Background
It has been reported that localized high velocity may be recorded by continuous-wave Doppler interrogation through the smaller central orifices of bileaflet mechanical heart valves (BMHV) and that this may result in overestimation of the transvalvular pressure gradient (TPG). However, the prevalence and clinical relevance of this phenomenon remain unclear, particularly for BMHVs in the mitral position. The objective of this in vitro study was to assess the presence and magnitude of localized high velocity in mitral BMHVs as well as its impact on TPG overestimation by Doppler.
Methods
Nine BMHVs were tested under nine different flow conditions (volumes and flow waveforms) in a simulator specifically designed to assess mitral valve hemodynamics. Flow velocity was measured at three different locations (leading edge, midleaflets, and trailing edge) within the central and lateral orifices of the BMHVs using pulsed-wave Doppler. TPG was measured by pulsed-wave and continuous-wave Doppler and by catheterization.
Results
The maximum flow velocity occurred within the central orifice of the BMHV in 61% of the 81 tested conditions. This locally higher velocity within the central orifice predominantly occurred at the leading edge of the prosthesis. Doppler overestimated mean TPG by an average of 5% to 10% compared with catheterization. The magnitude of the localized high velocity and ensuing overestimation of TPG by Doppler was more important at higher mitral flow volumes ( P < .0001) as well as in BMHVs with smaller internal ring diameters ( P < .0001).
Conclusions
This study shows that the flow velocity distribution within the three orifices of mitral BMHVs is not uniform and that higher velocity occurs more frequently, but not always, within the inflow aspect of the central orifice. In most mitral BMHVs and flow conditions, this localized high-velocity phenomenon causes small overestimation of TPGs (<2 mm Hg and <10%) by Doppler and is thus not clinically relevant. However, in small mitral BMHVs exposed to high flow rates, the overestimation of TPG due to localized high velocity could become more important and overlap with the range of gradients found in patients with prosthesis dysfunction or prosthesis-patient mismatch.
Doppler echocardiography is undoubtedly the method of choice to evaluate prosthetic valve function. In particular, the measurement of transvalvular velocity is crucial for the detection and quantification of prosthetic valve obstruction. However, high velocity or gradient is not necessarily proof of intrinsic prosthetic obstruction and may be secondary to prosthesis-patient mismatch, high flow conditions, prosthetic valve regurgitation, or localized high central jet velocity in bileaflet mechanical heart valves (BMHVs). Indeed, the fluid dynamics of BMHVs differ substantially from those of native or bioprosthetic valves. Because BMHVs have three orifices of different sizes, localized high velocity may be recorded by continuous-wave Doppler interrogation through the smaller central orifice. This phenomenon may lead to an overestimation of gradient and a false suspicion of prosthesis dysfunction. However, the exact prevalence as well as the hemodynamic and clinical relevance of this phenomenon remain unclear, particularly for BMHVs in the mitral position. In vivo, it is very difficult to obtain separate measurements of velocity at precise locations within the lateral and central orifices of BMHVs. The use of in vitro simulation and of pulsed activation duplicator in particular enables the independent study of each clinical parameter. The objective of this in vitro study was thus to assess the existence and magnitude of localized high velocity and transvalvular pressure gradients (TPGs) in mitral BMHVs tested under various flow conditions.
Methods
In Vitro Model
For the purposes of this study, several models and sizes of BMHVs were tested under various pulsed flow conditions in a pulse duplicator previously developed and validated to mimic physiologic and pathologic mitral and aortic flows. Briefly, two anatomically shaped deformable silicon molds of the left atrium and ventricle were enclosed inside two separate Plexiglas boxes (called activation boxes; Figure 1 A). The molds were activated (compressed or stretched) by adding or removing a volume of fluid into or from the activation boxes using two computer-controlled gear pumps. An additional gear pump simulated pulmonary artery outflow. Several models of compliance and resistance are added to the system to accurately replicate physiologic pulmonary arterial, left atrial (LA), left ventricular (LV), and systemic arterial pressure conditions. As designed, the pulse duplicator is capable of accurately reproducing several physiologic and pathologic flow and pressure conditions. The test fluid was a mixture of water and glycerol with a density of 1.060 kg/m 3 and dynamic viscosity of 3.8 ± 0.2 cP. Temperature is regulated through a specific thermostat system during the whole experiment to 37 ± 2°C. As the LA model in the duplicator is anatomically shaped, inflow from the pulmonary circulation is divided into four flows by the four asymmetric pulmonary veins ( Figure 1 B). The BMHVs have, by design, three orifices, one central and two lateral. In this study, the LatR orifice of any BMHV was along the atrial septum and right pulmonary veins, and the LatL orifice was along the left pulmonary veins.
Prosthetic Valves
In this study, we tested nine BMHVs inserted in the mitral position of the pulse duplicator ( Table 1 ): (1) five St. Jude Medical Master valves of sizes 19, 21, 23, 25, and 27 mm (St. Jude Medical, St. Paul, MN); (2) three St. Jude Medical Master HP valves of sizes 23, 25, and 29 mm; and (3) one On-X valve of size 27/29 mm (this valve is for both 27-mm and 29-mm mitral annular sizes; On-X Life Technologies, Inc., Austin, TX). Some of these valves (i.e. 19 and 21 mm) are used only in pediatric surgery but were nonetheless included in this study to examine the localized TPG phenomenon over a wide range of valve sizes. We used the internal ring diameter of the BMHV rather than the labeled valve size, given that the latter does not necessarily reflect the true internal diameter of the prosthesis ring. The internal ring diameter was measured from the inscribed circle of three points using a SmartScope contactless gauging machine (MVP200; Quality Vision International, Rochester, NY). We measured both lateral and central orifice areas at the inflow side of the prosthesis using the internal diameter and distances between leaflets when the valve is opened, assuming a circular shape of the prosthesis. The central/lateral orifice area ratio was then calculated from these data as the central orifice area divided by the area of the two lateral orifices.
| Internal ring diameter (mm) | Central/lateral orifice area ratio | Prosthesis model | Label size (mm) | Prosthesis internal ring diameter (mm) | Prosthesis central/lateral orifice area ratio |
|---|---|---|---|---|---|
| 16–19 | 0.28 to 0.31 | MJ-501 | 21 | 16.15 | 0.295 |
| 0.28 to 0.31 | MHPJ-505 | 19 | 16.37 | 0.302 | |
| 0.26 to 0.28 | MHPJ-505 | 21 | 18.5 | 0.277 | |
| 19–22 | 0.26 to 0.28 | MHPJ-505 | 23 | 19.88 | 0.262 |
| 0.26 to 0.28 | MJ-501 | 25 | 20.22 | 0.259 | |
| MHPJ-505 | 25 | 21.87 | 0.270 | ||
| 22–25 | 0.31 to 0.33 | ON-X | 27/29 | 23.4 | 0.328 |
| 0.26 to 0.28 | MHPJ-505 | 27 | 23.58 | 0.265 | |
| 0.28 to <0.31 | MJ-501 | 29 | 24.12 | 0.283 |
The BMHVs were separated into three categories according to ring internal diameter: 16 to 19, 20 to 22, and 23 to 25 mm. The BMHVs were also separated into three categories according to their central orifice/lateral orifice area ratios: 0.26 to <0.28, ≤0.28 to <0.31, and ≤0.31 to 0.33 ( Table 1 ).
Flow Conditions and Measurements
For each BMHV, three mitral flow stroke volumes (35, 50, and 70 mL, corresponding to mean transvalvular flow rates of 71 ± 4, 102 ± 1, and 139 ± 3 mL/sec) and three flow patterns (E/A ratios of 1.5, 1.0, and 0.5) were tested, thus resulting in nine different flow conditions at 70 beats/min, with a diastolic time of 58 ± 0.5% of the cardiac cycle duration. These conditions were denominated C i P j (where C represents mitral flow stroke volume variations [C 1 for 35 mL, C 2 for 50 mL, and C 3 for 70 mL] and P represents mitral flow pattern variations [P 1 for E/A = 0.5, P 2 for E/A = 1, and P 3 for E/A = 1.5) ( Table 2 ). The mean aortic pressure was held constant at 100 mm Hg. In total, there were thus 81 different experimental conditions (9 BMHVs × 9 flow conditions) tested in this study.
| Variable | Condition | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| C 1 P 1 | C 1 P 2 | C 1 P 3 | C 2 P 1 | C 2 P 2 | C 2 P 3 | C 3 P 1 | C 3 P 2 | C 3 P 3 | |
| Mitral diastolic flow volume (mL) | 35 | 35 | 35 | 50 | 50 | 50 | 70 | 70 | 70 |
| E/A ratio | 0.5 | 1.0 | 1.5 | 0.5 | 1.0 | 1.5 | 0.5 | 1.0 | 1.5 |
| Mean aortic pressure (mm Hg) | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 |
| Maximum transvalvular systolic pressure difference at mitral level (mm Hg) | 90 | 90 | 90 | 100 | 100 | 100 | 110 | 110 | 110 |
Measurement of mitral flow rate was performed using a flow probe (Carolina Medical Equipment, Lexington, SC). The probe was positioned immediately upstream from the BMHV, in the mitral position. This setting allows precise control of the flow volumes and patterns.
Catheter Pressure Measurements
LA, LV, and aortic pressure measurements were performed using three MPR-500 catheters (accuracy full scale ±0.5% for −50 to 300 mm Hg; Millar Instruments, Houston, TX). LA pressure measurements were performed 2.5 cm upstream from the mitral BMHV. LV pressure measurements were performed 3 cm downstream of the mitral BMHV’s trailing edge. Aortic pressure measurements were performed 3.5 cm downstream of the aortic valve (using a 23-mm Biocor prosthesis in all experiments; St. Jude Medical). Aortic pressure was maintained within the physiologic range (approximately 120/80 mm Hg) for all experiments ( Table 2 ). Catheter TPG was computed from pressure recordings as the mean difference between LA and LV pressures during the whole diastolic period.
Doppler Echocardiographic Measurements
Flow velocity measurements were performed using an ATL-HDI 5000 ultrasound system with a 2.1-MHz to 2.6-MHz probe (Philips Medical Systems, Andover, MA). This probe is a phased-array P4-2 probe with lateral accuracy ranging from 1 to 3 mm depending on the frequency, the number of piezoelectric components used, and the depth of the region of interest.
To avoid angle-related errors due to ultrasound beam positioning, the echocardiographic probe was fixed parallel to the mitral flow by a specific device. Continuous-wave and pulsed-wave Doppler velocity measurements were performed consecutively. Continuous Doppler measurements were performed by aligning the Doppler beam on the axis passing through the central orifice of the BMHV, as recommended in the guidelines of the American Society of Echocardiography. Maximum mitral flow velocity was measured at the peak of the E or A wave (when the E/A ratio was equal to 0.5) on the Doppler signal. Mean TPGs were derived from Doppler flow velocity measurements using the simplified Bernoulli equation (dP = 4 V 2 ) at each discrete point of the velocity-time integral.
Pulsed-wave Doppler measurements were performed by positioning the sampling volume (size, 1 mm) within each of the three orifices of the BMHV to assess the existence and magnitude of localized TPGs ( Figure 2 ). For each orifice, the sample volume was positioned at three different locations: (1) the leading edge of BMHV leaflets, (2) the trailing edge of leaflets, and (3) midway between the leading edge and the trailing edge of the BMHV leaflets. Pulsed-wave Doppler measurements were performed in pulsed repetition frequency mode at a frequency of about 3,700 Hz, and Doppler gain was adapted to optimize the visualization of velocity profiles. Displacement of the probe and of the sampling volume from one orifice to the other was precisely controlled using a micrometric system and adapted for each prosthesis to place the sampling volume at the middle of the orifice. For each condition and location, maximum velocity and mean TPG were acquired from three pulsed-wave Doppler measurements of velocity and velocity-time integral.
The ratio of central to lateral orifice velocity is defined as the ratio of the maximum central velocity to the average of the two maximum lateral velocities. The ratio of central to lateral orifice TPG is the ratio of the maximum central TPG to the average of the two maximum lateral TPGs.
Statistical Analysis
Data were compared using analysis of variance and t tests as appropriate and are expressed as mean ± SD. The relationship and agreement between the different Doppler and catheter measures of transvalvular flow velocities and TPGs were assessed using simple linear regression and Bland-Altman analyses. A multivariate analysis of variance was used to identify the independent determinants of the central/lateral flow velocity (or TPG) ratio. All statistical analyses were performed using the open-source software R (R Foundation for Statistical Computing, Vienna, Austria).
Results
Location of the Maximum Flow Velocity
Figure 3 shows the distribution of the occurrence of the maximum flow velocity within the three orifices of the BMHV according to BMHV model, BMHV internal ring diameter, mitral diastolic flow volume, and mitral flow E/A ratio. Overall, the maximum flow velocity occurred within the central orifice of the BMHV in 61% of the 81 tested conditions. The highest velocity was observed in the LatR orifice in 28% of the cases and in the LatL orifice in only 11% of the cases. The valve size as documented by the internal diameter of the valve ring and the valve model does not appear to influence the location of the maximum velocity ( Figure 3 A and 3 B). In contrast, the magnitude and pattern of mitral flow have a significant effect on the location of maximum flow velocity. The percentage of occurrence of maximum flow velocity within the central orifice indeed increased with mitral diastolic flow volume: 50% at 35 mL, 63% at 50 mL, and 70% at 70 mL ( Figure 3 C). Interestingly, there was also a shift of the location of maximum velocity from the lateral orifice (LatR) to the central orifice when the E/A ratio increased from 0.5 to 1.0 and to 1.5 (48% in LatR, 67% and 89% in the central orifice, respectively; Figure 3 D).
Figure 4 displays the occurrence of the maximal flow velocity at the three positions of the pulsed-wave Doppler sample volume within the central orifice of the BMHVs (i.e., the leading-edge, midleaflet, and trailing-edge positions). The maximum velocity was most often recorded at the leading edge of the BMHV. Overall, the location of the maximum velocity within the central orifice was 61% at the leading edge, 24% at midleaflet, and 15% at the trailing edge. The percentage of occurrence of maximum velocity at the leading-edge level varied with prosthesis model (67% for the On-X valve, 72% for the St. Jude Medical Master HP valve, and 48% for the St. Jude Medical Master valve; Figure 4 A). It decreased with increasing prosthesis size (71% for 16–19 mm, 67% for 19–23 mm, and 52% for 23–25 mm; Figure 4 B) and increases with increasing mitral diastolic flow volume (44% at 35 mL, 61% at 50 mL, and 78% at 70 mL; Figure 4 C). The E/A ratio had an inverted U–shaped effect on the location of maximum velocity within the central orifice at the leading edge (50% for E/A = 0.5, 81% for E/A = 1.0, and 52% for E/A = 1.5; Figure 4 D). In summary, the highest velocity inside the prosthesis occurred mainly at the leading edge level of the central orifice ( Figure 5 A), and values of the maximum velocities on average were higher in the central orifice at the leading-edge and midleaflet locations ( Figure 5 B) than at other locations in lateral orifices.
Determinants of the Central/Lateral Velocity Ratio
Figure 6 A shows the ratio of central to lateral orifice velocity as a function of BMHV type and internal diameter, mitral diastolic flow volume, and E/A ratio. The overall central/lateral velocity ratio was 1.07 ± 0.13, the central velocity overestimating the lateral velocity by about 7% on average. The central/lateral velocity ratio increased ( P < .0001) from 1.04 at a mitral diastolic flow volume of 35 mL to 1.12 at a mitral diastolic flow volume of 70 mL. The velocity ratio increased significantly ( P < .0001) with the E/A ratio, from 1.02 at E/A = 0.5 to 1.12 at E/A = 1.5. On the other hand, the velocity ratio slightly decreased ( P = .0007) with increasing prosthesis size (1.09 for 16–19 mm and 1.05 for 23–25 mm).
