Hemorrhagic and thrombotic complications are major concerns during neonatal extracorporeal membrane oxygenation (ECMO).1 Hemorrhagic complications are reported in 4–13% of neonatal patients placed on ECMO for respiratory failure (Extracorporeal Life Support Organization Registry). Thrombotic complications are also commonly reported (2–18.7% of neonatal respiratory ECMO runs); however, the clinical significance of small clots within the extracorporeal circuit is unknown. Bleeding can be overt at surgical sites or covert (intracerebral). Both can be troublesome, but the latter is likely to have a greater effect on patient outcome. The maintenance of the hemostatic mechanism as nearly normal as possible has a major role in the prevention of such events.2 Together with platelet transfusion, the restoration of factor activity (by the administration of blood products) is the mainstay of procoagulation therapy during ECMO.3 The effects of systemic inflammatory response, bypass, therapeutic anticoagulation, and development upon coagulation are complex.4–7
When an assist device is needed for recovery from heart failure, optimal conditions require: 1) careful choice of assistance, 2) absence of infection, and 3) control of hemostasis because of the strong and permanent aggression created in all the systems involved. Apart from single-center experience, no well defined consensus and/or protocol is available for neonatal and pediatric ECMO. Empirical antifibrinolytic, anticoagulant, or antiaggregant treatments do not prevent serious bleeding or frequent thromboembolic accidents.
The purpose of this report is to describe a novel anticoagulation strategy with continuous intravenous antithrombin III infusion and intermittent heparin infusion in order to reduce bleeding and hemorrhagic complication during the early stages of pediatric ECMO.
Materials and Methods
From November 2004 through February 2006, an observational study design was performed on 11 patients who required ECMO for postcardiotomy cardiorespiratory failure.
All circuits were phosphorylcholine-coated (Physio, Dideco, Mirandola) from venous to the arterial cannula. The starter priming (around 150 ml) was composed of saline solution 0.9% and circulated in a closed loop. Before connection to the patient, we replace it with homologous blood to obtain a hematocrit value of 35%. Overall mean duration time of support was 112 hours (range 68–192 hours). Age at surgery, body surface area, cardiopulmonary bypass time, crossclamp, and ECMO support time were not different in the study groups. All patients received a pulsatile ECMO with the use of Deltastream DP1 pump (Medos, Stolberg, Germany) and new-generation oxygenators characterized by the presence of a polymethylpentene hollow-fiber membrane (Dideco, Mirandola, Italy). The following anticoagulation protocol was used in the first five patients:
- Heparin neutralization was achieved with protamine (1:1) upon discontinuation of cardiopulmonary bypass (CPB), before switch to ECMO circuits.
- Continuous heparin infusion (10–20 UI/kg/h) was started via the oxygenator (target ACT was 180–200 seconds) at the end of surgery.
- Bolus of antithrombin III was administered according the following formula:
- (Desired level – Initial level) × Body weight (kg) Target antithrombin III level was not < 60%.
- Tranexamic acid was continuously infused 1–3 mg/kg/h up to 96 hours.
- Hematocrit levels were maintained between 35% and 40% by continuous infusion of red blood concentrate at 5 ml/h up to target value.
- Platelet counts were maintained at not < 100.000 mm3 by continuous infusion of concentrate platelets at 5 ml/h up to target value.
Since April 2005, anticoagulation management was modified as follows and used in the last 6 patients:
- Continuous antithrombin III infusion was started immediately after surgery based on the lab test. Antithrombin III target level was maintained higher than 100%. The following formula was used:
- (Desired level – Initial level) × Body weight (kg)
- Antithrombin value was checked at 4-hour intervals.
- Heparin infusion was started when the antithrombin III value was > 100%, and remained stable for more than 12 hours, and the amount of bleeding smaller to 2 ml/kg for more than 3 consecutive hours. Heparin infusion was started at 2 UI/kg/h via the venous line (target ACT was not < 150 seconds).
- Blood samples were taken before ECMO, on full flow, and after 4, 12, and 24 hours; platelet counts hematocrit, prothrombin time (PT), activated partial antithrombin time (aPTT), fibrinogen, plasmatic AT III, and activated clotting time (ACT) were measured. In all patients, thromboelastography (TEG, Figure 1) was used as a guide to anticoagulation administration every 12 hours with native and heparinase tracings (Figure 2).
Three patients in the first group died; one patient died on ECMO due to peritonitis and two died after 3 and 4 days, respectively, because of persistent pulmonary hypertension and major hemorrhagic neurologic complication. Eight patients were weaned and discharged. The third, fourth, and fifth patients of the first group required surgical revision for bleeding. One experienced minor neurologic sequelae. Neither surgical revision nor thromboembolic complications occurred in the second group treated under the new anticoagulation management protocol.
In all patients, during the first 24 hours of support, residual heparin was detected on thromboelastography (Figure 3). In the first group, trend analysis of plasmatic AT III value during the first 48 hours of support showed high variability of consumption. Bolus replacement of AT III in this group was associated with overt acute bleeding.
In the last six patients, according to thromboelastography evidence, we decided to use continuous AT III infusion with the highest plasmatic concentration target value (>100%) in order to avoid overworking of residual heparin and to obtain, on native blood sample TEG, an R-time ≥ 1200 seconds. This approach has also translated into a stable trend of AT III value during ECMO.
After ECMO removal, the exploration of circuits, pump, and oxygenator showed no evidence of macroscopic clots in high-risk sites, confirmed by macroscopic analysis.
During ECMO support, a compromise is required between the prevention of hemorrhagic complications and the prevention of thrombosis. Although the patients are anticoagulated, activation of clotting mechanisms occurs both within the extracorporeal circuit and within the vasculature as a consequence of endothelial dysfunction.8 Additionally, a balance exists among procoagulant, inhibitory, and fibrinolytic processes.
In our experience, which is based on the experience of La Pitiè,9 the additional use of thromboelastography as a guideline in anticoagulation management gave us the opportunity to study the different phases of coagulation. TEG (Haemoscope) is a whole blood test of viscoelastic blood clot formation that has been used in many different clinical scenarios to diagnose coagulation abnormalities. Within 10 to 20 minutes, information is obtained about the integrity of the coagulation cascade, platelet function, platelet-fibrin interactions, and fibrinolysis. Whole blood (360 μl) is placed into an oscillating cuvette. A piston connected to a transducer and oscillograph is immersed into the blood sample. The movement of the piston becomes coupled to the oscillating cuvette as the blood clots. Clotting generates a signature tracing with the following parameters: reaction time, coagulation time, angle, maximum amplitude, and amplitude 60 minutes after the maximal amplitude.
The R-time represents period of time of latency from start of test to initial fibrin formation. The R-time is affected by several factors: factor deficiency, anticoagulation (presence of heparin), severe hypofibrinogenemia, severe thrombocytopenia, and hypercoagulability syndromes. Postcardiotomy indication together with the use of coated circuits and oxygenators showed the presence, in all ECMO recipients, of a “high” residual rate of heparin, despite low ACT.
In a recent study10 conducted on 10 patients with a focus on coagulation factor activity during neonatal ECMO, the authors found low AT III levels. They reported a mean plasmatic AT III value during the first 24 hours of 27% with an overall incidence of overt or covert bleeding of 30% (3/10). In that series, intraventricular hemorrhage represented the major cause of death.
Administration of AT III has been proposed during neonatal ECMO,8,11 but no further data are available in current literature regarding its use during this type of mechanical circulatory support system.
Antithrombin III is an alpha2-glycoprotein of MW 58000, and is normally present in the human plasma at a concentration of approx. 12.5 mg/dl. It is a serpin (serine protease inhibitor, Figure 4) that inactivates a number of enzymes from the coagulation system, namely the activated forms of Factor X, Factor IX, Factor II (thrombin), Factor VII, Factor XI, and Factor XII. Its affinity for these molecules (i.e., its effectivity) is enhanced by heparin. It is a potent inhibitor of the coagulation cascade and a non–vitamin K-dependent protease. AT III activity is markedly potentiated by heparin; potentiation of its activity is the principle mechanism by which both heparin and low-molecular-weight heparin produce anticoagulation. The presence of heparin increases the activity of AT III by 104–105. According to this, in our experience, the choice to use continuous administration instead of bolus reduced the risk of activation of heparin and both overt and covert bleeding.
Although the name antithrombin implies that it works only on thrombin, it actually serves to inhibit virtually all of the coagulation enzymes to at least some extent. Its ability to limit coagulation through multiple interactions makes it one of the primary natural anticoagulant proteins. Antithrombin acts as a relatively inefficient inhibitor on its own. However, when it is able to bind with heparin, the speed with which the reaction that causes inhibition occurs is greatly accelerated; this makes the antithrombin-heparin complex a vital component of coagulation. This interaction is also the basis for the use of heparin and low-molecular-weight heparins as medications to produce anticoagulation.
The small number of patients in this study and observational design adopted combined with complexity of the clinical situations remains the major limitation of such a study. Initial experience with this anticoagulation strategy, using continuous intravenous antithrombin and intermittent heparin infusion, reduced significantly both overt and covert bleeding complications during the first 48 hours. This has translated into excellent overall outcomes.
1. Sell L, Cullin M, Grant, et al
: Hemorrhagic complications during extracorporeal membrane oxygenation: Prevention and treatment. J Pediatr Surg
2. Hirthler M, Blackwell E, Adde D: Coagulation parameter instability as an early predictor of intracranial hemorrhage during extracorporeal membrane oxygenation. J Pediatr Surg
27: 40–43, 1992.
3. Graves DF, Chermin JM, Kurusz M, Zwischberger JB: Anticoagulation practices during extracorporeal membrane oxygenation: Survey results. Perfusion
11: 461–466, 1996.
4. Andrew M, Paes B, Johnston M: Development of the human coagulation system in the neonate and young infants. Am J Pediatr Hematol Oncol
12: 95–104, 1990.
5. Dietrich W: Cardiac surgery and the coagulation system. Curr Opin Anaesthesiol
13: 27–34, 2000.
6. Kern F, Nicholas J, Sears J, et al
: Coagulation defects in neonates during cardiopulmonary bypass. Ann Thorac Surg
54: 541–546, 1992.
7. Cooper MM, Elliot MJ: Haemodiluition, in Jonas RA, Elliot MJ (eds), Cardiopulmonary Bypass in Neonates, Infants and Young Children.
Oxford, Butterworth-Heinemann, 1994, pp. 82–89.
8. Urlesberger B, Zobel G, Zenz, et al
: Activation of the clotting system during extracorporeal membrane oxygenation in term new born infants. J Pediatr
9. Szefner J: Control and treatment of hemostasis in cardiovascular surgery. The experience of La Pitiè Hospital with patients on total artificial hearts. Int J Artif Organs
18: 633–648, 1995.
10. Arnold P, Jackson S, Wallis J, et al
: Coagulation factor activity during neonatal extracorporeal membrane oxygenation. Intensive Care Med
27: 1395–1400, 2001.
11. Pollack ME, Owings JT, Gosselin RC: AT III replacement during infant extracorporeal support. Thromb Hemost
73: 936, 1995.