Heparin is the anticoagulant most commonly used for patients supported with extracorporeal membrane oxygenation (ECMO).1 Heparin potentiates antithrombin-mediated inactivation of factors IXa and Xa by more than three orders of magnitude. Despite systemic heparinization, circuit thrombosis and fibrin deposition in ECMO circuits are common.2 Generation of thrombi in the circuit can cause platelet consumption and “sick circuit syndrome” and has been speculated to be associated with multiorgan dysfunction.3 Alterations in plasma levels of multiple coagulation factors have been noted in patients supported with ECMO, including a relative deficiency of antithrombin III (ATIII).4 This has led to increased debate concerning whether the supplementation of ATIII can alter the variability in heparin response.
In the cardiopulmonary bypass (CPB) literature, increasingly it is recognized that ATIII administration can be used to increase activated clotting times (ACTs). Therapeutic ACTs have been reached by ATIII supplementation in states of relative antithrombin deficiency with a constant heparin dose5,6; however, it has been demonstrated that low antithrombin levels did not preclude therapeutic heparin levels or change the amount of thrombin-antithrombin complexes in the context of CPB when heparin dose was titrated to effect.7 However, the effect of ATIII supplementation on heparin dose among patients supported on ECMO has never been previously reported.
The Extracorporeal Life Support Organization (ELSO) in 2009 made recommendation to maintain ATIII in the normal range (80–100% control) among patients supported on ECMO by replacing with either ATIII concentrate or fresh frozen plasma.8 The evidence for these recommendations was based on few single case reports or small case series, which suggest that heparin infusion rates are reduced with ATIII supplementation.9–11 More recently, Niebler et al. 12 described a paucity of bleeding complications when ATIII was supplemented in patients supported on ECMO, adding credence to the safety of this therapy, although no other evidence suggesting efficacy were demonstrated outside higher ATIII levels the subsequent day. No studies in the context of ECMO had been reported at the time of the ELSO recommendations.
Based on the increasing number of case reports and the seemingly sound physiologic rationale, ATIII supplementation was incorporated into standard management of patients supported on ECMO at our institution in February 2008. Our study objective was to determine the effect of routine ATIII supplementation for activity <70% on 1) heparin dose immediately after ATIII treatment, 2) effective circuit life in ATIII supplemented versus nonsupplemented deployments, 3) unfractionated anti-Xa levels the day after supplementation, and 4) coincident blood product administration.
After receiving approval from the University of Arkansas for Medical Science Institutional Review Board, the hospital ECMO database was queried to identify the study population. All patients aged 0–18 years with a cardiac indication for ECMO between January 1, 2007, and December 31, 2008, were identified for inclusion in the study. Patients who were supported with ECMO for <72 hours were excluded from the study. Based on the inclusion and exclusion criteria, 42 patients who underwent 45 ECMO deployments were identified and their data were collated.
Control and Treatment Groups and Measurements
The study consisted of a patient-based analysis in which circuit life duration was compared—between those who were supplemented with ATIII compared with those who never received ATIII supplementation. For the patient-based analysis, the initial circuit duration was measured in hours. In those patients who received ATIII at some point during their deployment, a day-based analysis was performed—patients were analyzed based on whether they received ATIII that day, and how their heparin dose changed over the next 6 hours to maintain a steady state ACT, and how their anti-Xa level changed over the next 24 hours. On consecutive days, patients may be placed in the treatment group 1 day and the control day the next based on their ATIII levels for the day-based analysis. The change in heparin rate (units/kg/hour) was determined by comparing the mean rate of heparin zero to 3 hours before antithrombin infusion to the mean heparin infusion rate 3 to 6 hours after administration. If no antithrombin was administered, a sham administration time of 1,200 was observed where the mean heparin infusion rate from 0900 to 1200 was compared with mean heparin infusion rate from 1500 to 1800. Binary outcomes of a 10% reduction in heparin infusion rate or a 20% reduction in the heparin infusion rate were recorded.
All information was obtained from the patients electronic or paper medical records. Data regarding coagulation and hematologic laboratory data, hourly heparin dose, unfractionated heparin anti-Xa levels (UF anti-Xa), days on ECMO circuit, and total number of ECMO circuit failures were abstracted from the medical record. Circuit life was measured in hours from the time of ECMO cannulation. When a circuit was replaced, the underlying rationale for circuit change was notated.
Extracorporeal Membrane Oxygenation Deployment
At our institution during the study period, when a decision to institute ECMO support occurred, the in-house ECMO coordinator prepared a circuit consisting of a Stockert S3 or S5 pump (Stockert Instrument, Munich, Germany), a Jostra QuadroxD oxygenator (Maquet, Hirrlingen, Germany), and a heparin-bonded circuit with tubing diameter based on patient size with a length that is minimized and similar regardless of the pump selected. The cardiac surgeon and the surgical team performed ECMO cannulation. Precannulation anticoagulation consisted of heparin 100 units/kg at the time of vessel exposure. Once cannulated, initial ECMO flows varied usually between 100 and 150 ml/kg/minute, and ECMO flow rates were altered to maintain the desired blood pressures and systemic perfusion, while clinically monitoring cerebral and somatic near infrared spectroscopy values, oximetry, and blood lactate levels.
Effective anticoagulation was maintained with a continuous heparin infusion, which was titrated by ACTs between 180 and 220 seconds unless clinical bleeding dictated otherwise. Monitoring included routine ACTs, prothrombin time (PT), international normalized ratio (INR), platelet counts, antithrombin, UF anti-Xa, and fibrinogen levels. Platelet counts were maintained >80,000 unless clinical bleeding or chest tube output mandated a more liberal transfusion threshold. Fibrinogen levels and INR were usually kept in the normal range. Antithrombin III activity and UF anti-Xa levels were monitored daily. If the ATIII activity was <70% of normal values, ATIII (Thrombate III, Grifols Sant Cugat del Valles, Barcelona) was used to supplement to 100% at the discretion of the attending physician. Antithrombin dosage was determined by the following formula: ATIII dose = (100 − ATIII level)/1.4 + circuit factor (based on circuit’s volume).
Circuit failure was defined as mechanical dysfunction or hematologic abnormalities, which necessitated ECMO circuit change. Most commonly, this was recognized as a triad of disseminated clots throughout the ECMO circuit, increased blood product transfusion requirements, and decreasing platelet count, fibrinogen concentration, and rising INR. On occasions, the presence of unstable clots on the arterial limb of the circuit alone would lead to circuit change. The final decision to change the circuit was at the discretion of the clinical team and the ECMO support staff after taking into account the full clinical scenario.
Determining the effect of ATIII supplementation on measures of coagulation was assessed using general linear mixed model (GLMM) regression. Although data were collected on the patient level, several observations were made upon each patient throughout their ECMO course. Subsequently, the GLMM was used to determine differences “between” the treatment groups while accounting for the inherent correlation “within” repeated measurements on the same patient. Both coagulation measures of UF anti-Xa levels and PT could not be assumed to follow a Normal (i.e., Gaussian) distribution; generalized estimating equations assuming a strictly positive and severely skewed underlying distribution (i.e., Gamma) were used to model these two measures.
Effect of ATIII supplementation on the quantity of blood products administered during ECMO support was analyzed using a logistic mixed model regression, which was used to associate any treatment effect on the quantity of blood product administered. All regression models assessing the effect of ATIII supplementation on coagulation and blood product measures included age at ECMO deployment as a covariate; both heparin dose and platelets included premeasurement supplementation-specific covariates as well. Age was modeled using restricted cubic splines determined by four knots to relax the strict linear assumption between covariate and outcome.13
The Cox model also adjusted for age at ECMO and cardiac indication, which were defined as the following independent groups: 1) cardiomyopathy, 2) low cardiac output syndrome status after CPB (<72 hours), and 3) low cardiac output syndrome status after CPB (>72 hours), low cardiac output syndrome with respiratory distress in congenital heart disease without CPB, and low cardiac output because of rejection in cardiac heart transplant patients. Furthermore, the statistical interaction of age and cardiac indication was modeled to account for any differential effect of disease severity and age. Model fit was estimated used Harrell’s concordance statistic.13 Statistical analysis was completed using Stata v12.1 (College Station, TX) and R 2.12.2 (Vienna, Austria).14
Table 1 shows the demographic and deployment comparison of those with and without ATIII supplementation. There were 25 deployments in 21 patients with ATIII supplementation, which were compared with 20 ECMO deployments in 19 patients without ATIII supplemented. There was no significant difference between ATIII supplementation and nonsupplementation groups for age at deployment, duration of deployment, or indication for ECMO (Table 1).
Baseline ATIII levels were greater in the nonsupplemented group when compared with the supplemented group (68.3 vs. 57.0; p < 0.001). Supplementation with ATIII had no statistically significant effect on daily coagulation measures except for UF anti-Xa levels (Table 2). On ATIII supplemented days, on average, UF anti-Xa levels were 0.06 units lower than nonsupplementation days (95% confidence interval [CI] around difference [−0.10, −0.02], p = 0.001). However, supplementation with ATIII resulted in a 0.02 unit Δ anti-Xa (change in UF anti-Xa levels—defined as current UF anti-Xa level minus the immediately prior UF anti-Xa level increased with ATIII supplementation), while days not supplemented was associated with a −0.05 unit Δ anti-Xa. The difference between these units of change is 0.07 (95% CI [0.04, 0.11], p < 0.001). No difference in heparin rate reduction was present during the ATIII-supplemented days.
Administration of ATIII was not significantly associated with blood product administration (Table 3) when ATIII-supplemented days were compared with nonsupplemented days within the group of ATIII-supplemented patients. When ECMO deployments were analyzed as a whole depending on whether ATIII was supplemented at any point during the deployment, ATIII supplementation was independently associated with increased risk of circuit change compared with the nonsupplementation group (hazard ratio = 3.15, 95% CI [1.21, 8.16], p = 0.018) (Table 4).
Figure 1 shows that circuit change was more frequent with ATIII supplementation than without ATIII supplementation (Figure 1). Indication for ECMO of cardiomyopathy at younger ages was also significantly associated with risk of circuit change, but as patients aged any differential effect between cardiac indications disappeared (Figure 2).
Despite the physiologic rationale and prior case reports, we did not observe any significant decrease in heparin infusion rate in the day-based analysis or improved ECMO circuit life in the patient-based analysis with ATIII supplementation in children requiring ECMO for cardiac indications. These results question the rationale for recommendations of routine ATIII supplementation in pediatric ECMO patients requiring ECMO for cardiac indications.
Extracorporeal support is associated with reduction in levels of ATIII and other coagulation-related proteins, such as fibrinogen and von Willebrand factor.15 However, our results suggest that supplementing ATIII had no effect on heparin dosage in a shorter time frame (day-based analysis) and had no effect on circuit duration in those supplemented with ATIII compared with controls on a longer time frame (patient-based analysis). As a multitude of factors are involved in mediating the elaborate interplay of inflammation and hemostatic balance during ECMO support, supplementing a single protein may represent an over simplification of an extraordinarily complex interaction. It can be speculated that the effect of supplementing one coagulation protein may easily have been obviated with the effect of other competing procoagulant and anticoagulant factors within this complex interaction. Furthermore, following the current ELSO recommendations regarding ATIII supplementation would incur significant costs for an intervention for which our study shows no benefit. In addition, supplementing to achieve activity >80% in neonates in whom activity of >40% are considered normal may not be physiologic and will warrant further evaluation, given the lack of obvious benefit.16
Given the mechanism of action of ATIII, our results are contrary to the expected decrease in heparin dose response if the lack of therapeutic anticoagulation is because of low ATIII levels. We can speculate that ATIII supplementation did not result in decrease in heparin dose may be due to 1) increased consumption of supplemented ATIII,17 2) increased elimination/renal clearance of heparin, 3) higher ATIII activity may not be required for optimal heparin effect,18 4) activation of tissue factor pathways,8 5) variations in the concentrations of heparin-binding proteins, and 6) other unrecognized sources of variation.
Estimation of UF anti-Xa levels is increasingly being used as an alternative and possibly more optimal method of monitoring anticoagulation compared with traditional bedside ACT measurements on ECMO and with activated partial thromboplastin time in non-ECMO–related acquired ATIII deficiency.19–22 Monitoring with UF anti-Xa levels has even been implicated in providing a survival benefit.4 However, during the study period, our institutional clinical practice of managing anticoagulation on ECMO was primarily ACT-based heparin titration anticoagulation, which was associated with subtherapeutic UF anti-Xa levels (average UF anti-Xa of 0.22 units/ml; recommended 0.3–0.7 units/ml). Despite the subtherapeutic UF anti-Xa levels in our study cohort, we did find that on days when ATIII was supplemented, the UF anti-Xa levels were significantly lower before ATIII supplementation, suggesting a relatively lesser degree of anticoagulation, but this did not translate into any meaningful clinical difference. However, despite the overall lower UF anti-Xa levels at the time of ATIII supplementation, the Δ anti-Xa, levels actually increased, but again not by a clinically relevant measure. It is possible that a strategy targeting UF anti-Xa for heparin titration, the statistically significant change observed in the study could have translated to significant clinical effects (such as circuit survival).
Contrary to our proposed hypothesis, we demonstrated no improvement in ECMO circuit life with ATIII supplementation. Actually, earlier and significantly more frequent circuit changes were required in the ATIII-supplemented group. There are no other studies investigating the impact of ATIII on circuit life. We can speculate that there may be other contributing factors that have an impact on circuit life, including 1) subtherapeutic anti-Xa levels as observed in our study cohort, 2) inflammatory mediators that caused a prothrombotic state in subjects supplemented with ATIII even though white blood cell count was normal comparable with control subjects, and 3) variation in ECMO circuit beyond oxygenator and ECMO pump. Our initial speculation that age and diagnosis would impact circuit life was not supported by our Cox analysis. Our study could not attempt to account for all variables, which could affect coagulation in this critical population. The propensity toward circuit failure in the ATIII group could possibly be explained by the difference in the two eras’ populations that we were not able to account for outside of age and diagnosis.
This study has limitations: retrospective nature; small number of patients; no set criteria for circuit changes; and most importantly, no means to study the effects of ATIII supplementation on microcirculation. The retrospective nature of the study prevented the patient groups from being well controlled and mandated that data were missing at some time points (<3% of all data points). The nonstandard method of determining the immediate response to antithrombin by monitoring 10% and 20% reduction in the heparin dose is not previously reported; however, it allows for a rational means to determine an end-point on the order of hours following administration. The wide degree of variability in the anti-Xa levels was likely related to the manner in which we manage anticoagulation and has led to changes in anticoagulation management at our institution. We acknowledge that there are multiple factors that influence the decision to change ECMO circuits, which we made no attempt to control for in this study.
Based on our cohort of pediatric patients supported with ECMO for cardiac indications, daily antithrombin supplementation does not enhance coagulation-anticoagulation balance or circuit life. This study will hopefully allow for the clinical equipoise requisite for prospective randomized multicenter clinical studies to evaluate the effectiveness of antithrombin replacement regimens and to understand whether ATIII supplementation would be crucial or not to improve coagulation-anticoagulation balance.
1. Sievert AN, Shackelford AG, McCall MM. Trends and emerging technologies in extracorporeal life support: Results of the 2006 ECLS survey. J Extra Corpor Technol. 2009;41:73–78
2. Muntean W. Coagulation and anticoagulation in extracorporeal membrane oxygenation
. Artif Organs. 1999;23:979–983
3. Arnold P, Jackson S, Wallis J, Smith J, Bolton D, Haynes S. Coagulation factor activity during neonatal extra-corporeal membrane oxygenation. Intensive Care Med. 2001;27:1395–1400
4. Baird CW, Zurakowski D, Robinson B, et al. Anticoagulation and pediatric extracorporeal membrane oxygenation
: Impact of activated clotting time and heparin dose on survival. Ann Thorac Surg. 2007;83:912–919; discussion 919
5. Williams MR, D’Ambra AB, Beck JR, et al. A randomized trial of antithrombin
concentrate for treatment of heparin resistance. Ann Thorac Surg. 2000;70:873–877
6. Lemmer JH Jr, Despotis GJ. Antithrombin
III concentrate to treat heparin resistance in patients undergoing cardiac surgery. J Thorac Cardiovasc Surg. 2002;123:213–217
7. Koster A, Fischer T, Gruendel M, et al. Management of heparin resistance during cardiopulmonary bypass: The effect of five different anticoagulation strategies on hemostatic activation. J Cardiothorac Vasc Anesth. 2003;17:171–175
8. ELSO Guidelines for Cardiopulmonary Extracorporeal Life Support, Extracorporeal Life Support Organization, Version 1:1. April 2009 13. Available at http://www.elso.med.umich.edu/Guidelines.htm
. Accessed March 20, 2012
9. Sievert A, Uber W, Laws S, Cochran J. Improvement in long-term ECMO by detailed monitoring of anticoagulation: A case report. Perfusion. 2011;26:59–64
10. Muedra V, Bonanad S, Gómez M, Villalonga V, Sánchez F, Llopis JE. Relationships between antithrombin
activity, anticoagulant efficacy of heparin therapy and perioperative variables in patients undergoing cardiac surgery requiring cardiopulmonary bypass. Perfusion. 2011;26:487–495
11. Agati S, Ciccarello G, Salvo D, Turla G, Undar A, Mignosa C. Use of a novel anticoagulation strategy during ECMO in a pediatric population: Single-center experience. ASAIO J. 2006;52:513–516
12. Niebler RA, Christensen M, Berens R, Wellner H, Mikhailov T, Tweddell JS. Antithrombin
replacement during extracorporeal membrane oxygenation
. Artif Organs. 2011;35:1024–1028
13. Harrell FE Regression Modeling Strategies. 2001 New York, NY Springer:20–27–79–83
14. Harrell FE. Regression Modeling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. 2001 New York Springer-Verlag
15. Bilen O, Loftis L, Teruya J. Severe thrombotic and bleeding complications in a baby with heterozygous factor V Leiden and acquired von Willebrand disease on ECMO. J Extra Corpor Technol. 2011;43:64–69
16. Andrew M, Paes B, Milner R, et al. Development of the human coagulation system in the full-term infant. Blood
17. Stiller B, Lemmer J, Merkle F, et al. Consumption of blood
products during mechanical circulatory support in children: Comparison between ECMO and a pulsatile ventricular assist device. Intensive Care Med. 2004;30:1814–1820
18. Garvin S, Fitzgerald D, Muehlschlegel JD, et al. Heparin dose response is independent of preoperative antithrombin
activity in patients undergoing coronary artery bypass graft surgery using low heparin concentrations. Anesth Analg. 2010;111:856–861
19. Nankervis CA, Preston TJ, Dysart KC, et al. Assessing heparin dosing in neonates on venoarterial extracorporeal membrane oxygenation
. ASAIO J. 2007;53:111–114
20. Khaja WA, Bilen O, Lukner RB, Edwards R, Teruya J. Evaluation of heparin assay for coagulation management in newborns undergoing ECMO. Am J Clin Pathol. 2010;134:950–954
21. Levine MN, Hirsh J, Gent M, et al. A randomized trial comparing activated thromboplastin time with heparin assay in patients with acute venous thromboembolism requiring large daily doses of heparin. Arch Intern Med. 1994;154:49–56
22. Guervil DJ, Rosenberg AF, Winterstein AG, Harris NS, Johns TE, Zumberg MS. Activated partial thromboplastin time versus antifactor Xa heparin assay in monitoring unfractionated heparin by continuous intravenous infusion. Ann Pharmacother. 2011;45:861–868
Keywords:Copyright © 2014 by the American Society for Artificial Internal Organs
extracorporeal membrane oxygenation; thrombosis (extracorporeal circuit); antithrombin; blood; coagulation/anticoagulation