The Importance of Hemorheology and Patient Anatomy on the Hemodynamics in the Inferior Vena Cava

Abstract

Inferior vena cava (IVC) filters have been used for nearly half a century to prevent pulmonary embolism in at-risk patients. However, complications with IVC filters remain common. In this study, we investigate the importance of considering the hemorheological and morphological effects on IVC hemodynamics by simulating Newtonian and non-Newtonian blood flow in three IVC models with varying levels of geometric idealization. Partial occlusion by an IVC filter and a thrombus is also considered. More than 99% of the infrarenal IVC volume is found to contain flow in the nonlinear region of the shear rate-viscosity curve for blood (less than 100 s) in the unoccluded IVCs. Newtonian simulations performed using the asymptotic viscosity for blood over-predict the non-Newtonian Reynolds numbers by more than a factor of two and under-predict the mean wall shear stress (WSS) by 28-54%. Agreement with the non-Newtonian simulations is better using a characteristic viscosity, but local WSS errors are still large (up to 50%) in the partially occluded cases. Secondary flow patterns in the IVC also depend on the viscosity model and IVC morphological complexity. Non-Newtonian simulations required only a marginal increase in computational expense compared with the Newtonian simulations. We recommend that future studies of IVC hemodynamics consider the effects of hemorheology and IVC morphology when accurate predictions of WSS and secondary flow features are desired.

A comparative CFD study of four inferior vena cava filters

“…Note that the WSS levels in the posterior view in Figure 4B are considerably higher than they are in the anterior part of the wall, shown in Figure 4A. Foci with high or low WSS values are closely related to irregularities of the vein geometry, in agreement with previous results by Aycock et al 17 The corresponding local WSS distributions for other values of blood flow rates, either with the presence or the absence of a filter (not shown here for the sake of brevity) have a similar look as the location and shape of the high/low WSS spots basically coincide. The notion that the presence of a filter does not significantly alter the WSS levels on the IVC wall is confirmed by the surface averaged values, WSS vein , plotted in Figure 5A,B.…”

“…In the present calculations, we used both the Newtonian and a non‐Newtonian viscosity model. We found, in agreement with Aycock et al, that for any given combination of filter model and blood flow rate the non‐Newtonian viscosity model always yielded higher and F d values than it did the Newtonian model. Aycock et al reported that, on the average, the Newtonian model underestimated values by about 29 % .…”

“…For the smallest clot size investigated (250 mm 3 ) Singer et al reported that the presence of the clot increased the peak WSS values on the IVC wall by about a 40%. Aycock et al found that for a bigger clot of 1060 mm 3 the average WSS on the IVC wall increased from 2.22 to 2.53 Pa. Thus, it seems that clots smaller than 250 mm 3 would not significantly alter the blood flow and would produce just a modest increase on the WSS levels on the vein wall.…”

“…Moreover, the presence of the 2 blood clots produces a small increase of the averaged WSS on the whole IVC surface from 1.364 Pa (Figure B) to 1.414 Pa. Although we found, in agreement with previous studies,() that the presence of small clots induces just a small increase in WSS levels a different matter is how it affects the total drag force exerted by the blood flow. For the filter 2 alone this magnitude was F d =45.95 dyne in the nonexercise case, and F d =272.90 dyne in the exercise case (see Figure B,D).…”

“…Some previous studies 12,13,16,17 investigated the effects on the blood flow of a partial occlusion of the IVC caused by a blood clot trapped by the filter. The clot sizes in these studies are typically in the range 250 to 2000 mm 3 .…”