TY - JOUR
AU1 - Hartrumpf,, Martin
AU2 - Albes, Johannes, M.
AU3 - Krempl,, Tanja
AU4 - Rudolph,, Volker
AU5 - Wahlers,, Thorsten
AB - Abstract Objectives: Severe sclerosis of the native aortic annulus can result in a tilted implantation position of mechanical prostheses. In this study, the effects of tilting and rotation on the hemodynamic performance of standard bileaflet valves were assessed in an extracorporeal mock circulatory system. Methods: A pulsatile mock circulation driven by a Berlin Heart® system was developed. Main physiological components of the human circulation were mimicked. SJM®-AHPJ prostheses (21, 23, 25 mm) were mounted in an artificial aortic root containing physiologically oriented coronary ostia. All experiments were performed under constant conditions (stroke volume 60 ml, heart rate 70 bpm, systolic pressure 130 mmHg). Hydrostatic pressures were measured via fluid-filled catheters, transvalvular flow by ultrasonic probes. Data were digitally recorded at 50 Hz. Multiple pressure, volume, energy, and dimension parameters were derived off-line. Each valve was tested in a 0° (untilted) versus 20° (tilted) position at three axial rotation angles (0°, 45°, 90°). Tilting was performed independent of rotation by elevation of the prosthesis in the non-coronary sinus. Results: In all valves and all rotation angles, tilting resulted in a size-dependent significant increase of mean pressure gradient (range, 28–35% [21 mm valve], 59–96% [23 mm valve], 124–220% [25 mm valve]), valvular resistance (39–51, 84–121, 177–332%), regurgitation volume (84–148, 32–131, 93–118%), and systolic energy loss (113–146, 30–132, 69–213%), as well as a decrease of total stroke volume (2–5, 0–11, 3–10%), effective stroke volume (6–11, 9–14, 14–22%), cardiac output (6–11, 8–14, 13–22%), and effective opening area (16–24, 32–37, 47–57%). The strongest impairment of hemodynamic performance was seen at 90° rotation with reference to total and effective stroke volume, cardiac output, mean pressure gradient, and regurgitation fraction. Conclusions: Tilting of bileaflet valves resulted in a significant impairment of systolic and diastolic hemodynamics. Superiority of larger valves diminished in the tilted position. The strongest tilting effect was seen at 90° rotation. Such a position should therefore be avoided or surgically corrected by rotating the valve. Heart valves, Aortic valve, Hemodynamics, Pulsatile flow, Cardiovascular physiology, Cardiovascular models 1 Introduction A wide variety of mechanical heart valves, such as caged-ball prostheses, tilting disc valves, or bileaflet valves, as well as porcine and bovine tissue valves, have been developed in the past decades [1]. Currently, both mechanical and biological prostheses are used for aortic valve replacement. As for bioprostheses, there is no need for lifetime anticoagulation. However, these valves deteriorate becoming calcified and stenotic, so that half of the patients need redo surgery within 10–15 years. Mechanical valves, in contrast, require permanent anticoagulation with its associated risk of bleeding and thromboembolic events. The great advantage is their permanent durability. In patients less than 70 years, mechanical prostheses are therefore preferred [1,2]. Among the leading manufacturers of the mostly used bileaflet valves, St.-Jude-Medical and Sulzer-Carbomedics must be named [3,4]. Severe calcification and fibrosis of the aortic annulus and left ventricular hypertrophy can lead to a tilted implantation position of the prosthetic valve, particularly in patients with long-term disease. Tight attachment of the valve to the native annulus may promote distortion of the surrounding tissues. Moreover, the surgeon usually attempts to insert the largest prosthesis as possible to counteract a size mismatch [5]. Not rarely, an unphysiological tilt angle can be the consequence. Only few studies have been performed aiming at the relationship between valvular hemodynamics and implantation position. Wurzel [6] demonstrated the effect of implantation geometry of tilting disc valves on the mean pressure gradient in vitro while an influence of valve rotation in vivo was described by Laas [7] and Kleine [8]. However, no data yet exist regarding the tilting effects of bileaflet prostheses. The current study was designed to evaluate the effect of moderate (i.e. 20°) tilting on the hemodynamic performance of standard bileaflet valves (21, 23, and 25 mm) at three different axial rotation angles (0°, 45°, and 90°) in a newly developed mock circulation. 2 Materials and methods 2.1 Model description For the purpose of this study, a versatile mock circulatory system was developed (Fig. 1) . To approach physiological properties, all essential physical elements of the human circulation were incorporated in the model [9]. The circulation was operated with a pneumatical Berlin Heart® driving system (Berlin Heart AG, Berlin, Germany) to generate pulsatile flow conditions. We used a modified 80 ml polyurethane ventricle to simulate the left side of the heart. The artificial ventricle comprised two chambers separated by a membrane. The ‘air side’ was connected to the driving unit, while the ‘blood side’ was integral part of the mock circulatory system. To obtain appropriate compatibility of the ventricle, new inflow and outflow tracts were created. The former inflow valve was replaced by a 27 mm mitral valve prosthesis (St. Jude Medical Inc., St. Paul, MN, USA). The outflow tract was shortened and attached to the artificial ‘aortic root’ made of elastic silicone rubber tubing. Coronary ostia were inserted in a physiological position. The investigated valve prosthesis was fixed in a physiological position directly upstream of the coronary ostia using a tight external clamp. The sewing ring was not removed to achieve the closest proximity to the in-vivo-situation. Bileaflet Hemodynamic-Plus valves of the SJM® Masters series (SJM®-AHPJ-505) of three different sizes (21, 23 and 25 mm) were tested using different tubing diameters (22, 24 and 26 mm, respectively). These valves contain an artificial annulus, which is mounted in a Polyester sewing ring in a rotatable fashion. Two semilunar-shaped leaflets are symmetrically attached to the annulus providing an opening angle of 85°. Both annulus and leaflets consist of pyrolytic carbon, which is extremely durable and biocompatible and exhibits low thrombogenicity [3,4,10–13]. The system was completed by physical components representing the peripheral and venous circulation (Fig. 1). Downstream of the aortic root, which was vertically suspended, two compliance chambers were sequentially implemented to mimic the ‘Windkessel’ properties of the arterial system. A throttle served as an adjustable peripheral resistance. Venous reservoir function was simulated by a basin draining back into the ventricular inflow tract. A constant fluid level of 10 cm was maintained corresponding to central venous filling pressure. The model was filled with 0.9% saline solution at room temperature to allow for optimum visibility [3,4,9,10,14]. Fig. 1 Open in new tabDownload slide Schematic view of the mock circulatory model. Pulsatile flow is generated by the Berlin Heart® driven polyurethane ventricle. Fluid is ejected into the artificial aortic root containing the tested valve and the various pressure and flow probes. Arterial and peripheral circulation is mimicked by a double compliance chamber followed by a resistance element. Venous reservoir function is simulated by a constant fluid-level basin draining back into the ventricle. Fig. 1 Open in new tabDownload slide Schematic view of the mock circulatory model. Pulsatile flow is generated by the Berlin Heart® driven polyurethane ventricle. Fluid is ejected into the artificial aortic root containing the tested valve and the various pressure and flow probes. Arterial and peripheral circulation is mimicked by a double compliance chamber followed by a resistance element. Venous reservoir function is simulated by a constant fluid-level basin draining back into the ventricle. 2.2 Measuring equipment Hydrostatic pressures were measured in the ventricular outflow tract 2 cm upstream of the tested valve (PV) as well as 5 cm downstream of the valve in the aorta (PAO) [6–8,15]. Pressures were obtained via fluid-filled catheters and transformed into electrical signals. An AS/3 Compact Monitor (Datex-Ohmeda Division, Helsinki, Finland) allowed for instant visual control of the pressure curves. Transvalvular flow (QAO) was measured using perivascular ultrasonic flowprobes (A-Series, Transonic Systems Inc., Ithaca, NY, USA). To achieve optimum acoustic coupling, they were directly embedded in the circulation. According to tubing diameter, a 28 mm flowprobe was used with the 23 and 25 mm prostheses and a 24 mm probe with the 21 mm prosthesis. The probe was positioned 6 cm downstream of the valve and connected to a HT-207 dual flowmeter unit (Transonic Systems Inc., Ithaca, NY, USA). Instantaneous volume flow was provided by application of transit-time technology, which is insensitive to probe tilting and allows for accurate measurements especially in turbulent flow patterns. Mean and instantaneous flow were displayed on the unit. Subsequent to calibration, all pressure and flow channels were simultaneously recorded on a PC at 50 Hz over a 60-s period using a 12-bit-A/D-converter (PC16TR, BMC Messsysteme GmbH, Maisach, Germany). Digital data were then available for further off-line processing. 2.3 Measuring protocol The assist device was constantly operated at the following conditions: systolic pressure 130 mmHg, diastolic pressure −2.5 mmHg, systolic period 40%, heart rate 70 bpm. This resulted in an aortic pressure of 120/70 mmHg and a stroke volume of 60 ml in the mock circulation. For each valve size, data sets were recorded at three different axial rotation angles (0°, 45° and 90°) and two tilting positions (0°=’physiologic’, 20°=’tilted’). As depicted in Fig. 2 (top), 0° rotation was defined as a position with the valvular centerline pointing exactly between the coronary ostia. Ninety degree rotation indicated a position transverse to the ostia, whereas 45° indicated an intermediate alignment. Tilting was performed independent of rotation. Zero degree tilting was defined as the physiological position with the valve plane being precisely rectangular to the tube axis. The tilted 20° position was achieved by elevation of the prosthesis in the non-coronary sinus regardless of the rotation angle (Fig. 2, bottom). A goniometer was used to achieve maximum accuracy. For each valve size, the following measuring protocol was applied: (1) 0° rotation, 0° tilting; (2) 0° rotation, 20° tilting; (3) 45° rotation, 0° tilting; (4) 45° rotation, 20° tilting; (5) 90° rotation, 0° tilting; and (6) 90° rotation, 20° tilting. Prior to data acquisition, the circulatory model was adjusted to the above-mentioned parameters according to the protocol. A steady-state was always obtained by running the device over 10 min. Stability of the model was statistically confirmed by a complete sham measurement (data not shown). In each of the experimental settings, multiple data sets were subsequently recorded 11 times over 60 s. Fig. 2 Open in new tabDownload slide Top: Axial rotation angles of valve prosthesis used in the model. Left, 0° rotation with hinge axis pointing between coronary ostia. Middle, 45° rotation represented by an intermediate position. Right, 90° rotation with hinge axis transverse to the ostia. RCO, right coronary ostium; LCO, left coronary ostium. Bottom: Tilting angles of valve prosthesis used in the model. Left, 0° tilting defined as the physiological position. Right, 20° tilting performed by elevation of the prosthesis in the non-coronary sinus. Tilting was carried out independent of rotation. Fig. 2 Open in new tabDownload slide Top: Axial rotation angles of valve prosthesis used in the model. Left, 0° rotation with hinge axis pointing between coronary ostia. Middle, 45° rotation represented by an intermediate position. Right, 90° rotation with hinge axis transverse to the ostia. RCO, right coronary ostium; LCO, left coronary ostium. Bottom: Tilting angles of valve prosthesis used in the model. Left, 0° tilting defined as the physiological position. Right, 20° tilting performed by elevation of the prosthesis in the non-coronary sinus. Tilting was carried out independent of rotation. 2.4 Data processing Off-line analysis of the data was performed using self-developed software. Time axes of the simultaneously recorded pressure and flow channels were synchronized when necessary. The single cardiac cycles were separated and superimposed so that an averaged beat could be generated to serve as a basis for all subsequent calculations. Peak-to-peak (ΔPmax) and mean (ΔPmean) transaortic pressure gradients were derived from the systolic pressure curves. Cardiac output (CO) corresponded to the measured transaortic volume flow. Systolic ejection time (Tsys) and heart rate (HR) were determined from the flow curve. Total stroke volume (SVtot) was computed by the systolic time integral of transvalvular flow. The diastolic part of the flow curve revealed information about valvular regurgitation. Closure time was represented by the negative peak following the ejection period while leakage time was determined by a subsequent slightly negative flow prior to the next systole. Transvalvular closure (VCl) and leakage volume (VL) were calculated by the corresponding time integrals, summation of which resulted in the total regurgitation volume (VR) [9,12]. Effective stroke volume (SVeff) corresponds with the resulting volume which is finally ejected into the peripheral circulation. It is calculated by subtracting the regurgitation from the total stroke volume (Eq. (1), see Appendix A for equations). Regurgitation fraction (RF) is an important parameter to characterize valvular hemodynamics, determined by the ratio of regurgitation to total stroke volume (Eq. (2)) [12,13]. Every time when volume flow across the valve is present, some energy loss occurs due to fluid friction imposing additional work load on the heart. Eq. (3) shows the calculation of systolic transvalvular energy loss (ΔEsys) [9]. Effective (functional) opening area (EOA) was obtained by the classic Gorlin formula for aortic valves (Eq. (4)). The discharge coefficient (Cd), defined as the ratio between EOA and the geometrical opening area (GOA), is a measure of how effectively the valvular aperture is utilized by flow (Eq. (5)) [12,13,16–18]. Finally, aortic valvular resistance (RAO) was calculated from transvalvular pressure gradient and flow reflecting the opposition to the ejected volume. Only the systolic period with wide open leaflets is considered (Eq. (6)). RAO gives an idea of the additional work load demanded from the ventricle [13]. 2.5 Statistics and data presentation Results are arranged in three categories to facilitate synopsis: (1) pressure gradients; (2) flow and volume parameters; and (3) transvalvular energy loss and valve-specific parameters. All data are displayed as mean±standard deviation and sorted by valve size (21/23/25 mm), axial rotation (0°/45°/90°) as well as tilting angle (0°/20°). Data were analyzed for normal distribution. Each parameter was then subjected to one-way ANOVA using Tukey's post hoc comparison (honestly significant difference). Significance is indicated for 20° versus 0° tilting with respect to the 5, 1, and 0.1% error levels: *P≤0.05, **P≤0.01, ***P≤0.001. A brief comment is given to each data table to focus on the principal conclusions. 3 Results 3.1 Pressure gradients (Table 1) Table 1 Open in new tabDownload slide Pressure gradients Table 1 Open in new tabDownload slide Pressure gradients Mean pressure gradient of the untilted valves showed an inverse correlation with the respective valve size. Tilting of the aortic valve prosthesis always led to a highly significant increase of mean pressure gradient at all sizes independent of rotation. The intensity of this effect was size-dependent with a range of 28–35% in the 21 mm valve, 59–96% in the 23 mm valve and 124–220% in the 25 mm valve. The gradients and their changes were highest in the 90° rotation except for the 23 mm valve (Fig. 3) . Peak gradient, conversely, showed inconsistent results among the valves in terms of rotation. Only the 21 mm valve yielded a constant increase with the strongest effect seen at 90° rotation. The 23 mm valve also showed a marked increase at 90° whereas a slight drop was seen in the 25 mm prosthesis. Fig. 3 Open in new tabDownload slide Tilting-related changes of mean values with respect to the different rotation angles (0°, 45° and 90°). ΔPmean, mean pressure gradient; CO, cardiac output. Data are separately displayed for each valve size and include standard deviation bars. All changes are statistically significant (20° versus 0° tilting, P≤0.001). Fig. 3 Open in new tabDownload slide Tilting-related changes of mean values with respect to the different rotation angles (0°, 45° and 90°). ΔPmean, mean pressure gradient; CO, cardiac output. Data are separately displayed for each valve size and include standard deviation bars. All changes are statistically significant (20° versus 0° tilting, P≤0.001). 3.2 Flow and volume parameters (Table 2) Table 2 Open in new tabDownload slide Flow, volume and related parameters Table 2 Open in new tabDownload slide Flow, volume and related parameters Total stroke volume was significantly reduced (2–4%) when tilting the 21 and 25 mm valves with the strongest alterations seen at 90° rotation (5–10%). Tilting of the 23 mm valve showed a drop in SVtot only at 45° and 90° rotation. Effective stroke volume and cardiac output revealed a consistent decrease in all tested prostheses with a focus on the 90° position. The highest drop (13–22%) occurred in the largest valve, the lowest drop (5–11%) in the smallest valve (Fig. 3). Closure volume also showed a size-dependent behavior. While tilting of the larger valves (23 and 25 mm) resulted in an increase which faded from 0° to 90° rotation, the 21 mm prosthesis yielded the strongest increment at 90° rotation whereas it was not significant at 0°. Likewise, leakage volume consistently increased at 20° tilting among all valve sizes (80–100% on the average). At 90° rotation, this effect was enhanced in the 21 and 25 mm prostheses (132 and 150%, respectively) but attenuated in the 23 mm valve (32%). According to the results of VCl and VL, regurgitation volume and fraction exhibited a significant increase at all rotations and sizes showing specific accentuations (Fig. 4) . In particular, tilting of the 23 mm valve led to the strongest rise in these parameters at 0° (130–135%) and the smallest rise at 90° rotation (32–47%). The 21 mm valve, however, showed the clearest effect at 90° rotation (133–148%) whereas in the largest prosthesis, the increase appeared to be rather independent of the axial rotation (93–138%). Fig. 4 Open in new tabDownload slide Tilting-related changes of mean values with respect to the different rotation angles (0°, 45° and 90°). VR, regurgitation volume; EOA, effective opening area. Data are separately displayed for each valve size and include standard deviation bars. All changes are statistically significant (20° versus 0° tilting, P≤0.001). Fig. 4 Open in new tabDownload slide Tilting-related changes of mean values with respect to the different rotation angles (0°, 45° and 90°). VR, regurgitation volume; EOA, effective opening area. Data are separately displayed for each valve size and include standard deviation bars. All changes are statistically significant (20° versus 0° tilting, P≤0.001). 3.3 Transvalvular energy loss and valve-specific parameters (Table 3) Table 3 Open in new tabDownload slide Transvalvular energy loss and valve-specific parameters Table 3 Open in new tabDownload slide Transvalvular energy loss and valve-specific parameters Tilting of the prosthetic valve caused a significant rise in systolic energy loss in all valve sizes without exhibiting a clear size dependence. The strongest effect occurred at 90° rotation in both the 21 and the 23 mm prostheses but at 45° rotation in the 25 mm valve. Tilting significantly decreased effective opening area and Cd to a comparable extent in each prosthesis regardless of rotation. However, differences were seen among the various valve sizes. EOA was positively correlated with the valve size regarding the untilted position (average 1.72, 2.45, and 3.34 cm2, respectively). The maximum change appeared when tilting the 25 mm valve (47–57%) followed by the 23 mm prosthesis (31–37%). The least reduction was noticed in the 21 mm valve (16–19%) with a focus on the 90° rotation (23–24%) (Fig. 4). Resistance changed inversely to the valve size showing the greatest increment in the 25 mm valve (up to 332%) compared with the 21 mm valve (up to 51%). 4 Discussion Mechanical valves have often been studied in vitro [3,4,6,9–12,14–16,18–21] and in vivo [8] in the past. Many details are known about energetics [3,4,9], flow profiles and shear stresses [8,10,11,16,21] as well as opening dynamics and regurgitation [3,4,9,12,20]. Typical dimensions exist for each mechanical valve type in terms of diameter, area, or excursion angles. The investigated SJM® bileaflet valves have the following opening areas: 2.55 (21 mm), 3.09 (23 mm), and 3.67 cm2 (25 mm) with a large opening angle of 85° and closing angle of 30° [10,15]. Such valves exhibit a tripartite flow profile with two major lateral portions, which represent 78% of the area available for forward flow, and a central minor portion. The shear stresses are elevated at the tube wall but moderate in magnitude. The SJM® valve shows the lowest pressure gradient but a somewhat higher regurgitation compared with tilting disc or caged-ball prostheses [3,4,10,11,15,16]. However, most in vitro studies about valvular hemodynamics have been conducted by engineers with little reference to clinical practice [4,9,10,12,14,15,20]. There are very few studies discussing the influence of implantation conditions [6–8]. Since valve replacement, aside from bypass surgery, represents the second largest number of cardiosurgical procedures [2], it is important to know how valvular function can be affected by the implantation technique. The effect of tilting of bileaflet valves has not yet been discussed. The current study addresses an important issue in cardiac surgery because a tilted implantation result must often be accepted in clinical practice. The hemodynamic effects of moderate tilting will be discussed in the following. 4.1 Size dependence of untilted valve Our results with the untilted implantation position demonstrate good correlation with the database of Walker [15]. Mean pressure gradient and valvular resistance showed a significant increase with the reduction of valve diameter. Effective opening area gives another estimate of valvular stenosis. In all prostheses, the values revealed a typical size dependence and were consistently lower than the corresponding geometrical area [15,16,18]. The difference between EOA and GOA was most accentuated in the 21 mm valve and least in the 25 mm valve implying that the valvular aperture can be more effectively utilized by flow in the larger prostheses. The reason lies in an augmented vortex formation in small prostheses impairing forward flow through the orifice [12,16,17]. This was also reflected in the discharge coefficient, which was, on the average, 35% higher in the 25 mm valve compared with the 21 mm valve. A similar behavior is described by Flachskampf [17] and Walker [16]. However, the stenotic effect of an artificial valve is fairly low compared with a diseased native valve. Work load on the left ventricle can therefore considerably be reduced by aortic valve replacement with subsequent regression of myocardial hypertrophy [13]. However, implantation of too small a valve can result in a size mismatch in terms of cardiac output. Ventricular function and long-term survival may substantially be impaired in this situation [5]. As for regurgitation, a typical size-dependent behavior was also noticed, which is consistent with recent observations [3,15,16]. Closure volume increases with growing valve diameter due to higher inertial forces of the leaflets [3,14]. Leakage volume, in contrast, is dependent on the ‘insufficiency area’, which is made up by tiny clefts between annulus and leaflets. These values were found to grow with rising valve diameter, which has also been described by other authors [16]. Diastolic leakage causes elevated shear-stress on corpuscular blood elements with promotion of thrombus formation and hemolysis [9,16,20]. On the other hand, a favorable wash-out effect is induced in the hinge region in order to attenuate thrombogenic potency [12,16,20]. 4.2 Rotation dependence of untilted valve The longitudinal axis of the left ventricle forms an angle of 140°–150° with its outflow tract in vivo. Consequently, the flow pattern of the ejected bloodstream shows the highest velocities near the non-coronary sinus [7]. On this background, studies have determined an optimum rotation position for tilting disc valves [7,8]. Bileaflet valves are less susceptible to rotation [7,8,11]. Our results accordingly showed no preference for a particular rotation in the untilted position. 4.3 Valve tilting Specific surgical conditions may require a tilted implantation of the valve. A moderate angle of 20°, as opposed to the untilted position, was assessed in this study. In most parameters, significant alterations occurred in the tilted position indicating global impairment of valvular hemodynamics with increased stenosis and regurgitation. 4.4 Systolic dysfunction A marked systolic dysfunction was expressed by a rise in mean pressure gradient and valvular resistance leading to a decrease of total stroke volume and cardiac output. This suggests obstruction of the outflow tract due to an impaired opening movement of the leaflets. A drop in effective opening area and Cd was consistently seen in all valves, which is also indicative of increased stenosis and may reflect flow acceleration and augmented turbulences near the valve [8] as well as a reduction of the axially projected area of the tilted prosthesis. As a consequence, systolic energy loss was increased in most cases, imposing additional work load on the ventricle [9]. This in turn delays recovery of myocardial hypertrophy which occurs in long-term valvular disease [5,13]. Tilting might additionally promote a size mismatch of the prosthesis. These deleterious effects may finally reduce life-expectancy [5]. 4.5 Diastolic dysfunction Tilting caused a remarkable increase in diastolic regurgitation which resulted in a decline of effective stroke volume and cardiac output. This effect has two contributions. First, tilting led to an increase of closure volume due to a delayed and asymmetrical closure of the leaflets. The latter is determined by a passive movement that correlates with fluid velocity, maximum opening angle, friction, and inertial forces [3,9,19]. However, gravity also affects valve behavior showing little effect as long as the valve is horizontally positioned. Wu [19] demonstrated in vitro that oblique implantation of a bileaflet valve leads to delayed closure of the upper leaflet due to unequal influence of gravitational forces. In our model, the tilted valve might have been similarly affected showing asynchronous leaflet closure. Furthermore, the tilted valve was exposed to the bloodstream at a different angle compared to the untilted position. Asynchronous closure may therefore be additionally explained by a change in the flow pattern resulting in higher turbulences; yet we did not apply any visualization techniques such as PIV [21]. Second, a moderate rise in leakage volume was also observed in the tilted position. The reason for this remains speculative since no visualization data were collected. No gain in leakage interval occurred so that incomplete leaflet closure may account for this result. Besides, leakage may be overestimated since we used low-viscosity fluid [3,4,9]. Augmented diastolic regurgitation further increases transvalvular energy loss and reduces effective stroke volume, thus contributing to the drop in cardiac output [9]. Ventricular volume load is increased compared with the untilted position, which further impairs postoperative remodeling. Besides, higher regurgitation causes damage to corpuscular blood elements resulting in hemolysis and thromboembolic events [16,20,22,23]. 4.6 Influence of valve size Although deterioration of hemodynamics was seen in all valve sizes, there was a remarkable size dependence in some parameters. The increase of mean transvalvular gradient and resistance was strongest in the 25 mm and smallest in the 21 mm valve. This indicates that the advantage of larger prostheses is severely counterbalanced by tilting. Size-related changes were also observed in the reduction of effective stroke volume and cardiac output suggesting higher reflux in the larger prostheses, which was reflected by corresponding changes in the regurgitation parameters. Tilting showed the greatest loss in effective opening area and Cd in the 25 mm valve. Utilization of the valvular orifice by forward flow is therefore more restricted in the larger prostheses, which might be explained by augmented turbulence formation in the tilted position. No considerable size dependence was seen regarding systolic energy loss. 4.7 Influence of valve rotation Ninety degree rotation proved to be most detrimental because total and effective stroke volume as well as cardiac output showed the strongest impairment in this position, especially in the larger prostheses. Values of mean pressure gradient and resistance were also highest at 90° rotation except for the 23 mm valve. The same is true for the increase of regurgitation fraction. Systolic energy loss showed the strongest changes at 90° in the 21 and 23 mm prostheses. The rotation effect was most pronounced in the 21 mm valve. The 25 mm valve also yielded the most detrimental results in opening area, discharge coefficient, and transvalvular resistance at 90° rotation. However, some exceptions were seen in the 23 mm valve. Forty-five degree rotation showed an intermediate attitude in many, but not all, parameters. Overall, the current data suggest a significant inferiority of the 90° position. In this rotation, the leaflets of the tilted valve are differently exposed to bloodstream and gravity. Opening of the lower leaflet occurs against gravity while closure of the upper leaflet is delayed due to its nearly vertical position. Thus, valvular stenosis and insufficiency are emphasized in this position, whereas at 0° rotation, acting forces seem to be symmetrically distributed. 4.8 Limitations of the study Every in-vitro model, as used in our study, is no more than a rough approximation to the human circulation. Especially the unnaturally shaped artificial ventricle with the straight aortic root does not generate an ideally natural flow profile, although in contrast to many steady-flow models [6,21], the pulsatile flow in our model is an acceptable compromise. Additionally, all mandatory physical components of the circulation were incorporated to achieve physiological conditions. We operated the model using 0.9% saline instead of blood for hygienic reasons and optical transparency [3,14]. As a Newtonian fluid, saline follows exact physical laws. Blood, in contrast, is a Non-Newtonian fluid showing variable viscosity dependent on flow velocity leading to unpredictable changes in flow behavior. Additionally, the opacity impedes visualization of leaflet function. Regarding other viscous media, such as glycerol solution, Werner [14] demonstrated only little influence of the fluid viscosity on systolic parameters. Using 0.9% saline, however, pressure gradients may be slightly underestimated and leakage volumes overestimated [3,4,9,14]. Despite the limitations, an artificial model remains an essential element in circulation research representing a universal instrument with interchangeable components. It provides physically exact measurements free from biological fluctuations and reduces the need for animal experiments [3,14]. 4.9 Conclusions We conclude from the study that even moderate tilting of a bileaflet valve, which often occurs in clinical practice, leads to significant impairment of its hemodynamic properties in terms of systolic and diastolic function. Deleterious effects were demonstrated for the 21, 23, and 25 mm SJM®-AHPJ prostheses. Superiority of large-sized valves, as shown for the regular position, diminished in the tilted position. In contrast to the untilted valve, rotation becomes an important factor in the tilted valve. The tilting effect was most pronounced at 90° rotation but faded at 45° and 0° rotation. Whenever possible, a 90° position should therefore be avoided or surgically corrected. The use of rotatable valves is hence recommended. This project was in part supported by: (1) St. Jude Medical GmbH (Eschborn, Germany); (2) Berlin Heart AG (Berlin, Germany); (3) a 12-month Fellowship by the Foundation of the Friedrich-Schiller-University Jena (supported by Jenoptik AG, Jena, Germany); and (4) the Research Workshop of the University Hospital Jena. 1 2 3 4 5 6[13] Symbols used in the equations: ρ=fluid density; ΔEsys=systolic energy loss; ΔPmean=mean pressure gradient; Cd=discharge coefficient; CO=cardiac output; EOA=effective opening area; GOA=geometrical opening area; HR=heart rate; PV=ventricular pressure; PAO=aortic pressure; QAO=aortic flow; RAO=aortic valvular resistance; SVeff=effective stroke volume; SVtot=total stroke volume; Tsys=systolic time; vAO=aortic flow velocity; VR=regurgitation volume; and RF=regurgitation fraction. Dr J. Hasenkam (Aarhus, Denmark): Did you investigate the effect of tilting the other way around so you had the valve tips in the superannular position in the anterior part of the valve? Dr Hartrumpf: We tried to keep it as simple as possible. We thought it was most likely that the valve was elevated in the non-coronary sinus, because the surgeon usually pays attention to the coronary ostia. There is a study in literature in which negative tilting was investigated, however, just to reduce the number of results we compared positive tilting, which I have shown to you, because we thought it most often occurs in the epiannular implantation position, which we routinely employ at our institution. Dr Hasenkam: If you tilt the valve with the anterior part more downstream than the posterior part, you might even find a positive impact of tilting the valve. Dr Hartrumpf: No, we did not find any positive effects of tilting. We saw an increased obstruction due to the reduced effective opening area, so the systolic function was impaired, and diastolic regurgitation was also increased, and these results have to be regarded as deleterious effects in all cases. Dr F. Mohr (Leipzig, Germany): May I just ask a simple surgical question. Whole changes occur because there is a change in terms of flow direction, is that right, because you tilt the valve and all of a sudden you change the orifice area of the valve in terms of the flow direction of the left ventricular outflow tract? Is that the point? Dr Hartrumpf: Yes. The tilted valve is differently hit by the bloodstream. So if you look at the valve from the view of the bloodstream, the orifice area decreases to an elliptic form, and so the projected valve area becomes smaller. I think the bloodstream is probably deviated by the tilted valve, which might lead to more turbulences. Dr Mohr: However, I have to say that human nature does not always reflect the situation in your model, because very often in those patients who have had long-standing aortic valve disease they have a dilated aorta and the situation is such that the non-coronary sinus most often is dilated, and it may have a real impact. I think the clinical relevance of these findings, not talking about the narrow outflow tract but of a normal sized or even an enlarged aortic root, I think these findings would not really matter from a surgical standpoint. Would you agree with that? Dr Hartrumpf: You are completely right. What we did was to perform a study with an artificial model. Dr Mohr: It is a nice study, no question about that. Dr Hartrumpf: For example, we used straight tubes and we did not account for the complex geometry of the left ventricle, the aortic arch is curved, and so all this remains to be considered in subsequent studies. However, there are data in literature saying that the axis of the left ventricle shows an angle of 140–150° towards the outflow tract and that is why flow is deviated when traveling through the aortic valve. This leads to an asymmetric flow profile, and this will, of course, further affect the results in human. Dr E. Akl (Cairo, Egypt): In practice do you recommend enlarging the root in the clinical status where you have the two options whether to put the valve in a tilted position in the non-coronary sinus or enlargement of the root? Dr Hartrumpf: We would recommend this if it is technically feasible. It is our clinical practice that we sometimes perform a root enlargement in very small aortic roots. However, in the first place we would recommend the use of hemodynamically enhanced valves, such as hemodynamic plus valves. You are right, because size mismatch has to be avoided. There is a study in literature where the orifice area is related to the body surface area, and this is a very important issue. And so I think this is more important than the changes that may occur when tilting the valve. So, therefore, if the root is really small we would recommend to do a root enlargement. Dr A. Moritz (Frankfurt, Germany): I think this data is clinically very important because we know after a number of valve implantations, once you control them, various degrees of gradients exist. We don't know exactly why those gradients develop and your observation may very well bring in a point to explain how this gradients develop, 21 or 23 valve gradients of 50 mm mercury may exist and others have 20. All this develops despite subvalvular myectomies. The angluation of the valve may play a more important role than we think generally. References [1] Edmunds L.H. Jr . Evolution of prosthetic heart valves , Am Heart J , 2001 , vol. 141 5 (pg. 849 - 855 ) Google Scholar Crossref Search ADS PubMed WorldCat [2] Baumgartner F.J. . Baumgartner F.J. . Valvular heart disease , Cardiothoracic surgery , 1999 2nd ed. Georgetown, TX Landes Bioscience (pg. 63 - 96 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [3] Butterfield M. , Fisher J. , Davies G.A. , Spyt T.J. . Comparative study of the hydrodynamic function of the CarboMedics valve , Ann Thorac Surg , 1991 , vol. 52 4 (pg. 815 - 820 ) Google Scholar Crossref Search ADS PubMed WorldCat [4] Fisher J. . Comparative study of the hydrodynamic function of the size 19 and 21 mm St. Jude Medical Hemodynamic plus bileaflet heart valves , J Heart Valve Dis , 1994 , vol. 3 1 (pg. 75 - 80 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [5] Pibarot P. , Dumesnil J.G. . Hemodynamic and clinical impact of prosthesis-patient mismatch in the aortic valve position and its prevention , J Am Coll Cardiol , 2000 , vol. 36 4 (pg. 1131 - 1141 ) Google Scholar Crossref Search ADS PubMed WorldCat [6] Wurzel D. , Panidis I. , Gonzales R. . In vitro continuous wave Doppler gradients of mechanical valves in less than optimal orientations , ASAIO Trans , 1991 , vol. 37 3 (pg. M448 - M451 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [7] Laas J. , Kleine P. , Hasenkam M.J. , Nygaard H. . Orientation of tilting disc and bileaflet aortic valve substitutes for optimal hemodynamics , Ann Thorac Surg , 1999 , vol. 68 3 (pg. 1096 - 1099 ) Google Scholar Crossref Search ADS PubMed WorldCat [8] Kleine P. , Perthel M. , Nygaard H. , Hansen S.B. , Paulsen P.K. , Riis C. , Laas J. . Medtronic Hall Versus St. Jude Medical Mechanical Aortic Valve: downstream turbulences with respect to rotation in pigs , J Heart Valve Dis , 1998 , vol. 7 5 (pg. 548 - 555 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [9] Knott E. , Reul H. , Knoch M. , Steinseifer U. , Rau G. . In vitro comparison of aortic heart valve prostheses. Part 1: mechanical valves , J Thorac Cardiovasc Surg , 1988 , vol. 96 6 (pg. 952 - 961 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [10] Hanle D.D. , Harrison E.C. , Yoganathan A.P. , Allen D.T. , Corcoran W.H. . In vitro flow dynamics of four prosthetic aortic valves: a comparative analysis , J Biomech , 1989 , vol. 22 6–7 (pg. 597 - 607 ) Google Scholar Crossref Search ADS PubMed WorldCat [11] Chandran K.B. . Pulsatile flow past St. Jude Medical Bileaflet Valve. An in vitro study , J Thorac Cardiovasc Surg , 1985 , vol. 89 5 (pg. 743 - 749 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [12] Heiliger R. , Lambertz H. , Geks J. , Mittermayer C. . Hydrodynamic investigation of mechanical bileaflet valves , Artif Organs , 1988 , vol. 12 5 (pg. 431 - 443 ) Google Scholar Crossref Search ADS PubMed WorldCat [13] Braunwald E. . , Heart disease: a textbook of cardiovascular medicine , 1997 5th ed. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [14] Werner S. , Wendt M.O. , Schichl K. , Pohl M. , Koch B. . Testing the hydrodynamic properties of heart valve prostheses with a new test apparatus , Biomed Tech (Berl) , 1994 , vol. 39 9 (pg. 204 - 210 ) Google Scholar Crossref Search ADS PubMed WorldCat [15] Walker D.K. , Scotten L.N. . A database obtained from in vitro function testing of mechanical heart valves , J Heart Valve Dis , 1994 , vol. 3 5 (pg. 561 - 570 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [16] Walker P.G. , Yoganathan A.P. . In vitro pulsatile flow hemodynamics of five mechanical aortic heart valve prostheses , Eur J Cardiothorac Surg , 1992 , vol. 6 Suppl. 1 (pg. S113 - S123 ) Google Scholar Crossref Search ADS PubMed WorldCat [17] Flachskampf F.A. , Weyman A.E. , Guerrero J.L. , Thomas J.D. . Influence of orifice geometry and flow rate on effective valve area: an in vitro study , J Am Coll Cardiol , 1990 , vol. 15 5 (pg. 1173 - 1180 ) Google Scholar Crossref Search ADS PubMed WorldCat [18] Dumesnil J.G. , Yoganathan A.P. . Theoretical and practical differences between the Gorlin Formula and the continuity equation for calculating aortic and mitral valve areas , Am J Cardiol , 1991 , vol. 67 15 (pg. 1268 - 1272 ) Google Scholar Crossref Search ADS PubMed WorldCat [19] Wu Z.J. , Hwang N.H. . Asynchronous closure and leaflet impact velocity of bileaflet mechanical heart valves , J Heart Valve Dis , 1995 , vol. 4 Suppl. 1 (pg. S38 - S49 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [20] Ellis J.T. , Healy T.M. , Fontaine A.A. , Weston M.W. , Jarret C.A. , Saxena R. , Yoganathan A.P. . An in vitro investigation of the retrograde flow fields of two bileaflet mechanical heart valves , J Heart Valve Dis , 1996 , vol. 5 6 (pg. 600 - 606 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [21] Lim W.L. , Chew Y.T. , Chew T.C. , Low H.T. . Steady flow dynamics of prosthetic aortic heart valves: a comparative evaluation with PIV techniques , J Biomech , 1998 , vol. 31 5 (pg. 411 - 421 ) Google Scholar Crossref Search ADS PubMed WorldCat [22] Hung T.C. , Hochmuth R.M. , Joist J.H. , Sutera S.P. . Shear-induced aggregation and lysis of platelets , Trans Am Soc Artif Intern Organs , 1976 , vol. 22 (pg. 285 - 291 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [23] Leverett L.B. , Hellums J.D. , Alfrey C.P. , Lynch E.C. . Red blood cell damage by shear stress , Biophys J , 1972 , vol. 12 3 (pg. 257 - 273 ) Google Scholar Crossref Search ADS PubMed WorldCat © 2002 Elsevier Science B.V. Elsevier Science B.V.
TI - The hemodynamic performance of standard bileaflet valves is impaired by a tilted implantation position
JF - European Journal of Cardio-Thoracic Surgery
DO - 10.1016/s1010-7940(02)00804-7
DA - 2003-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/the-hemodynamic-performance-of-standard-bileaflet-valves-is-impaired-NF9XQVbdqC
SP - 283
EP - 291
VL - 23
IS - 3
DP - DeepDyve
ER -