C, Bos J. Aortic valve replacement with stentless porcine aortic bioprosthesis. J Thorac Cardiovasc Surg 1990;99: Cohen G, Christakis GT, Buth KJ, et al. Early experience with stentless versus stented valves. Circulation 1997;96:II Rao V, Christakis GT, Sever J, et al. A novel comparison of stentless versus stented valves in the small aortic root. J Thorac Cardiovasc Surg 1999;117: Del Rizzo DF, Abdoh A, Cartier P, Doty D, Westaby S. The effect of prosthetic valve type on survival after aortic valve surgery. Semin Thorac Cardiovasc Surg 1999;11: Collinson J, Henein M, Flather M, Pepper JR, Gibson DG. Valve replacement for aortic stenosis in patients with poor left ventricular function: comparison of early changes with stented and stentless valves. Circulation 1999;100:II Vrandecic M, Fantini FA, Filho BG, de Oliveira OC, da Costa Junior IM, Vrandecic E. Retrospective clinical analysis of stented vs. stentless porcine aortic bioprostheses. Eur J Cardiothorac Surg 2000;18: Westaby S, Horton M, Jin XY, et al. Survival advantage of stentless aortic bioprostheses. Ann Thorac Surg 2000;70: David TE, Puschmann R, Ivanov J, et al. Aortic valve replacement with stentless and stented porcine valves: a case-match study. J Thorac Cardiovasc Surg 1998;116: Del Rizzo DF, Goldman BS, Christakis GT, David TE. Hemodynamic benefits of the Toronto stentless valve. J Thorac Cardiovasc Surg 1996;112: Edmunds LH Jr, Clarke RE, Cohn LH, Grunkemeier GL, Miller DC, Weisel RD. Guidelines for reporting mortality and morbidity after cardiac valvular operations. J Thorac Cardiovasc Surg 1996;112:

7 Ann Thorac Surg LUCIANI ET AL 2002;74: STENTLESS VERSUS STENTED XENOGRAFTS Walther T, Falk V, Langebartels G, et al. Prospectively randomized evaluation of stentless versus conventional biological aortic valves: impact on early regression of left ventricular hypertrophy. Circulation 1999;100:II Williams RJ, Muir DF, Pathi V, MacArthur K, Berg GA. Randomized controlled trial of stented and stentless aortic bioprostheses: hemodynamic performance at 3 years. Semin Thorac Cardiovasc Surg 1999;11: David TE, Ivanov J, Armstrong S, Feindel CM, Cohen G. Late results of heart valve replacement with the Hancock II bioprosthesis. J Thorac Cardiovasc Surg 2001;121: Gelsomino S, Frassani R, Porreca L, Morocutti G, Morelli A, Livi U. Early and midterm results of model 300 CryoLife- O Brien stentless porcine aortic bioprosthesis. Ann Thorac Surg 2001;71(Suppl):S Yun YL, Jamieson WR, Khonsari S, Burr LH, Munro AI, Sintek CF. Prosthesis-patient mismatch: hemodynamic comparison of stented and stentless aortic valves. Semin Thorac Cardiovasc Surg 1999;11: Del Rizzo DF, Abdoh A, Cartier P, Doty D, Westaby S. Factors affecting left ventricular mass regression after aortic valve replacement with stentless valves. Semin Thorac Cardiovasc Surg 1999;11: Levy D, Garrison RJ, Savage DD, et al. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322: Thurbrikar M, Deck JD, Aouad J, Nolan SP. Role of mechanical stress in calcification of aortic bioprosthetic valves. J Thorac Cardiovasc Surg 1983;86: Jin XY, Westaby S. Aortic root geometry and stentless porcine valve competence. Semin Thorac Cardiovasc Surg 1999;11: Luciani GB, Auriemma S, Casali G, Barozzi L, Lamascese N, Mazzucco A. Experience with stentless aortic xenografts. J Heart Valve Dis 2000;9: Bach DS, LeMire MS, Eberhart D, Armstrong WF, Deeb GM. Impact of intraoperative post-pump aortic regurgitation with stentless aortic bioprostheses. Semin Thorac Cardiovasc Surg 1999;11: Jamieson WRE, Lemieux MD, Sullivan JA, Munro I, Metras J, Cartier PC. Medtronic Intact porcine bioprosthesis experience to twelve years. Ann Thorac Surg 2001;71(Suppl): S Hvass U, Chatel D, Assayag P, et al. The O Bien-Angell stentless porcine valve: early results with 150 implants. Ann Thorac Surg 1995;60(Suppl):S Schoof PH, Baur LH, Kappetein AP, Hazekamp MG, van Rijk-Zwikker GL, Huysmans HA. Dehiscence of the Freestyle stentless bioprosthesis. Semin Thorac Cardiovasc Surg 1999;11: Luciani GB, Bertolini P, Vecchi B, Mazzucco A. Reoperation on stentless aortic xenografts. Ann Thorac Surg 1998; 66(Suppl):S Appendix Variables entered in the multivariate analysis included age (years), gender (female), BSA (m 2 ), diagnosis (aortic stenosis), prior aortic procedure, associated coronary artery disease, other associated cardiac disease, NYHA functional class, aortic anulus diameter (mm), size of xenograft valve (mm), type of prosthetic device (stentless), model of stentless xenograft (Biocor PSB, Toronto SPV, Cryolife-O Brien), duration of aortic cross-clamp time (minutes), duration of cardiopulmonary bypass time (minutes), and duration of intensive care stay (days).