Abstract
Abstract 5201
Left ventricular assist devices (LVAD) are used to improve circulation in heart failure patients until an available transplant. Despite the benefits of these pumps, they have also been shown to cause a higher rate of adverse effects than the customary medical treatment. Specifically, these implants have been shown to heighten the incidence of both bleeding events (such as intestinal bleeds) and thromboembolic events. Previous studies have attributed these episodes to shear stress caused by blood flow through the pump and to contact of the blood with the artificial surfaces of the device. Contact of blood with artificial surfaces can activate the clotting cascade, while shear stress can directly activate platelets and can impact the functionality of von Willebrand factor. Therefore, a testing regimen that can predict the occurrence of adverse events could lead to enhanced clinical benefit of ventricular assist devices.
Patients implanted with the HeartMate II LVAD at Loyola University Medical Center were enrolled in the study. Blood samples are drawn on a monthly basis and at time of an adverse event. These blood samples were analyzed using flow cytometry to assess P-selectin expression and PAC-1 binding (activated GPIIb/IIIa), the ICHOR PlateletWorks assay (Helena Laboratories; Beaumont, TX) and thrombelastography using the Platelet Mapping assay (Haemoscope; Niles, IL). Patient demographic and clinical data was obtained from medical records. Additionally, blood drawn from 5 individuals per group of warfarin alone, aspirin alone, and normal individuals served as controls.
Patients included 15 males and 5 females with a mean age of 53.3 ±13.5 years (range: 23.5 to 70 years) at the time of enrollment. Of the patients 35% were diabetic, 36.7 % were hypertensive, and 45% were hyperlipidemic. Of the 20 patients 18 were on warfarin (mean starting dose of 4.6±2.5 mg) and 19 of 20 patients were on aspirin (81–325 mg/day). Additionally, 60% of patients were treated with an ACE inhibitor, 80% with a β-blocker and 80% with a diuretic. Of the 20 patients, 2 patients had experienced adverse events during the study period. One had a small bowel bleed requiring multiple transfusions. The second had a 3-day GI bleed requiring multi-unit transfusion. By flow cytometry, both patients with bleeding episodes exhibited less of an activation response to platelet agonists ADP, arachidonic acid and TRAP in terms of P-selectin expression and PAC-1 binding than the mean of the stable LVAD group or the normal individuals. In the Platelet Mapping assay, the stable LVAD population exhibited a 31.5 ± 18.4% (mean ± SD) inhibition of ADP-induced activation (p<0.05 vs normals at 15.1 ± 4.8%); the patient with the small bowel bleed exhibited a 65.6% inhibition. In the ICHOR PlateletWorks assay, the stable LVAD population exhibited a 37.3 ± 19.7% inhibition of collagen-induced aggregation (p=0.004 vs normals at 7.9 ± 1.9%); the patients experiencing bleeding events had inhibition levels of 63.7% and 74.2%, respectively. In the PlateletWorks assay, while the stable LVAD population exhibited a 21.2 ± 18.2% inhibition of ADP-induced aggregation (p<0.004 vs normals at 6.1 ± 2.6%), the patient with the GI bleed exhibited an inhibition level of 72.6%. In the PlateletWorks and PlateletMapping assays, several patients exhibited results indicative of low platelet reactivity, but had not experienced a hemorrhagic event to date. These patients will be followed with periodic monitoring of platelet function to determine the predictive nature of each assay for any coagulation disorder.
Through these studies a baseline of normal platelet function has been established for our LVAD patient population. Further, while early in our evaluation the selected assays of PlateletWorks, PlateletMapping, and flow cytometry suggest that it may be possible to identify patients at risk of LVAD-associated bleeding complications.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
