Decreased erythrocyte aggregation in Glenn and Fontan: univentricular circulation as a rheologic disease model

Abstract

Background: In the Fontan palliation for single ventricle heart disease (SVHD), pulmonary blood flow is non-pulsatile/passive, low velocity, and low shear, making viscous power loss a critical determinant of cardiac output. The rheologic properties of blood in SVHD patients are essential for understanding and modulating their limited cardiac output and they have not been systematically studied. We hypothesize that viscosity is decreased in single ventricle circulation.

Methods: We evaluated whole blood viscosity, red blood cell (RBC) aggregation, and RBC deformability to evaluate changes in healthy children and SVHD patients. We altered suspending media to understand cellular and plasma differences contributing to rheologic differences.

Results: Whole blood viscosity was similar between SVHD and healthy at their native hematocrits, while viscosity was lower at equivalent hematocrits for SVHD patients. RBC deformability is increased, and RBC aggregation is decreased in SVHD patients. Suspending SVHD RBCs in healthy plasma resulted in increased RBC aggregation and suspending healthy RBCs in SVHD plasma resulted in lower RBC aggregation.

Conclusions: Hematocrit corrected blood viscosity is lower in SVHD vs. healthy due to decreased RBC aggregation and higher RBC deformability, a viscous adaptation of blood in patients whose cardiac output is dependent on minimizing viscous power loss.

Impact: Patients with single ventricle circulation have decreased red blood cell aggregation and increased red blood cell deformability, both of which result in a decrease in blood viscosity across a large shear rate range. Since the unique Fontan circulation has very low-shear and low velocity flow in the pulmonary arteries, blood viscosity plays an increased role in vascular resistance, therefore this work is the first to describe a novel mechanism to target pulmonary vascular resistance as a modifiable risk factor. This is a novel, modifiable risk factor in this patient population.

Fig. 1. Whole blood viscosities as measured by two different viscometers were not significantly different between Glenn, Fontan and healthy control groups.

Solid lines = healthy, dotted lines = Glenn, dash-dot lines = Fontan. The x-axes of both graphs are log transformed. a Whole blood viscosity at native hematocrit from 34 Glenn, 61 Fontan and 40 healthy controls was measured by an upright U-tube viscometer (Rheolog), over a range of shear rates from 1 to 1000 s⁻¹. There was not a significant difference between each group. b Whole blood viscosity at native hematocrit from 12 Glenn, 63 Fontan and 40 healthy controls was measured using a cone/plate viscometer (Brookfield) over a range of shear rates from 37.5 to 1500 s⁻¹. There was not a significant difference between healthy and Fontan or between Glenn and Fontan at any shear rate, however, Glenn viscosity was significantly lower than healthy viscosity at shear rates 75, 150, and 750 s⁻¹ (*p < 0.05) and trended to be lower at shear rate of 300 s⁻¹ (#p = 0.0704).

Fig. 2. Viscosities in SVHD groups were lower than in healthy controls when adjusted to the same hematocrits.

Solid lines = healthy, dash lines = Glenn, dash-dot lines = Fontan. Whole blood viscosity was measured at 20, 30, 40, 50 and 60% hematocrits at each shear rate in Brookfield cone/plate viscometer. In all graphs, viscosity in healthy group was significantly higher than both Fontan and Glenn groups (*p < 0.0002). a At a shear rate of 35 s⁻¹, Fontan viscosity was higher than Glenn at hematocrits of 40 and 60% (#p = 0.0536, ‡p = 0.0055). b At a shear rate of 75 s⁻¹, Fontan viscosity was higher than Glenn at hematocrits of 20, 50, and 60% (#p = 0.0405, %p = 0.0599, ‡p = 0.0266). c At shear rate of 150 s⁻¹, Fontan blood viscosity was higher than Glenn at 60% hematocrit (#p = 0.0548). d, e At shear rates of 300 and 750 s⁻¹, Fontan and Glenn viscosities were not significantly different. f At a shear rate of 1500 s⁻¹, Fontan blood viscosity was significantly higher than Glenn at 20% hematocrit (#p = 0.0393).

Fig. 3. Hematocrit to viscosity ratios (HVR) were higher in SVHD groups than in healthy controls.

Solid lines = healthy, dash lines = Glenn, dash-dot lines = Fontan. HVR is calculated by dividing the hematocrit by the viscosity measured at each shear rate in Brookfield cone/plate viscometer. In all graphs, HVR in healthy group was significantly lower than both Fontan and Glenn groups (*p < 0.0002). a At a shear rate of 35 s⁻¹, Glenn HVR was significantly higher than Fontan at hematocrits of 40, 50, and 60% (#p = 0.059, ‡p = 0.0043). b At a shear rate of 75 s⁻¹, Glenn HVR was significantly higher than Fontan at hematocrits 20, 50, and 60% (%p = 0.0332, #p = 0.0561, ‡p = 0.0158). c At shear rate of 150 s⁻¹, Glenn HVR was significantly higher than Fontan at a hematocrit of 60% (#p = 0.05). d, e At shear rates of 300 and 750 s⁻¹, Glenn and Fontan HVR were not significantly different. f At a shear rate of 1500 s⁻¹, Glenn HVR was significantly higher than Fontan at a hematocrit of 20% (#p = 0.019).

Fig. 4. RBC deformability in SVHD were significantly higher than in healthy group at most shear stresses between 0.5 and 50 Pa.

40 healthy, 62 Fontan and 33 Glenn subjects are included here. a % denotes p ≤ 0.05 comparing healthy and Fontan, * denotes p ≤ 0.05 comparing healthy and Glenn, and ‡ denotes p ≤ 0.05 comparing Glenn and Fontan. Elongation index (EI) of each RBC was measured at shear stresses of 0.5, 0.89, 1.58, 2.81, 8.89, 15.81, 28.12, 50 Pa at 37 °C. b Shear stress at half maximal deformation (SS_1/2) is plotted here with each dot representing a single subject. c While there is no significant difference between healthy and Fontan groups, Glenn EImax (maximum EI) was significantly higher than the two groups.

Fig. 5. RBC aggregation in SVHD groups were both lower than healthy controls.

Each dot represents an individual subject. Each subject was measured at 40% hematocrit with their RBCs in autologous plasma and also in 3% dextran. RBC aggregation index was measured using Myrenne cone/plate aggregometer at stasis, M (a) and low-shear, 3 s⁻¹, M1 (b). 39 healthy, 62 Fontan and 30 Glenn subjects were plotted in these graphs. c RCB aggregation index was measured using Lorrca ektacytometry. 37 healthy, 47 Fontan and 12 Glenn subjects were plotted in these graphs.

Fig. 6. RBC aggregation was lower in both Glenn (panels a and c) and Fontan (panels b and d) and was increased when RBC from Glen and Fontan was resuspended in plasma from healthy controls.

Each line represents a paired comparison within an individual subject. All RBC aggregation measurements are done with 40% hematocrit. Total number of subjects were 12 in each group; however, only data from blood-type compatible matches are included here. Paired t-test was used to evaluate the significance in changes in RBC aggregation.