European Journal of Cardio-Thoracic Surgery (2013) 1 7 doi:10.1093/ejcts/ezt367 ORIGINAL ARTICLE Haemodynamic performance of a new pericardial aortic bioprosthesis during exercise and recovery: comparison with pulmonary autograft, stentless aortic bioprosthesis and healthy control groups Thorsten Hanke a, *, Efstratios I. Charitos a,, Hauke Paarmann b, Ulrich Stierle a and Hans-H. Sievers a a b Clinic for Cardiac and Thoracic Vascular Surgery, University Clinic of Schleswig-Holstein, Lübeck, Germany Clinic for Anaesthesiology, University Clinic of Schleswig-Holstein, Lübeck, Germany * Corresponding author. Clinic for Cardiac and Thoracic Vascular Surgery, University Clinic of Schleswig-Holstein, Campus Lübeck, Ratzeburger Allee 160, Lübeck 23538, Germany. Tel: +49-451-5002108; fax: +49-451-5002051; e-mail: thorsten.hanke@uksh.de (T. Hanke). Received 26 January 2013; received in revised form 6 June 2013; accepted 17 June 2013 Abstract OBJECTIVES: Since blood flow impairment by aortic valve prosthesis is characteristically dynamic, this dynamic component is best and thoroughly appreciated by exercise Doppler echocardiography. We sought to determine the haemodynamics of a new pericardial aortic bioprosthesis [Trifecta -aortic valve bioprosthesis (T-AVB), St Jude Medical, MN, USA] at rest and during exercise and a 10-min recovery period in comparison with alternative aortic valve prostheses, e.g. Ross operation (RO), stentless aortic valve [Medtronic freestyle-aortic valve bioprosthesis (MF-AVB)] and a healthy control group (CO). METHODS: Haemodynamics at rest and during supine exercise stress testing and a 10-min recovery period were evaluated in 32 patients (mean age: 70.8 ± 6.7 years) with T-AVB (mean follow-up: 5 ± 2 months), 49 with RO (mean age: 43.5 ± 13.7 years), 39 with an MF-AVB (mean age: 64.6 ± 9.4 years) and 26 healthy patients (mean age: 39 ± 9 years). Measurements included mean outflow tract gradient (δpmean, mmhg), effective orifice area index (EOAI, cm 2 /m 2 ) and valvular resistance (vr, dyn s cm 5 ). RESULTS: Mean body surface area for T-AVB was 1.93 ± 0.24 m 2 (median 1.97 m 2 ). Mean δpmean at rest was 7.2 ± 3.4 mmhg, mean EOAI 0.86 ± 0.23 cm 2 /m 2 and mean vr 50.7 ± 23.2 dyn s cm 5. Supine stress testing did increase the mean EOAI to 0.98 ± 0.27 cm 2 /m 2,themean vr to 62.6 ± 25.3 dyn s cm 5 and the mean δpmean to 10.21 ± 4.7 mmhg, respectively (P < 0.05 for all comparisons). During the post-exercise recovery period, δpmean, EOAI and vr showed a prompt normalization within 5 min of cessation of exercise. At all the three measurement points, T-AVB and MF-AVB revealed low gradients, satisfactory EOAI and low vr. Compared with the RO and a healthy control group, both groups showed significantly inferior performance throughout the exercise and post-exercise study protocol (P < 0.05). In comparison with T-AVB, patients with an MF-AVB only showed significant inferior performance throughout series with respect to a higher vr, indicating a smaller increase in the EOAI during exercise. During the 10-min post-exercise period, T-AVB recovered significantly earlier than MF-AVB. CONCLUSIONS: When comparing two different types of aortic valve bioprostheses with a gold standard group (RO) and a healthy population, both aortic valve bioprostheses perform inferior but reveal promising haemodynamics during exercise. During post-exercise haemodynamic recovery, only the T-AVB revealed a nearly physiological recovery pattern compared with the RO and a healthy control group. Keywords: Exercise stress test Haemodynamics Aortic valve bioprosthesis INTRODUCTION When implanting a stented aortic valve prosthesis, a remaining gradient across the valve is inevitable in general due to the valve s design, resembling an obstruction to blood flow through the left ventricular outflow tract [1]. The fairly new valve design concept of supra-annular positioning, e.g. fixation of the valve above the sewing ring with consecutive implantation of the valve distal to the native aortic valve annulus, theoretically avoids blood flow impairment in the left ventricular outflow tract (LVOT), since the aortic valve prosthesis does not reach into the LVOT lumen. The same theoretical positive influence of reduced blood flow These authors have contributed equally to the manuscript. impairment by valve design also accounts for stentless aortic valve bioprostheses, since a flow-reducing metal stent in these prostheses does not exist [2]. This haemodynamic aspect, although not being assessed scientifically in larger studies, might prevent structural valve deterioration among these two valve types [3]. The pulmonary autograft procedure [Ross operation (RO)] avoids the impact of LVOT blood flow impairment due to its natural and physiological implantation with excellent haemodynamic results [4]. Mostly up to now, the echocardiographic assessment of heart valve prosthesis performance is performed with the patients at rest. However, this may not reflect the haemodynamic conditions in valve performance during everyday activities (mild-to-moderate exercise). Therefore, these haemodynamic changes are best and The Author 2013. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
2 T. Hanke et al. / European Journal of Cardio-Thoracic Surgery thoroughly appreciated by exercise Doppler echocardiography [5]. Furthermore, the echocardiographic observation shortly after cessation of exercise might be useful for further risk stratification in patients undergoing heart valve surgery [6]. During this period, the endurance of load stress to the left ventricle (LV), as indicated by elevated haemodynamic parameters, can be observed. Although not being examined in larger studies, these pathologically elevated parameters after moderate or even severe exercise might encounter further LV performance depression [7]. The St Jude Medical (SJM) Trifecta (MN, USA) aortic valve bioprosthesis (T-AVB) is a pericardial valve bioprosthesis designed for supra-annular placement, incorporating certain design features, which may improve haemodynamics. This valve received CE approval in 2010 and was evaluated in an international multicentre preclinical trial, starting in 2007. The Medtronic freestyle-aortic valve bioprosthesis (MF-AVB) is a stentless porcine aortic root with an intact aortic valve fixed with glutaraldehyde at zero pressure and treated with α-amino-oleic acid to reduce tissue degeneration. The first clinical use was started in 1992 and it was approved for clinical use in the USA in 1997. Until now, no peer-viewed published data of the T-AVB are available with respect to haemodynamics during exercise. This even more accounts for haemodynamic assessment post-exercise in addition with a comparison of other assumingly haemodynamically superior aortic valve substitutes. Thus, it was the purpose of the conducted study to evaluate the haemodynamic parameters after standardized implantation of a T-AVB at rest, during moderate exercise and during post-exercise recovery and to compare this set of data with a group of patients with an implanted stentless aortic valve bioprosthesis (Medtronic Freestyle, Medtronic, MN, USA), a group of patients with an implanted pulmonary autograft (the RO) as described by Sievers et al.[4] and a healthy group with non-operated aortic heart valves. MATERIALS AND METHODS Patients Between July 2010 and October 2012, 320 patients underwent aortic valve replacement with a T-AVB at our institution. During the early phase of T-AVB implantation in our clinic ( July 2007 till January 2008), eligible patients for the exercise stress test study were stratified with the following inclusion criteria: regular normal sinus rhythm, a moderate or normalized ventricular ejection fraction (EF 45%) and stable clinical conditions. Patients with moderate or severe mitral valve regurgitation post-cardiac surgery, with an implanted pacemaker, a known history of peripheral vascular disease and the inability to exercise were excluded. Thus, 32 patients with a mean follow-up of 5 ± 2 months and a mean age of 70.8 ± 6.7 years were enrolled as the final exercise study T-AVB cohort. Data from historical groups of 49 Ross-operated patients with a mean age of 43.5 ± 13.7 years, 39 patients after MF-AVB replacement (mean age: 64.6 ± 9.4 years and mean follow-up 26 ± 18 months) and 26 healthy individuals (mean age 39 ± 9 years) with no signs of heart valve disease or any other cardiac disease, serving as a control group (CO), were chosen for comparison. All of these patients and the control group were examined with exactly the same exercise echocardiographic protocol as the T-AVB patients. The institutional ethics committee approved the study and patients gave their informed consent. Preoperative data of each operated and non-operated group are summarized in Table 1. Table 1: Preoperative patient characteristics Trifecta Stentless Ross Control Number of patients 32 39 49 25 Age at operation (years) 71 ± 7 65 ± 10 44 ± 13 39 ± 9 Sex Male 19 (59.4) 27 (69.2) 40 (81.6) 14 (56) Female 13 (40.6) 12 (30.8) 9 (18.4) 11 (44) Body surface area (m 2 ) 1.9 ± 0.21 1.9 ± 0.2 1.94 ± 0.21 1.88 ± 0.18 Body mass index (kg/m 2 ) 27.0 ± 3.7 27.0 ± 3.8 24.6 ± 3.5 24.1 ± 2.8 New York Heart Association classification Class I 1 (3.1) 1 (2.6) 16 (32.7) 25 (100) Class II 10 (31.3) 20 (51.3) 23 (46.9) Class III 18 (56.3) 14 (35.9) 10 (20.4) Class IV 3 (9.4) 4 (10.3) Type of aortic valve disease Insufficiency 7 (21.9) 3 (7.7) 6 (12.2) Stenosis 16 (50.0) 10 (25.6) 17 (34.7) Combined 9 (28.1) 25 (64.1) 26 (53.1) Aortic dissection Type A 1 (2.6) Pressure gradient across aortic valve δp max (mmhg) 56.23 ± 27.71 75.87 ± 25.53 54.14 ± 26.33 5.41 ± 1.71 (n = 25) δp mean (mmhg) 36.16 ± 20.23 48.34 ± 17.62 42.44 ± 20.93 3.03 ± 0.91 (n = 27) Left ventricle ejection fraction (%) 59.00 ± 13.08 61.48 ± 12.02 63.03 ± 17.74 70.60 ± 5.78 (n = 27) Additional conditions Coronary artery disease 19 (59.4) 26 (66.7) 1 (2.0) Hypertension 29 (90.6) 28 (71.8) 17 (34.7) Diabetes mellitus 9 (28.1) 11 (28.2) 1 (2.0)
T. Hanke et al. / European Journal of Cardio-Thoracic Surgery 3 Table 2: Perioperative patient characteristics Trifecta Stentless Ross Control Size of prosthesis (mm) 19 3 (9.4) 1 (2.6) 21 8 (25.0) 6 (15.4) 23 15 (46.9) 7 (17.9) 25 6 (18.8) 14 (35.9) 27 8 (20.5) 29 3 (7.7) Mean 22.5 24.6 Median 23 25 Perfusion time (min) 150.22 ± 54.77 141.56 ± 45.94 202.02 ± 24.15 Cross-clamp time (min) 118.97 ± 47.53 112.59 ± 37.85 167.92 ± 21.46 Concomitant procedures None 9 (28.1) 23 (59.0) 28 (57.1) Coronary artery bypass graft/ internal mammary artery 15 (46.9) 10 (25.6) Mitral valve repair 6 (18.8) 2 (5.1) 2 (4.1) Tricuspid valve repair 1 (3.1) 1 (2) Others 11 (34.4) 7 (17.9) 19 (38.8) Surgical procedure For all aortic valve replacement procedures, standard cardiopulmonary bypass was used with mild hypothermic cardiac arrest (34 C), using cold blood cardioplegia. After decalcification of the aortic valve annulus, the original valve sizer was used to choose for correct valve size. The T-AVB was implanted in a standardized fashion using pledgeted 2-0 interrupted non-inverting mattress sutures. The MF-AVB, in the patients being chosen for stress testing, was implanted in a sub-coronary fashion using a 3-0 Prolene running suture for fixation of the proximal annulus suture line and a 5-0 Prolene running suture for the distal sinus suture line. All ROs were performed in a sub-coronary fashion, as described previously by Sievers et al. [4]. Further perioperative data are summarized in Table 2. Echocardiographic follow-up Transthoracic Doppler echocardiography was performed at rest and continuously during exercise and a 10-min post-exercise recover period directly after cessation of exercise, with either a Hewlett Packard/Philips Sonos 5500 System (Philips/Hewlett-Packard, Andover, MA, USA) with a 2.5- to 4.0-MHz ultrasound transducer (MF-AVB, CO and RO) or a Vivid 7 System (GE Healthcare, WI, USA). Further details of the echocardiographic evaluation have been reported previously [4]. Exercise protocol Exercise was performed in a supine fashion with treadmill testing (Bosso Ergofit 877, Germany). A modified lead I electrocardiogram was continuously recorded. Blood pressure was measured every 2 min by cuff sphygmomanometer (Dinamap, Siemens, Germany). Transthoracic echocardiography was performed at rest and continuously during exercise and a 10-min recovery period after cessation of exercise. Patients were encouraged to exercise up to a maximum workload of 100 W, which resembles walking up two stairways at a time. Exercise was started at 25 W and was increased every 2 min by 25 W steps. The test was aborted in case of severe systolic hypertension (systolic blood pressure >220 mmhg), new onset of arrhythmias or exhaustion. Throughout exercise and post-exercise recovery, flow velocities across the aortic valve (continuous wave doppler) and in the LVOT were measured after 90 s after each 2-min step. Examinations were recorded on S-VHS videotapes and digitally on magneto-optical discs, and evaluated at separate time points. Investigations and analyses of the acquired echocardiographic data were performed by one investigator (U.S.). Calculations of mean pressure gradient (δpmean, mmhg), effective orifice area index (EOAI, cm 2 /m 2 ) and valvular resistance (vr, dyn s cm 5 ) have been described previously [8]. Statistical analysis Categorical data are given as total numbers and relative frequencies. Continuous data are given as mean ± standard deviation. Comparisons between groups were made using t-tests, analysis of variance, χ 2 or Mann Whitney U-test methods where appropriate according to the type and the distribution of the investigated variables. Time to restoration of haemodynamics (within a 10% margin of measurement error) were analysed with time-to-event methods. A P-value of <0.05 was considered significant. The Bonferroni correction was employed for multiple comparisons with a correction factor of 6 (the number of post hoc comparative combinations between the groups CO, RO, T-AVB and MF-AVB). Statistical analyses were performed using R version 2.15.2. RESULTS Haemodynamic parameters at rest, maximum exercise and at the end of recovery period for the four different groups studied are displayed in Table 3. For both aortic valve bioprosthetic valve types, single-digit mean gradients at rest were observed. T-AVB mean systolic pressure gradients as well as EOAI and vr at all the three measurement points showed almost similar values
4 T. Hanke et al. / European Journal of Cardio-Thoracic Surgery Table 3: Haemodynamic values for the three different patient groups and the healthy control group at rest, maximum exercise and after cessation of recovery Group Rest Maximum exercise End of recovery δp (mmhg) Control 3.03 ± 0.93 (1.5 4.7) 6.05 ± 1.77 (2.8 9.5) 2.79 ± 0.92 (1 4.6) Ross 3.05 ± 1.65 (1.25 8.84) 4.64 ± 2.51 (1.62 13.45) 3.32 ± 1.65 (1.17 7.3) Stentless 8.67 ± 4.51 (3 24.35) 12.37 ± 5.39 (4.13 28.8) 11.02 ± 6.25 (4.16 33.2), * Trifecta 7.21 ± 3.36 (2.19 15.2) # 10.21 ± 4.65 (3.58 21.6) # 8.07 ± 4.01 (1.52 16.5) #, * vr (dyn s cm 5 ) Control 19.88 ± 5 (11.29 29.4) 26.68 ± 7.4 (14.93 41.31) 18.69 ± 5.04 (6.2 28.99) Ross 17.15 ± 6.1 (7.34 34.27) 24.15 ± 8.67 (8.22 52.78) 18.17 ± 6.36 (5.87 31.87) Stentless 60.75 ± 29.52 (19.98 134.45) 82.14 ± 43.55 (30.86 195.8), * 71.2 ± 34.76 (25.82 147.21), * Trifecta 50.75 ± 23.15 (20.43 115) # 62.58 ± 25.3 (26 130) #, * 54.72 ± 29.29 (15 129) #, * EOAI (cm 2 /m 2 ) Control 1.36 ± 0.32 (0.99 2.19) 1.53 ± 0.38 (1.06 2.79) 1.43 ± 0.36 (0.95 2.62) Ross 1.53 ± 0.39 (0.97 2.49) 1.62 ± 0.42 (1.01 2.56) 1.51 ± 0.39 (0.88 2.56) Stentless 0.8 ± 0.18 (0.54 1.23) 0.82 ± 0.26 (0.38 1.4) 0.74 ± 0.23 (0.3 1.26) Trifecta 0.84 ± 0.23 (0.44 1.31) # 0.98 ± 0.27 (0.53 1.49) # 0.88 ± 0.24 (0.47 1.44) # *T-AVB vs MF-AVB P < 0.05, MF-AVB vs RO P < 0.05, # T-AVB vs RO P < 0.05. vr: valvular resistance; EOAI: effective orifice area index. Figure 1: Mean pressure gradients for the entire exercise stress testing and recovery period for all patient groups studied. compared with the MF-AVB, but numerically lower with respect to mean gradient and vr, the latter being indicative of a larger T-AVB EOAI at peak exercise. In contrast, T-AVB and MF-AVB differed significantly from the RO and the control groups (higher mean gradients, smaller EOAI and lower vr, P < 0.05). Haemodynamics during exercise and recovery period: mean pressure gradients (δpmean) The mean pressure gradients of the T-AVB group increased significantly up to a maximum workload of 100 W and decreased back Figure 2: Percentage of patients with complete recovery of pressure gradients (δp) during the entire recovery period. All the patients were followed for 10 min after cessation of exercise. At 10 min in the CO, RO, T-AVB and MF-AVB, 0, 2, 3 and 10 patients had not reached complete recovery of pressure gradients, respectively. to resting values after the fourth recovery minute (Fig. 1). This haemodynamic behaviour during the post-exercise recovery period was comparable with that of the RO and control groups. Among these, a normalization of mean pressure gradient after cessation of exercise was reached at 4 min. In contrast, although the values at maximum exercise did not differ significantly, the MF-AVB group did not reach initial value throughout the entire recovery period (P < 0.05 rest vs 10-min recovery). A complete recovery of the elevated δpmean was reached by 93% of the T-AVB patients (Fig. 2).
T. Hanke et al. / European Journal of Cardio-Thoracic Surgery 5 Figure 3: Changes in the EOAI for the entire exercise stress testing and recovery period for all the patient groups studied. Figure 5: Percentage of patients with complete recovery of vr during the entire recovery period. All the patients were followed for 10 min after cessation of exercise. At 10 min in the CO, RO, T-AVB and MF-AVB, 0, 0, 2 and 15 patients had not reached complete recovery of valve resistance, respectively. in the MF-AVB group with a stable orifice area index throughout the entire exercise and recovery period (Fig. 3). Haemodynamics during exercise and recovery period: valvular resistance Under exercise conditions, vr for the T-AVB group as well as for the MF-AVB group, the RO and the control groups increased significantly (P < 0.05). At maximum exercise as well as at the end of recovery (10 min), MF-AVB in comparison with T-AVB revealed significantly higher values (P < 0.05). During the post-exercise period, T-AVB, RO and control groups revealed recovery patterns similar to initial values during the fourth minute after cessation of exercise (P < 0.05 until the fifth recovery minute). A great proportion of patients in the MF-AVB group did not recover till the end of the 10-min observation period (Fig. 4). The percentage of T-AVB patients that completely recovered from an elevated vr was 97% (Fig. 5). Figure 4: Changes in the vr for the entire exercise stress testing and recovery period for all the patient groups studied. Haemodynamics during exercise and recovery period: indexed effective orifice area During exercise, the EOAI in the T-AVB group increased significantly until maximum exercise (+17%, P = 0.02) and decreased back to normal within the first recovery minute. Among the RO and control groups, the EOAI also increased significantly until maximum exercise and decreased within the first recovery minute back to the initial value. In contrast, there was no change in EOAI DISCUSSION Many studies have been performed to evaluate aortic valve bioprosthesis after CE market or FDA approval. Most of the haemodynamic echocardiographic examinations of these implanted aortic valve prostheses are primarily performed under resting conditions, as this is the method of choice for prosthetic heart valve evaluation [9, 10]. This method of choice will inevitably miss the haemodynamic changes during everyday activities, which are characterized by an increase/decrease in the transvalvular pressure gradients, the changes of EOAI and an increase/decrease in vr according to cardiovascular demands. Furthermore,
6 T. Hanke et al. / European Journal of Cardio-Thoracic Surgery functioning or malfunctioning aortic valve prostheses may produce similar resting gradients. Echocardiographic follow-up protocol performed solely at rest can not detect a pathological gradient increase during exercise, e.g. > 20 mmhg for the aortic position, which can cause either moderate or severe inter-patient -prosthesis mismatch (EOAI 0.85 or 0.65 cm 2 /m 2, respectively) [10, 11]. Thus, for better detection of possible malfunctioning aortic valve prostheses, an exercise stress echocardiography is recommended, and hence, it provides the clinician with diagnostic and prognostic information that can contribute to subsequent clinical decisions [5, 9]. Only few studies have performed a comparative stress echocardiography, either with exercise or with inotropic pharmacological substances, in order to better evaluate a new prosthetic heart valve [10, 12, 13], but none of these has studied the early recovery period right after cessation of exercise. We focused on the early recovery period in order to investigate for how long the sustained LV stress persist after cessation of exercise, in other words, when after cessation of exercise a haemodynamic normalization has been achieved. In order to get the most informative comparative results, implanted heart valve bioprostheses are compared with healthy control groups, and thus the effect of possible blood flow impairment by valve prosthesis is best documented. The pulmonary autograft reveals an excellent haemodynamic behaviour at rest and also during exercise very much similar to that of healthy control groups, and thus, others and we believe that this cohort might be used as a surgical gold standard for haemodynamic comparison after aortic valve replacement therapies [14 17]. Therefore, in order to evaluate the SJM Trifecta aortic valve bioprosthesis performance most thoroughly, we choose a moderate comparative exercise echocardiography protocol in combination with a 10-min post-exercise recovery observational period. To the best of our knowledge, we describe the first comparative exercise study of this newly developed bovine pericardial aortic bioprosthetic heart valve. Aortic bioprosthetic valve function The mean pressure gradients at all the three measurement points (rest, maximum exercise and ending of the recovery period) of the T-AVB were higher than between the RO and the control groups, and similar to those of the MF-AVB, but this difference will not be clinically relevant. However, this difference in our opinion becomes notable, when one takes into consideration that the mean aortic labeled valve size of the T-AVB was 23 ± 2 mm and the labeled size of the MF-AVB was 25 ± 2 mm. Furthermore, it is worth mentioning that the T-AVB and the MF-AVB mean values at rest were low one-digit numbers, as this is expected for stentless aortic valve bioprosthesis but somewhat unexpected in stented pericardial bioprosthesis. Similar findings at rest for the T-AVB are observed by Bavaria et al. [18], who describe a mean pressure gradient after 2 years of follow-up for the Trifecta heart valve of 7.3 ± 4.6 mmhg. In accordance with others, the gradients at rest and maximum exercise for the RO, albeit a different exercise protocol was used, were similar or even lower compared with a healthy control group [15 17, 19]. Although clinically not relevant, the MF-AVB revealed the numerically highest gradients at all the three measurement points (Table 3). Additionally, Cordovil et al. observed similar findings [20]. Their findings can be explained due to the usage of a prototype stentless aortic valve bioprosthesis. In a large meta-analysis, however, for some stentless aortic valve bioprosthesis, gradients similar to those in our group were reported [21]. The fact, that the sub-coronary implantation technique with possible narrowing of the aortic root at the level of the annulus, as confirmed by the small EOAI and an elevated vr in this group, was chosen for the MF-AVB might explain the valve prosthesis performance, as this phenomenon was also reported by Matsue et al.[22]. Dynamic aortic bioprosthetic valve function When analysing the dynamic behaviour of the T-AVB and the MF-AVB, certain unexpected findings were observed. In comparison with the physiological RO and the control groups, the T-AVB and MF-AVB showed very similar satisfying gradient and vr development during exercise. Only the EOAI and vr changes of the T-AVB differed significantly from the MF-AVB with a similar recovery pattern during the post-exercise period, when compared with the RO and the control groups. Since the leaflets of the T-AVB are mounted on the outside of a titanium stent, this design might possibly enable the leaflets to open up wider during higher flow through the LVOT, as it is the case with raising exercise levels and shown by the increase in the EOAI. On the contrary, the aortic root geometry of the MF-AVB seems to be somewhat rigid, as shown by a constant EOAI over the entire exercise and recovery period. The flexibility of the T-AVB titanium stent and the leaflet opening as well as the rigid structure of the MF-AVB are supported by the development of vr in both bioprosthetic valve types. As described by the group of Pibarot, vr increases and the EOA does not change in more or less fixed stenotic valves, whereas in more flexible geometric settings, vr does not significantly increase due to an increase in the valve area (Figs 3 and 4)[23]. STUDY LIMITATIONS The non-randomization of bioprosthetic aortic valve implantation might cause a certain bias with respect to results, but a randomization of three different valve types (RO, T-AVB and MF-AVB) is difficult to perform, especially when the Ross procedure is predominantly performed in young patients, as shown in the mean age of our patient group. Implanting a bioprosthetic valve in young patients is associated with a high reoperation rate within the first 10 years of surgery, and therefore difficult to support [24]. Additionally, the four groups are somewhat heterogenic, but for the study endpoints, e.g. valve gradient, resistance and EOAI, we believe that this heterogeneity will not have a significant impact on the development of these parameters, especially between the T-AVB and the MF-AVB groups. The difference in follow-up time especially between the T-AVB and MF-AVB might have had an influence on the valve performance, since a longer postoperative time span might have had an impact on valve deterioration in the MF-AVB group. However, we believe that this possible aspect is negligible, since the mean follow-up in the MF-AVB was 26 months. During this time frame, degeneration due to calcification of the leaflets is unlikely to occur. Another limitating aspect is the lack of comparison with another new third generation bovine aortic bioprosthesis, but it was the primary goal of our study to compare a new bioprosthetic heart valve with another modern non-stented aortic valve bioprosthesis and secondly with an operated gold standard as presented by the RO, as well as with a healthy control group. In a
T. Hanke et al. / European Journal of Cardio-Thoracic Surgery 7 recent publication though, the Sorin Mitroflow pericardial bioprosthesis, a third generation bioprosthetic aortic valve with a similar design to the T-AVB, showed numerically higher gradients throughout a similar exercise protocol for all valve sizes studied [25]. In addition, although this again was not the primary aim of the study protocol, both tested aortic valve bioprostheses do perform significantly inferior in comparison with the pulmonary autograft. In fact, the Ross-operated patients haemodynamically very much behave like a healthy control group. Thus, this surgical aortic valve replacement procedure, especially in young patients, at least from a haemodynamic standpoint, ought to be considered as a surgical aortic valve replacement alternative. CONCLUSIONS The Trifecta pericardial aortic valve bioprosthesis and the MF-AVB offer promising haemodynamic results at rest and during moderate exercise with an almost physiological dynamic pattern. The post-exercise haemodynamic recovery pattern of the T-AVB, as well as the dynamic adaption of the EOA to haemodynamic load during exercise, although still inferior to Ross-operated patients, are similar to this haemodynamic surgical gold standard and healthy individuals. Whether this might translate into better clinical outcome and longer durability needs to be evaluated by longer and larger follow-up studies. ACKNOWLEDGEMENTS We thank Kathrin Meyer, Bettina Schröder, Petra Lingens and Jana Peise for their secretarial support. Conflicts of interest: Thorsten Hanke, Hans-H. Sievers and Efstratios I. Charitos received lecture honoraria by SJM (<US$10000). Thorsten Hanke is a consultant for SJM with minor honoraria (<US$10 000 per year). REFERENCES [1] Rashtian MY, Stevenson DM, Allen DT, Yoganathan AP, Harrison EC, Edmiston WA et al. Flow characteristics of bioprosthetic heart valves. Chest 1990;98:365 75. [2] Reardon MJ, David TE. Stentless xenograft aortic valves. Curr Opin Cardiol 1999;14:84 9. [3] Badano LP, Pavoni D, Musumeci S, Frassani R, Gianfagna P, Baldassi M et al. Stented bioprosthetic valve hemodynamics: is the supra-annular implant better than the intra-annular? J Heart Valve Dis 2006;15:238 46. [4] Sievers H-H, Stierle U, Charitos EI, Hanke T, Gorski A, Misfeld M et al. Fourteen years experience with 501 subcoronary Ross procedures: surgical details and results. J Thorac Cardiovasc Surg 2010;140:816 22, 822.e1 5. [5] Picano E, Pibarot P, Lancellotti P, Monin JL, Bonow RO. The emerging role of exercise testing and stress echocardiography in valvular heart disease. J Am Coll Cardiol 2009;54:2251 60. [6] Lee R, Haluska B, Leung DY, Case C, Mundy J, Marwick TH. Functional and prognostic implications of left ventricular contractile reserve in patients with asymptomatic severe mitral regurgitation. Heart 2005;91:1407 12. [7] Ruel M, Rubens FD, Masters RG, Pipe AL, Bédard P, Hendry PJ et al. Late incidence and predictors of persistent or recurrent heart failure in patients with aortic prosthetic valves. J Thorac Cardiovasc Surg 2004;127:149 59. [8] Sievers H-H, Stierle U, Charitos EI, Hanke T, Misfeld M, Matthias Bechtel JF et al. Major adverse cardiac and cerebrovascular events after the Ross procedure: a report from the German-Dutch Ross Registry. Circulation 2010; 122:S216 23. [9] Vahanian A, Alfieri O, Andreotti F, Antunes MJ, Barón-Esquivias G, Baumgartner H et al. Guidelines on the management of valvular heart disease (version 2012): the Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Eur J Cardiothorac Surg 2012;42:S1 44. [10] Pibarot P, Dumesnil JG. Hemodynamic and clinical impact of prosthesispatient mismatch in the aortic valve position and its prevention. J Am Coll Cardiol 2000;36:1131 41. [11] Pibarot P, Dumesnil JG. Prosthetic heart valves: selection of the optimal prosthesis and long-term management. Circulation 2009;119:1034 48. [12] Silberman S, Shaheen J, Merin O, Fink D, Shapira N, Liviatan-Strauss N et al. Exercise hemodynamics of aortic prostheses: comparison between stentless bioprostheses and mechanical valves. Ann Thorac Surg 2001;72: 1217 21. [13] Chambers J, Rimington H, Rajani R, Hodson F, Blauth C. Hemodynamic performance on exercise: comparison of a stentless and stented biological aortic valve replacement. J Heart Valve Dis 2004;13:729 33. [14] Hobson NA, Wilkinson GAL, Wheeldon NM, Lynch J. Hemodynamic performance of the Ultracor and Carpentier-Edwards aortic prostheses using exercise and dobutamine stress echocardiography. J Heart Valve Dis 2005; 14:828 34. [15] Phillips JR, Daniels CJ, Orsinelli DA, Orsinelli MH, Cohen DM, Brown DA et al. Valvular hemodynamics and arrhythmias with exercise following the Ross procedure. Am J Cardiol 2001;87:577 83. [16] Pibarot P, Dumesnil JG, Briand M, Laforest I, Cartier P. Hemodynamic performance during maximum exercise in adult patients with the Ross operation and comparison with normal controls and patients with aortic bioprostheses. Am J Cardiol 2000;86:982 8. [17] Sievers HH, Schmidtke C, Graf B. Hemodynamics of semilunar valves at rest and exercise at an average of more than two years after the Ross procedure. J Heart Valve Dis 2001;10:166 9; discussion 169 70. [18] Bavaria J. Two-year Results of the St Jude Medical Trifecta Pericardial Aortic Valve Bioprosthesis. http://shvd.org/abstracts/2011/c54_55.cgi. [19] Porter GF, Skillington PD, Bjorksten AR, Morgan JG, Yapanis AG, Grigg LE. Exercise hemodynamic performance of the pulmonary autograft following the Ross procedure. J Heart Valve Dis 1999;8:516 21. [20] Cordovil A, Filho OC, de Andrade JL, Rodrigues ACT, Gerola LA, Moises V et al. Exercise echocardiography in cryopreserved aortic homografts: comparison of a prototype stentless, a stented bioprosthesis, and native aortic valves. Echocardiography 2009;26:1204 10. [21] Rosenhek R, Binder T, Maurer G, Baumgartner H. Normal values for Doppler echocardiographic assessment of heart valve prostheses. J Am Soc Echocardiogr 2003;16:1116 27. [22] Matsue H, Sawa Y, Matsumiya G, Matsuda H, Hamada S. Mid-term results of freestyle aortic stentless bioprosthetic valve: clinical impact of quantitative analysis of in vivo three-dimensional flow velocity profile by magnetic resonance imaging. J Heart Valve Dis 2005;14:630 6. [23] Blais C, Pibarot P, Dumesnil JG, Garcia D, Chen D, Durand LG. Comparison of valve resistance with effective orifice area regarding flow dependence. Am J Cardiol 2001;88:45 52. [24] Chan V, Malas T, Lapierre H, Boodhwani M, Lam B-K, Rubens FD et al. Reoperation of left heart valve bioprostheses according to age at implantation. Circulation 2011;124:S75 80. [25] Bleiziffer S, Eichinger WB, Hettich IM, Ruzicka D, Badiu CC, Guenzinger R et al. Hemodynamic characterization of the Sorin Mitroflow pericardial bioprosthesis at rest and exercise. J Heart Valve Dis 2009;18:95 100.