An experimental and computational study of the inferior vena cava hemodynamics under respiratory-induced collapse of the infrarenal IVC

Highlights

• Particle image velocimetry experiments and numerical simulations predict that different levels of inferior vena cava (IVC) collapse that occur physiologically during normal breathing and Valsalva influence IVC hemodynamics.

• A high degree of IVC collapse (50%) generates large WSS (i.e. 5–10 Pa) that have been shown to initiate thrombus formation and platelet deposition in vitro studies.

• Furthermore, computational simulations points out that IVC hemodynamics are sensitive to small variations in the iliac vein flow splitting.

• From a clinical perspective, the results could further elucidate the respiratory-induced IVC hemodynamic changes and the occurrence of IVC filter tilt and migration, and potentially contribute to the determination of the optimal IVC filter design and robust criteria for its positioning.

Abstract

Inferior vena cava (IVC) filters have been used for over five decades as an alternative to anticoagulation therapy in the treatment of venous thromboembolic disease. However, complications associated with IVC filters remain common. Though many studies have investigated blood flow in the IVC, the effects of respiration-induced IVC collapse have not been evaluated. Our hypothesis is that IVC collapse may have an influence on IVC filter performance. Therefore, we herein investigate the hemodynamics in uncollapsed and collapsed IVC configurations using in vitro flow experiments and computational simulations.

Particle image velocimetry (PIV) is used to measure the hemodynamics in an idealized, compliant model of the human IVC made of silicone rubber. Flow is studied under uncollapsed and collapsed scenarios, with the minor diameter of the IVC reduced by 30% in the collapsed state. Both rest and exercise flow conditions are investigated, corresponding to suprarenal flow rates of 2 lpm and 5.5 lpm, respectively. Finite element analysis simulations are carried out in a computational model of the undeformed, idealized IVC to reproduce the 30% collapse configuration and an additional 50% collapse configuration. Computational fluid dynamics (CFD) simulations are then performed to predict the flow in the uncollapsed and collapsed scenarios, and CFD results are compared to the experimental data.

The results show that the collapsed states generate a higher velocity jet at the iliac junction that propagates farther into the lumen of the vena cava in comparison to the jet generated in the uncollapsed state. Moreover, 50% collapse of the IVC causes a shift of the jet away from the IVC wall and towards the center of the vena cava lumen. The area of maximum wall shear stress occurs where the jet impacts the wall and is larger in the collapsed scenarios. Secondary flow is also more complex in the collapsed scenarios.

Interestingly, this study demonstrates that a small variation in the flow rate distribution between the right and left iliac veins induces significant variations in the flow characteristics. We speculate that asymmetries in the flow may generate unbalanced forces on the IVC wall and on placed IVC filters that could promote filter tilting and migration, although this requires further investigation. If unbalanced forces are present in vivo, the forces should be considered when determining the optimal placement positions and geometric features for IVC filters. Therefore, these findings motivate further investigation of the in vivo hemodynamics in the infrarenal IVC.

Introduction

Pulmonary embolism (PE) is the third most common cause of death from cardiovascular disease, after heart attack and stroke [1]. Usually associated with deep vein thrombosis (DVT), pulmonary embolism occurs when a venous thrombus embolizes and passes through the inferior vena cava (IVC) to the pulmonary arteries. Emboli occluding blood flow to one or both lungs can lead to impaired oxygenation, heart strain, and death [2].

IVC filters have been used for over five decades as an alternative to anticoagulation therapy to prevent PE. These devices are designed to intercept and trap large emboli before they reach the lungs while allowing blood to bypass freely. However, complications associated with IVC filters remain common, including failure in embolus capture, wall perforation, filter tilt or migration, and filter fracture [3].

As demonstrated by several in vitro [4], [5], [6] and computational [7], [8], [9], [10], [11], [12], [13] studies, an in-depth understanding of the IVC hemodynamics is crucial to assess both potential risks and benefits of IVC filters. On the one hand, in vitro studies [4], [5], [6] have investigated the fluid dynamics in IVCs partially occluded by a placed IVC filter and captured emboli. These studies revealed stagnant flow and recirculation regions downstream of the partially occluded IVC, which may promote further thrombus formation. On the other hand, computational studies have simulated the hemodynamics in unoccluded and partially occluded IVCs [7], [8], [9], [10], [11], [12], embolus transport and capture [8], filter positioning in idealized [13] and patient-specific IVC geometries [10], [13]. Two studies [14], [15] combined both in vitro experiments and computational fluid dynamics (CFD) to deepen the understanding of the hemodynamic performance of different IVC filters. Notably, Stewart and colleagues [14] were the first to use a compliant vena cava model for particle image velocimetry (PIV) flow visualization. They compared flow velocities between PIV and CFD, using the latter also to quantify IVC wall shear stresses (WSS).

Nonetheless, the hemodynamics during respiratory-induced partial collapse of the IVC that occurs during normal breathing and Valsalva maneuver have not yet been evaluated. We hypothesize that partial collapse of the IVC may have an important influence on the IVC hemodynamics. Indeed, during inspiration, positive pressure in the abdomen causes the infrarenal IVC to diminish in size, while during expiration a negative pressure causes the infrarenal IVC to expand [16]. Valsalva has an effect similar to that of inspiration, with even greater positive pressure and, thereby, greater collapse of the IVC [16].

The IVC collapsibility index (IVC-CI) quantifies the effect of normal breathing and Valsalva on the IVC cross-sectional shape and is defined as the difference between the end-expiratory (D_exp) and end-inspiratory (D_insp) minor diameter of the IVC, divided by the D_exp [16], [17], [18], [19]. Clinically, the IVC-CI is used to evaluate a patient's intravascular volume status and can range from zero (no collapse during respiration, a possible indication of hypervolemia) to unity (full collapse during breathing, a possible indication of hypovolemia) [20]. Reported healthy (i.e., asymptomatic) IVC-CI values are approximately 0.3 to 0.4 during normal breathing (e.g., see control groups in [21], [22]) and 0.50 + /‒ 0.04 during Valsalva [16].

We investigate the impact of different levels of IVC collapse under both resting and exercise flow conditions by performing PIV measurements on a compliant model of the human IVC accompanied by numerical CFD simulations. In all experiments and simulations, the complex dynamics of IVC collapse are simplified by approximating the vessel walls as rigid and the flow as quasi-steady. The combination of in vitro and computational approaches allows us to partially validate our numerical CFD results, thus providing greater confidence in the additional fields (e.g wall shear stress [WSS]) that are relatively easily extracted from CFD results, but are difficult to obtain directly through PIV.