764

Abstract

Introduction: Heparin-induced thrombocytopenia (HIT) is a serious adverse effect of heparin treatment characterized by formation of antibodies to heparin–platelet factor 4 (H–PF4) complexes. Clinically, patients with HIT demonstrate a fall in platelet count of more than fifty percent, occurring 5–10 days after the start of heparin treatment; some also develop HIT related thrombosis (HITT). Cardiopulmonary bypass (CPB) is associated with platelet activation, thrombocytopenia and the release into the plasma of large amounts of PF4, as well as exposure to high doses of unfractionated heparin (UFH). Up to 50% of patients develop antibodies to H–PF4 complexes following CPB. However, less than 3% of those patients develop actual HITT, characterized by a second episode of thrombocytopenia 5-10 days after CPB ("P1 pattern" reported by Pouplard et al.). ECMO is a modified CPB circuit that critically ill patients may be exposed to for days and weeks. The incidence of HIT in patients undergoing ECMO is unknown. Furthermore, clinical scores used to diagnose HIT may not be useful in ECMO patients given the high prevalence and persistence of thrombocytopenia caused by artificial circuit and underlying disease. We sought to determine the incidence of HIT/HITT and platelet patterns potentially suggestive of HIT in a cohort of ECMO patients at our institution Methods: After IRB approval, a prospectively collected data set encompassing all ECMO patients from 2009 to 2012 at our institution was reviewed. To compare platelet trends, we only included patients who survived longer than 48 hours after ECMO initiation. Patients were tested for HIT based on presence of thrombocytopenia and clinical suspicion. Institutional criteria defines patients as "HIT positive" if they meet any of the following serological criteria: HIT ELISA Ab(+)/SRA(+); HIT ELISA Ab(-)/SRA(+); HIT ELISA(+) with optical density (OD)>1. "HIT positive" patients with evidence of thrombosis were considered as having HITT. Patients who didn't meet these criteria, or were not tested due to lack of clinical suspicion, were considered "HIT negative". Daily platelet counts were recorded for 10 days inclusive of day of ECMO initiation. Statistical analysis was done using the Wilcoxon rank sum test and Fisher exact test as appropriate for the data sets. Results: 85 patients underwent ECMO from 2009 to 2012. 68 (80%) patients survived longer than 48 hours after ECMO initiation. Of those, 50 patients(73.5%) were tested for HIT ELISA, SRA or both. 4 patients were deemed "HIT positive" and 3 of those patients had documented thrombo-embolic events (HITT) based on predefined criteria (HIT and HITT incidence in the study group of 5.9% and 4.4% respectively). HIT positive patients were significantly younger (53.1 vs. 37.5 years, p=0.048) and had a trend toward higher average HIT ELISA OD (1.246 vs. 0.258, p=0.056). Only significantly different average platelet count between "HIT negative" and "HIT positive group" was at day 10 (163 vs. 53.7, p=0.024).This difference remained significant when we excluded patients from "HIT negative" group who were not tested for HIT ELISA, SRA or both (127.9 vs. 53.7, p=0.033). No other platelet parameter, such as average highest platelet count, platelet nadir, days to nadir and percentage platelet drop, differed significantly between the groups. However, plots of average daily platelet counts suggests a potentially revealing pattern in "HIT positive" group, where a gradual drop to platelet nadir (post ECMO day 5-6) is followed by a period of persistently depressed platelet counts (days 6-10). Conclusions: Incidence of both HIT and HITT is relatively high in patients who survive ECMO for longer than 48 hours at our institution. Traditional clinical scores for diagnosis of HIT such as "4 Ts", or presence of "P 1" platelet pattern (typical of HIT in cardiac surgery patients) are not useful in ECMO patients. Pattern of depressed platelet counts after nadir may be suggestive of HIT in these patients.