Characterization of Autoantibodies to PF4/Heparin in a Rat Model of Cardiopulmonary Bypass.

Patients undergoing cardiopulmonary bypass (CPB) represent a high-risk population for developing life-threatening complications of Heparin-Induced Thrombocytopenia and Thrombosis (HITT). Approximately 30-70% of patients undergoing CPB develop heparin dependent autoantibodies to PF4/heparin, which can provoke disease in susceptible individuals. To address the clinical relevance of bypass associated anti-PF4/heparin, we have created a rat model of autoimmune PF4/heparin in the setting of CPB. Sprague-Dawley rats were prepared for CPB using cannulae in the right external jugular vein for venous return to the CPB circuit and the ventral tail artery for arterial inflow. Rats were administered 0.3–0.4U/g heparin i.v. after cannulation, and were then connected to a miniaturized CPB system complete with a membrane oxygenator primed with 40 ml of blood from 2 syngeneic donors (each receiving 0.3U/g i.v. heparin). CPB was maintained for 75 minutes with a flow between 160–180 mls/kg/min (normal cardiac output in the rat) after which animals were decannulated and sacrificed for study over the ensuing 6 weeks. Control rats were subjected to a sham operation for which they received the same dose of heparin, as well as arterial and venous catheterization, but were not placed on CPB. Blood was collected from control animals or rats after CPB and tested for autoantibodies to rat PF4/heparin (rP+H) by ELISA. Rats were studied either after sham operation (n=6) or after CPB at the following time points: 15 minutes (n=15), 3 days (n=6), 16 days (n=25), and at 6 weeks (n=8). Rats undergoing CPB had markedly elevated anti-rP+H levels (as gauged by mean A450 nm ± SEM) 16 days after the procedure as compared to rats receiving sham operation (0.211 ± 04, 16 day CPB vs. 0.01 ± 0.01sham; P = 0<.0001 by ANOVA). High titer antibody responders from day 16 {Rat 77 (A450nm=0.84), Rat 80 (A450nm=0.81) and Rat 70 (A450nm=0.46)} were characterized for reactivity to rPF4 alone, rP+H or rP+H in the presence of excess heparin. Anti-rP+H from all three animals showed maximal reactivity to rP+H, but reduced binding to rPF4 alone or rat P+H in the presence of excess heparin. We additionally tested functional effects of anti-rP+H on rat platelets and human platelets, using a 14C-SRA. Rat platelets were not activated by anti-rP+H or a human HIT-like monoclonal antibody, KKO confirming the absence of activating FcγRII on rat platelets. Human platelets were activated by KKO (60% release), but not rat anti-P+H (<5% release). In conclusion, we demonstrate that autoantibodies occur in heparinized animals subjected to the platelet activating effects of CPB. This animal model will be critical for studying the many facets of bypass- related heparin sensitization, including: the immunobiology of HITT, non-FcR mediated thrombotic effects of anti-P+H, and for developing alternative therapies for HITT patients undergoing CPB.