Assessment of the Flow Field in the HeartMate 3 Using Three-Dimensional Particle Tracking Velocimetry and Comparison to Computational Fluid Dynamics

Turbulent flow field in maglev centrifugal blood pumps of CH-VAD and HeartMate III: secondary flow and its effects on pump performance

Secondary flow path is one of the crucial aspects during the design of centrifugal blood pumps. Small clearance size increases stress level and blood damage, while large clearance size can improve blood washout and reduce stress level. Nonetheless, large clearance also leads to strong secondary flows, causing further blood damage. Maglev blood pumps rely on magnetic force to achieve rotor suspension and allow more design freedom of clearance size. This study aims to characterize turbulent flow field and secondary flow as well as its effects on the primary flow and pump performance, in two representative commercial maglev blood pumps of CH-VAD and HeartMate III, which feature distinct designs of secondary flow path. The narrow and long secondary flow path of CH-VAD resulted in low secondary flow rates and low disturbance to the primary flow. The flow loss and blood damage potential of the CH-VAD mainly occurred at the secondary flow path, as well as the blade clearances. By contrast, the wide clearances in HeartMate III induced significant disturbance to the primary flow, resulting in large incidence angle, strong secondary flows and high flow loss. At higher flow rates, the incidence angle was even larger, causing larger separation, leading to a significant decrease of efficiency and steeper performance curve compared with CH-VAD. This study shows that maglev bearings do not guarantee good blood compatibility, and more attention should be paid to the influence of secondary flows on pump performance when designing centrifugal blood pumps.

Transient Performance Analysis of Centrifugal Left Ventricular Assist Devices Coupled With Windkessel Model: Large Eddy Simulations Study on Continuous and Pulsatile Flow Operation

Computational fluid dynamics (CFD) simulations are widely used to develop and analyze blood-contacting medical devices such as left ventricular assist devices (LVADs). This work presents an analysis of the transient behavior of two centrifugal LVADs with different designs: HeartWare VAD and HeartMate3. A scale-resolving methodology is followed through Large Eddy Simulations, which allows for the visualization of turbulent structures. The three-dimensional (3D) LVAD models are coupled to a zero-dimensional (0D) 2-element Windkessel model, which accounts for the vascular resistance and compliance of the arterial system downstream of the device. Furthermore, both continuous- and pulsatile-flow operation modes are analyzed. For the pulsatile conditions, the artificial pulse of HeartMate3 is imposed, leading to a larger variation of performance variables in HeartWare VAD than in HeartMate3. Moreover, CFD results of pulsatile-flow simulations are compared to those obtained by accessing the quasi-steady maps of the pumps. The quasi-steady approach is a predictive tool used to provide a preliminary approximation of the pulsatile evolution of flow rate, pressure head, and power, by only imposing a speed pulse and vascular parameters. This preliminary quasi-steady solution can be useful for deciding the characteristics of the pulsatile speed law before running a transient CFD simulation, as the former entails a significant reduction in computational cost in comparison to the latter.

Effect of impeller rotational phase on the FDA blood pump velocity fields

BACKGROUND

The Food and Drug Administration (FDA) blood pump is an open-source benchmark cardiovascular device introduced for validating computational and experimental performance analysis tools. The time-resolved velocity field for the whole impeller has not been established, as is undertaken in this particle image velocimetry (PIV) study. The level of instantaneous velocity fluctuations is important, to assess the flow-induced rotor vibrations which may contribute to the total blood damage.

METHODS

To document these factors, time-resolved two-dimensional PIV experiments were performed that were precisely phase-locked with the impeller rotation angle. The velocity fields in the impeller and in the volute conformed with the previous single blade passage experiments of literature.

RESULTS

Depending on the impeller orientation, present experiments showed that volute outlet nozzle flow can fluctuate up to 34% during impeller rotation, with a maximum standard experimental uncertainty of 2.2%. Likewise, the flow fields in each impeller passage also altered in average 33.5%. Considerably different vortex patterns were observed for different blade passages, with the largest vortical structures reaching an average core radii of 7 mm. The constant volute area employed in the FDA pump design contributes to the observed velocity imbalance, as illustrated in our velocity measurements.

CONCLUSIONS

By introducing the impeller orientation parameter for the nozzle flow, this study considers the possible uncertainties influencing pump flow. Expanding the available literature data, analysis of inter-blade relative velocity fields is provided here for the first-time to the best of our knowledge. Consequently, our research fills a critical knowledge gap in the understanding of the flow dynamics of an important benchmark cardiovascular device. This study prompts the need for improved hydrodynamic designs and optimized devices to be used as benchmark test devices, to build more confidence and safety in future ventricular assist device performance assessment studies.

Haemodynamic significance of extrinsic outflow graft stenoses during HeartMate 3™ therapy

Aims

The HeartMate 3 (HM3) implantable left ventricular assist device connects the left ventricle apex to the aorta via an outflow graft. Extrinsic obstruction of the graft (eOGO) is associated with serious morbidity and mortality and recently led to a Food and Drug Administration Class 1 device recall of HM3. This study aimed to provide a better understanding of the haemodynamic impact of extrinsic stenoses.

Methods and results

Computed tomography (CT) images of two retrospectively identified patients with eOGO (29 and 36% decrease in cross-sectional area, respectively, by radiological evaluation) were acquired with a novel photon-counting CT system. Numerical evaluations of haemodynamics were conducted using a high-fidelity 3D computational fluid dynamics approach on both the patient-specific graft geometries and in two virtually augmented stenotic severities and three device flows. Visual analysis identified increased velocity, pressure, and turbulent flow in the outer anterior curvature of the outflow graft; however, changes in graft pressure gradients were slight (1-9 mmHg) across the range of stenosis severities and flow rates tested.

Conclusion

Evidence of eOGO during HM3 support and the recent device recall can provoke clinical apprehension and interventions. The haemodynamic impact of a stenosis detected visually or by quantification of cross-sectional area reduction may be difficult to predict and easily overestimated. This numerical study suggests that, for clinically encountered flow rates and stenosis severities below 61% in cross-sectional area decrease, eOGO may have low haemodynamic impact. This suggests that patients without symptoms or signs consistent with haemodynamically significant obstruction might be managed expectantly.

In-Vitro Flow Validation of Third-Generation Ventricular Assist Devices: Feasibility and Challenges

Computational fluid dynamics (CFD) is a powerful tool for the in-silico evaluation of rotodynamic blood pumps (RBPs). Corresponding validation, however, is typically restricted to easily accessible, global flow quantities. This study showcased the HeartMate 3 (HM3) to identify feasibility and challenges of enhanced in-vitro validation in third-generation RBPs. To enable high-precision acquisition of impeller torques and grant access for optical flow measurements, the HM3 testbench geometry was geometrically modified. These modifications were reproduced in silico , and global flow computations validated along 15 operating conditions. The globally validated flow in the testbench geometry was compared with CFD-simulated flows in the original geometry to assess the impact of the necessary modifications on global and local hydraulic properties. Global hydraulic properties in the testbench geometry were successfully validated (pressure head: r = 0.999, root mean square error [RMSE] = 2.92 mmHg; torque: r = 0.996, RMSE = 0.134 mNm). In-silico comparison with the original geometry demonstrated good agreement ( r > 0.999, relative errors < 11.97%) of global hydraulic properties. Local hydraulic properties (errors up to 81.78%) and hemocopatibility predictions (deviations up to 21.03%), however, were substantially affected by the geometric modifications. Transferability of local flow measures derived on advanced in-vitro testbenches toward original pump designs is challenged by significant local effects associated with the necessary geometrical modifications.

A mathematical model for assessing shear induced bleeding risk

PURPOSE

The objective of this study is to develop a bleeding risk model for assessing device-induced bleeding risk in patients supported with blood contact medical devices (BCMDs).

METHODS

The mathematical model for evaluating bleeding risk considers the effects of shear stress on von Willebrand factor (vWF) unfolding, high molecular weight multimers-vWF (HMWM-vWF) degradation, platelet activation and receptor shedding and platelet-vWF binding ability. Functions of the effect of shear stress on the above factors are fitted/employed and solved by the Eulerian transport equation. An axial flow-through Couette device and two clinical VADs which are HeartWare Ventricular Assist Device (HVAD) and HeartMate II (HM II) blood pump were employed to perform the simulation to evaluate platelet receptor shedding (GPIbα and GPIIb/IIIa), loss of HWMW-vWF, platelet-vWF binding ability and bleeding risk for validating the accuracy of our model.

RESULTS

The platelet-vWF binding ability after being subjected to high shear region in the axial flow-through Couette device predicted by our bleeding model was highly consistent with reported experimental data. As indicated by our CFD simulation results in the axial flow-through Couette device, it can find that an increase in shear stress led to a decrease in the adhesion ability of platelets on vWF, while the binding ability of vWF with platelets first increase and then decrease as shear stress elevates gradually beyond a threshold. The factor of exposure time can enhance the effect of shear stress. Additionally, the shear-induced bleeding risk predicted by our model increases with increasing shear stress and exposure time in an axial flow-through Couette device. As indicated by our numerical model, the bleeding risk in HVAD was higher than HMII, which is highly consistent with the meta-analysis based on clinical statistics. Our simulation investigations in these two clinical VADs also found that HVAD caused a higher rate of platelet receptor shedding and lower damage to HWMW-vWF than HeartMate II. The high shear stress generated in the narrow and turbulent regions of both VADs was the underlying cause of device-induced bleeding.

CONCLUSION

In this study, the shear-induced bleeding risk predicted by our bleeding model in axial flow-through Couette device and two clinical VADs is consistent or highly correlated with experimental and clinical findings, which proves the accuracy of our bleeding model. Our bleeding model can be used to aid the development of new BCMDs with improved functional characteristics and biocompatibility, and help to reduce risk of device-induced adverse events in patients.

Hemocompatibility and hemodynamic comparison of two centrifugal LVADs: HVAD and HeartMate3

Mechanical circulatory support using ventricular assist devices is a common technique for treating patients suffering from advanced heart failure. The latest generation of devices is characterized by centrifugal turbopumps which employ magnetic levitation bearings to ensure a gap clearance between moving and static parts. Despite the increasing use of these devices as a destination therapy, several long-term complications still exist regarding their hemocompatibility. The blood damage associated with different pump designs has been investigated profoundly in the literature, while the hemodynamic performance has been hardly considered. This work presents a novel comparison between the two main devices of the latest generation-HVAD and HM3-from both perspectives, hemodynamic performance and blood damage. Computational fluid dynamics simulations are performed to model the considered LVADs, and computational results are compared to experimental measurements of pressure head to validate the model. Enhanced performance and hemocompatibility are detected for HM3 owing to its design incorporating more conventional blades and larger gap clearances.

Preclinical evaluation of the fluid dynamics and hemocompatibility of the Corheart 6 left ventricular assist device

BACKGROUND

Corheart 6 (Corheart) is a newly developed magnetically levitated continuous-flow left ventricular assist device currently undergoing multicenter clinical trials in China. Featuring a small size, minimal weight, and low power consumption, the Corheart aims to improve pump hemocompatibility, reduce adverse events, and enhance the quality of life of heart failure patients.

METHODS

Computational simulations assessed flow field, shear stress, and washout, while in vitro and in vivo experiments were performed to further demonstrate hemocompatibility.

RESULTS

Numerical results show that the flow path in the Corheart blood pump is well designed. There is no significantly high shear stress in the majority of the flow domain. Short secondary flow paths and small pump size (small priming volume) provide good washing (0.049 and 0.165 s to remove 55% and 95% old blood, respectively), allowing low hemolysis levels both in computational and in vitro hemolysis tests (in vitro hemolysis index ranges from 0.00092 ± 0.00006 g/100 L to 0.00134 ± 0.00019 g/100 L). Good hemocompatibility was further evidenced by ten 60-day sheep implants tested with relatively low flow rates of 2.0 ± 0.2 L/min; the results showed no hemolysis or thrombosis.

CONCLUSIONS

Numerical and experimental results shed light on the fluid dynamics characteristics and hemocompatibility of the Corheart. It is believed that the Corheart will provide more promising possibilities for minimally invasive implantation techniques and for those patients with a small body surface area.

CFD analysis of the HVAD's hemodynamic performance and blood damage with insight into gap clearance

Mechanical circulatory support using ventricular assist devices has become commonplace in the treatment of patients suffering from advanced stages of heart failure. While blood damage generated by these devices has been evaluated in depth, their hemodynamic performance has been investigated much less. This work presents the analysis of the complete operating map of a left ventricular assist device, in terms of pressure head, power and efficiency. Further investigation into its hemocompatibility is included as well. To achieve these objectives, computational fluid dynamics simulations of a centrifugal blood pump with a wide-blade impeller were performed. Several conditions were considered by varying the rotational speed and volumetric flow rate. Regarding the device's hemocompatibility, blood damage was evaluated by means of the hemolysis index. By relating the hemocompatibility of the device to its hemodynamic performance, the results have demonstrated that the highest hemolysis occurs at low flow rates, corresponding to operating conditions of low efficiency. Both performance and hemocompatibility are affected by the gap clearance. An innovative investigation into the influence of this design parameter has yielded decreased efficiencies and increased hemolysis as the gap clearance is reduced. As a further novelty, pump operating maps were non-dimensionalized to highlight the influence of Reynolds number, which allows their application to any working condition. The pump's operating range places it in the transitional regime between laminar and turbulent, leading to enhanced efficiency for the highest Reynolds number.

Flow assessment as a function of pump timing of tubular pulsatile pump for use as a ventricular assist device in a left heart simulator

INTRODUCTION

Although mechanical circulatory support saved many lives during the last decade, clinical observations have shown that the continuous flow pumps are associated with a much higher incidence of gastrointestinal bleeding and kidney problems, among others, compared with the earlier generation pulsatile pumps. However, the presence of several moving mechanical components made pulsatile pumps less durable, bulky, and prone to malfunction, ultimately leading to bias in favor of continuous flow designs.

OBJECTIVE

The aim of the current work is to create a prototype tubular pulsatile pump and to test the timing of the pump in a left heart simulator.

METHODS

A left heart simulator to mimic pumping from a failing heart was created. This was used to experimentally test the output of a prototype ventricular assist device relative to a failing heart in the form of flow and pressure. The effect of pulsation timing was quantified.

RESULTS

A failing heart was simulated with an average flow rate of 1.1 L/min and a systolic pressure of 47 mmHg. With the pump, the flow rate increases to 4.8 L/min and a systolic pressure of 110mmHg, in a copulsation mode, while activating for 300-400 ms. If the activation time is reduced, or increased, the pump becomes less effective. Load on the heart is reduced when the pump operates in a counterpulsation mode.

CONCLUSION

A pulsatile pump, like the one proposed, provides adequate output for mechanical circulatory support, while minimizing the number of moving parts that could otherwise lead to tribological wear.

Hemolytic Footprint of Rotodynamic Blood Pumps

OBJECTIVE

In preclinical examinations, rotodynamic blood pumps (RBPs) are predominantly evaluated at design-point conditions. In clinical practice, however, they run at diversified modes of operation. This study aimed at extending current preclinical evaluation of hemolytic profiles in RBPs toward broader, clinically relevant ranges of operation.

METHODS

Two implantable RBPs the HeartMate 3 (HM3) and the HeartWare Ventricular Assist Device (HVAD) were analyzed at three pump speeds (HM3: 4300, 5600, 7000rpm; HVAD: 1800, 2760, 3600rpm) with three flow rates (1-9L/min) per speed setting. Hemolysis measurements were performed in heparinized bovine blood. The delta free hemoglobin (dfHb) and the normalized index of hemolysis (NIH) served as hemolytic measures. Statistical analysis was performed by multiple comparison of the 9 operating conditions. Moreover, computational fluid dynamics (CFD) was applied to provide mechanistic insights into the interrelation between hydraulics and hemolysis by correlating numerically computed hydraulic losses with in-vitro hemolytic measures.

RESULTS

In both devices, dfHb increased toward increasing speeds, particularly during low but also during high flow condition. By contrast, in both RBPs magnitudes of NIH were significantly elevated during low flow operation compared to high flow conditions (p<0.0036). Maps of hemolytic metrics revealed morphologically similar trends to in-silico hydraulic losses (r>0.793).

CONCLUSIONS

While off-design operation is associated with increased hemolytic profiles, the setting of different operating conditions render a preclinical prediction of clinical impact with current hemolysis metrics difficult.

SIGNIFICANCE

The identified increase in hemolytic measures during episodes of off-design operation is highlighting the need to consider worst-case operation during preclinical examinations.

Validation of Numerically Predicted Shear Stress-dependent Dissipative Losses Within a Rotary Blood Pump

Computational fluid dynamics find widespread application in the development of rotary blood pumps (RBPs). Yet, corresponding simulations rely on shear stress computations that are afflicted with limited resolution while lacking validation. This study aimed at the experimental validation of integral hydraulic properties to analyze global shear stress resolution across the operational range of a novel RBP. Pressure head and impeller torque were numerically predicted based on Unsteady Reynolds-averaged Navier-Stokes (URANS) simulations and validated on a testbench with integrated sensor modalities (flow, pressure, and torque). Validation was performed by linear regression and Bland-Altman analysis across nine operating conditions. In power loss analysis (PLA), in silico hydraulic power losses were derived based on the validated hydraulic quantities and balanced with in silico shear-dependent dissipative power losses. Discrepancies among both terms provided a measure of in silico shear stress resolution. In silico and in vitro data correlated with low discordance in pressure (r = 0.992, RMSE = 1.02 mmHg), torque (r = 0.999, RMSE = 0.034 mNm), and hydraulic power losses (r = 0.990, RMSE = 0.015W). PLA revealed numerically predicted dissipative losses to be up to 34.4% smaller than validated computations of hydraulic losses. This study confirmed the suitability of URANS settings to predict integral hydraulic properties. However, numerical credibility was hampered by lacking resolution of shear-dependent dissipative losses.

Hemodynamic evaluation and in vitro hemolysis evaluation of a novel centrifugal pump for extracorporeal membrane oxygenation

Background

The STM CP-24 I centrifugal pump is a newly developed centrifugal pump for extracorporeal membrane oxygenation equipment. This study aimed to combine hydraulic experiments, hemodynamic numerical simulations, and standard in vitro hemolysis experiments to investigate the comprehensive performance of this centrifugal pump.

Methods

In vitro experiments were first done to obtain the pressure-flow data of the centrifugal pump in its working range to evaluate its hydraulic performance. Next, the commonly used clinical working points were selected as boundary conditions, and a computational fluid dynamics method was applied to evaluate its hemodynamic performance. The blood pressure distribution, blood flow fields, and high-wall-shear-stress zones in the centrifugal pump were determined as indicators for hemodynamic evaluation. Finally, standard in vitro hemolysis experiments were performed to test the blood compatibility of this centrifugal pump (n=3 blood samples). In addition, its blood compatibility was evaluated in the form of the normalized index of hemolysis (NIH).

Results

The pressure-flow curve of the centrifugal pump showed that the head pressure and flow of the centrifugal pump showed a mostly linear relationship within the whole working range. When the rotation speed of the centrifugal pump was 5,500 rpm, it achieved a hydraulic performance of 550 mmHg head pressure and 8 L/min output flow, which could meet the clinical needs of extracorporeal membrane oxygenation. Analysis of computational fluid dynamics data indicated that the centrifugal pump had excellent hemodynamic performance: even distribution of blood pressure in the pump, no blood flow stagnation zone or dead zone in the overall flow field, and secondary flows in the gap between the rotor and the volute that significantly reduced the volume of the low-blood-flow zone close to the impeller. There was no obvious high-shear-stress zone on the surface of the volute or the impeller, which will effectively reduce the risk of thrombosis. In vitro hemolysis experiments indicated that the centrifugal pump had excellent blood biocompatibility, with a NIH =0.0125±0.0022 g/100 L.

Conclusions

The STM CP-24 I centrifugal pump has excellent hydraulic performance, a reasonable design of the hemodynamic structure of the blood pump, and excellent blood compatibility. Therefore, it can meet clinical needs.

Thrombotic Risk of Rotor Speed Modulation Regimes of Contemporary Centrifugal Continuous-flow Left Ventricular Assist Devices

Contemporary centrifugal continuous-flow left ventricular assist devices (LVADs) incorporate dynamic speed modulation algorithms. Hemocompatibility of these periodic unsteady pump operating conditions has been only partially explored. We evaluated whether speed modulation induces flow alterations associated with detrimental prothrombotic effects. For this aim, we evaluated the thrombogenic profile of the HeartWare ventricular assist device (HVAD) Lavare Cycle (LC) and HeartMate3 (HM3) artificial pulse (AP) via comprehensive numerical evaluation of (i) pump washout, (ii) stagnation zones, (iii) shear stress regimens, and (iv) modeling of platelet activation status via the platelet activity state (PAS) model. Data were compared between different simulated operating scenarios, including: (i) constant rotational speed and pump pressure head, used as reference; (ii) unsteady pump pressure head as induced by cardiac pulsatility; and (iii) unsteady rotor speed modulation of the LC (HVAD) and AP (HM3). Our results show that pump washout did not improve across the different simulated scenarios in neither the HVAD nor the HM3. The LC reduced but did not eliminate flow stagnation (-57%) and did not impact metrics of HVAD platelet activation (median PAS: +0.4%). The AP reduced HM3 flow stagnation by up to 91% but increased prothrombotic shear stress and simulated platelet activation (median PAS: +124%). Our study advances understanding of the pathogenesis of LVAD thrombosis, suggesting mechanistic implications of rotor speed modulation. Our data provide rationale criteria for the future design optimization of next generation LVADs to further reduce hemocompatibility-related adverse events.

High-speed visualization of ingested, ejected, adherent, and disintegrated thrombus in contemporary ventricular assist devices

Biocompatibility of ventricular assist devices (VADs) has been steadily improving, yet the rate of neurological events remains unacceptably high. Recent speculation for elevated stroke rates centers on ingestion of thrombi originating upstream of the pump, such as in the ventricle or left atrial appendage. These thrombi may be ejected by the VAD or become deposited within the blood flow pathway, presenting serious complications to the patient. This study was performed to visualize and quantify the degree of disruption, adherence, and disintegration of thrombi that are ingested by the three most implanted VADs: the HeartMate II, HeartMate 3, and HVAD. Clot analogs of varying microstructure compositions (red, white) and sizes (0.5, 1, 2 cm³ ) were synthesized in vitro based on clinical explant data. These were introduced individually into an in vitro flow loop with a transparent replica of the HMII, HM3, and HVAD operated at nominal steady flow (2.3-4.0 L/min). High-speed videography (up to 10 000 fps) revealed the ingestion, disruption, ejection, and adherence of thrombus fragments. Thromboemboli of varying compositions and sizes were observed mechanically attaching to components in all 3 VAD models. In some instances, ingested thrombi physically obstructed portions of the blood flow path; 18% (3 of 17 total) of red thrombi adhered to the inflow straightener of the transparent HMII. In the HVAD model, fewer than 4% of clots were adherent or trapped within the pump, irrespective of microstructure or initial volume. In comparison, 100% (4 of 4 total) of 1-cm³ white (fibrin) clots became lodged within the transparent HM3 while, in contrast, less than 5% of macerated red clots (3 of 63 total) of the same volume were adherent inside the pump. A significant proportion of ingested thrombi were macerated into infinitesimal fragments; 84% and 74% of 2-cm³ red thrombi in the HVAD and HM3 models, respectively, were found to have disintegrated upon ingestion. However, large emboli were also discharged from both centrifugal VADs; these fragments, ranging from 0.01 to 0.29 cm³ regardless of microstructure and original volume, may be capable of occluding an intracranial vessel. Therefore, ingested thrombus may explain, in part, elevated stroke rates in contemporary blood pumps in the absence of adherent pump thrombosis.