Coagulation factor deficiencies during initiation of extracorporeal membrane oxygenation

Coagulation factor deficiencies during initiation of extracorporeal membrane oxygenation Michael L. McManus, MD, Sherwin V. Kevy, MD, Lynne K. Bower, RRT,a n d Paul R. Hickey, MD From the Multidisciplinary intensive Care Unit and the Departments of Laboratory Medicine (Hemotherapy Division) and Anesthesia, Children's Hospital, and the Department of Pediatrics, Harvard Medical School, Boston, Massachusetts

Objective: We examined the hypothesis that critically ill patients receiving extracorporeal membrane oxygenation (ECMO) have reduced clotting factor levels, which may contribute to the risk of hemorrhagic complications. Methods: Blood samples were collected from 19 patients before and I hour after initiation of ECMO. Heparin present in samples was removed by ECTEOLA (epichlorohydrin triethanolamine) cellulose resin adsorption, and coagulation factors were assayed by automated techniques. Factor deficiency was defined as levels at least 2 SD less than published age-adjusted reference values. Results: Thirteen patients (68%) had deficiencies of two or more factors before ECMO. Despite inclusion of factor-containing blood products in the ECMO priming solution, 10 patients (53%) had deficiencies of two or more factors after initiation of ECMO. Four patients had intracranial hemorrhages and were found to be deficient in five or more factors at the time of cannulation. Conclusions: Severe coagulation factor deficiencies are often present in patients requiring ECMO, and coagulation factors provided through the circuit prime are insufficient to ensure correction of these deficiencies. Deficiency of multiple coagulation factors may contribute to the risk of intracranial hemorrhage during ECMO; the practice of excluding factor-containing solutions from the circuit prime should be examined prospectively. (J PEDIATR1995; 126:900-4)

Hemorrhage is among the most frequent and serious complications encountered in the use of extracorporeal membrane oxygenation.l-4 The reported incidence of intracranial hemorrhage during ECMO has ranged from as high as 52% in early studies 2 to approximately 15% today.5 Together with chronic lung disease, neurologic injury associated with intracranial hemorrhage is the primary determinant of longterm morbidity in survivors of ECMO, 6 and concern regarding hemorrhage has hampered widespread application of this Submitted for publication Aug. 31, 1994; accepted Dec. 28, 1994. Reprint requests: Michael L. McManus, MD, Multidisciplinary Intensive Care Unit Office, Farley 517, Children's Hospital, 300 Longwood Ave., Boston, MA 02115. Copyright © 1995 by Mosby-Year Book, Inc. 0022-3476/95/$3.00 + 0 9120162988

900

therapy in the care of premature infants and patients with a preexisting coagulopathy.7, 8 Predisposition to hemorrhage in patients receiving ECMO is presumed to be related largely to heparin anticoagulation, but other factors, including preexisting hypoxia, acidosis, carotid ligation, and impaired cerebral autoregulation, 9 may ECMO 1VH PT

Extracorporeal membrane oxygenation Intraventricular hemorrhage Prothrombin time

contribute. Additionally, neonates, the most frequent recipients of ECMO, possess coagulation factor levels substantially less than adult levels, 1°12 so that these patients are particularly vulnerable to hemorrhage on further reduction of coagulation factor concentrations.

The Journal of Pediatrics Volume 126, Number 6

McManus et aI.

901

T a b l e I. Coagulation factor levels* for 19 patients before and after cannulation for ECMO After cannulafion (U/ml)

Before cannulation (U/ml) Factor 1I V VII VIII IX X XI XII

Mean _+SD 0.39 0.50 0.49 0.70 0.42 0.42 0.32 0.24

-+ 0.27 +- 0.35 -+ 0.26 + 0.50 -+ 0.26 -+ 0.24 + 0.19 +_0.18

Mean _+SD

Range 0.05-0.91 0.10-1.47 0.12-0.96 0.03-2.06 0.04-1.04 0.07-0.83 0.01-0.76 0.02-0.64

0,53 0.46 0.50 0.48 0.41 0.52 0.37 0.22

-+ 0.19 -+ 0.27 -+ 0.25 + 0.20 - 0.t5 + 0.19 +_0.16 _+0.13

Range 0.14-1.00 0.10-1.06 0.16-1.18 0.18-0.87 0.12-0.61 0.14-0.94 0.11-0.72 0.02-0.53

*Normal pooled plasma has a factor level of 1.0 U/ml.

Despite the importance of the coagulation system, surprisingly little is known regarding its alterations during ECMO. We hypothesized that, as one basis for predisposition toward hemorrhage, patients receiving ECMO have severe clotting factor deficiencies that precede, and are potentially exacerbated by, initiation of ECMO. We therefore measured the concentrations of clotting factors in plasma taken from patients before and 1 hour after initiation of bypass for ECMO.

METHODS Subjects. Twenty patients, from neonates (>36 weeks of gestational age) to children 3 years of age, who had been referred to our intensive care unit for ECMO were studied. After initial examination the decision to proceed with ECMO was made by the primary medical-surgical team on the basis of institutional criteria (oxygenation index [(Fraction of inspired oxygen x Mean airway pressure) + Arterial oxygen pressure] >0.4, despite maximal therapy). Study consent and consent for ECMO were then obtained from parents in accordance with protocol established by the Children's Hospital Committee on Clinical Investigation. Patients receiving factor-containing blood products, other than those administered according to ECMO initiation protocol, during the 2 hours preceding or hour after initiation of ECMO, were exchided. Extraeorporeal membrane oxygenation. The ECMO circuit priming solution contained 150 ml fresh-frozen plasma, 400 ml packed erythrocytes, 12 ml calcium gluconate, 75 ml tromethamine and 600 U heparin. During cannulation 0.25 U/kg platelets suspended in plasma ( - 1 0 m]/kg) were administered intravenously. Activated clotting time was measured soon after initiation of ECMO, and systemic heparin was administered to maintain activated clotting time between 180 and 200 seconds. No blood products other than packed erythrocytes were given during cannulation or during the first hour of ECMO. Laboratory tests. Blood was obtained from each patient

by way of indwelling arterial catheters before cannulation and 1 hour after the start of ECMO. Samples were collected into citrated tubes on ice and separated by centrifugation at 1700g. Plasma was then treated with ECTEOLA (epichlorohydrin Iriethanolamine) cellulose resin (Sigma Diagnostics, St. Louis, Mo.) for in vitro heparin adsorption before storage at -80 ° C. In control experiments with plasma from normal volunteers, it has been demonstrated that in vitro heparin adsorption successfully reverses the effects of heparin. 13 This was also verified in our laboratory with plasma from normal volunteers (data not shown). Each patient sample was assayed for prothrombin time, fibrinogen, and activities of coagulation factors II, V, VII, VIII, IX, X, XI, and XII with the ACL 300 automated coagulation laboratory (Insmamentation Laboratories, Lexington, Mass.). Before all assays, calibration was accomplished by way of internally generated R values with commercial control plasma and then further validated through parallel analysis of samples fi'om normal volunteers whose factor levels were verified by other techniques. Data analysis. Unless otherwise noted, factor activities are expressed as units per milliliter, where pooled plasma contains 1.0 U/ml. For reference, activities were compared with published age-based control values 1°-12 and deficiency defined as levels less than or equal to 2 SD below the mean for age. Values for pre-ECMO and post-ECMO PT and fibrinogen are presented as means (n = 19) +- SD. Comparisons of mean pre-ECMO and post-ECMO values were made with the two-tailed Student t test.

RESULTS Blood samples were taken from 20 patients; one patient was excluded from analysis because of incomplete data. Fourteen newborn and five pediatric subjects receiving ECMO were enrolled, encompassing a variety of primary diagnoses including congenital diaphragmatic hernia (five patients), meconium aspiration with persistent pulmonary hypertension (four), pneumonia or sepsis (five), acute respi-

902

McManus et aL

The Journal of Pediatrics June 1995

T a b l e II, Deficient coagulation factors before and after

initiation of ECMO Deficient coagulation factors Patient No. 1"

2 3 4 5 6* 7 8* 9 10 11 12 13 14 15 16" 17 18 19

Before cannulation

After cannulation

II, VIII, IX, X, XI, XII II, X, XII, F VIII, XII VII, XII F lI, V, VII, VIII, IX, X, XI, XII VIII, XII II, V, VII. VIII, IX, X, XI, XII F II VIII, IX, XII, F II, V, VII, X, F II, V, VII, F -II, VIII 1I, V, VII, X, F II, V, F

VIII, IX, XI, XII II, X, XI, XII V VII, F V, VIII, F F XII II, V, VII, VIII, IX, XII -V V, IX, XI/, F F V,F XII XII V, VII, XII, F V,F

F

F

XII

V, VII, VIII, XI, XII

Measurements were made from samplestaken immediatelybefore and 1

hour after initiationof ECMO. Deficiencywas definedas levels at least 2 SD less than publishedmeans for age.1° F, Fibrinogen. *Patientswith hemorrhages.Patient6 had a gradeIIVHbeforecannulation but had no further extension(see text). ratory distress syndrome (four), and persistent air leak (one). Overall survival rate was 63%; three patients (16%) died as a direct result of intracranial hemorrhage. Of these, two were newborn infants having signs of hemorrhage by intracranial ultrasound study within 24 hours of cannulation. In the third patient, 3 years of age, a large brain parenchymal hemorrhage developed later in the course of ECMO. One patient with congenital diaphragmatic hernia was cannulated despite the presence of a grade I intraventricular hemorrhage. Although hemorrhage extension was not observed, the patient ultimately died as a result of pulmonary hypoplasia. All newborn infants received vitamin K before ECMO. Mean factor concentrations for all 19 patients before and after initiation of ECMO are presented in Table I. Wide variability in factor concentrations was observed, as evidenced by large ranges and SD. For the population, differences in factor levels before and after ECMO cannulation did not reach statistical significance, and no consistent global pattern of factor deficiency or desmaction was revealed. However, examination of individual patient data revealed frequent and marked abnormalities of coagulation. As detailed in Table II, specific factor deficiencies were identi-

fled in nearly all patients before and after initiation of ECMO. Before cannulation 18 patients (95%) were deficient in at least one clotting factor or fibrinogen, and 13 (68%) in more than one factor as defined by levels that were at least 2 SD less than mean for age. Approximately half the patients studied were deficient in factors lI and XII; one third were found to be deficient in factors V, VII, and X. All four patients with hemorrhagic complications had concurrent deficiencies in five or more factors (range, five to eight factors), whereas all but one of those without hemorrhage were deficient in four or fewer (range, zero to five factors). One hour after initiation of ECMO, deficiencies of at least one factor persisted in 17 patients and a new deficiency of factor IX appeared in one patient (Table II). The overall frequency of most specific factor deficiencies decreased after initiation of ECMO; the incidence of deficiency increased only for factor V. However, 10 of 19 patients (53%) continued to lack adequate concentrations of at least two factors, with the mildest deficiencies (e.g., cases 5 and 19) often worsening. Because of large interpatient variability, overall changes in hematologic profile were not reflected in significant (p <0,05) alterations of the mean PT (18.4 _+ 5.9 seconds vs 18.6 -+ 4.8 seconds) or plasma fibrinogen concentration (1.96 _4-1.20 gm/L vs 1.98 -+ 1.20 gm/L). However, as a gross measure of overall clotting, the number of patients with PT prolonged more than 2 SD beyond the mean for age increased from 10 (53%) to 16 (84%). DISCUSSION

These data demonstrate that critically ill infants and children receiving ECMO often have extensive coagulation factor deficiencies and that the initiation of bypass according to a generally accepted protocol is not consistently a remedy. These heretofore unrecognized factor deficiencies are of sufficient breadth and magnitude as to be clinically worrisome. A role for coagulopathy in the pathogenesis of neonatal intracranial hemorrhage has been suggested, 14-16 yet these discussions have been largely limited to premature infants. Chessells and Wigglesworth 14 found coagulopathies to be more prevalent in newborn infants with IVH than in those without IVH. Later, McDonald et alj6 observed hypocoagulability (defined as hypofibrinogenemia, thrombocytopenia, or prolonged clotting time) in 11 of 15 infants in whom IVH occurred and in only 5 of 35 with no IVH. Few data are available concerning the coagulation status of patients receiving prolonged extracorporeal support, but extrapolation from studies such as these have led many institutions to consider the presence of a coagulopathy to be a contraindication to ECMO. 7, 8 Nevertheless, numerous other factors

The Journal of Pediatrics Volume 126, Number 6

are believed to contribute to the risk of intracranial hemorrhage during ECMO, including prolonged hypoxia, ischemia, acidosis, carotid ligation, and fluctuations in cerebral blood flow superimposed on impaired cerebral autoregulation. 9 Although it is likely that all such factors play a role, their individual relative contributions are difficult to ascertain. Although usually having "normal" surgical hemostasis, healthy neonates, as compared with adults, have markedly diminished concentrations of most procoagulant proteins. For some factors the lower 95% confidence limit of normal values reaches 10% to 15% of adult values. Clinically these differences combine with other factors to produce prolongation of clotting times and, it has been suggested, a sensitiv ity to further minor reductions in factor concentrations, t2 For our study factor deficiency was defined as levels at least 2 SD less than published means for age, to ensure that this term was applied only to very severe factor reductions. It is likely, however, that even less severe reductions, particularly when present across multiple factors, will result in significant impairment of coagulation, t2, t7, 18 For example, to ensure surgical hemostasis in infants and children with hemophilia, it is generally recommended that factor VIII levels be maintained at 50% or more of normal adult levels. 19 In this investigation patients receiving ECMO were found to have coagulation factor levels substantially less than even the normally depleted levels of early infancy. With the use of a factor-containing prime solution, the initiation of ECMO was found to have variable effects on the coagulation profiles of individual patients. Qualitative examination of the data suggests that the priming protocol employed in our institution tended to benefit the patients with the most severe depletion and to lower factor levels in the patients with less depletion. Because our protocol includes administration of platelets in plasma and a circuit prime solution containing small amount of fresh-frozen plasma, it might be expected that the most severe factor deficiencies before ECMO would be ameliorated somewhat by transfusion during ECMO initiation. Conversely, because the priming solution is factor depleted as compared with normal whole blood, patients with relatively normal factor profiles might be expected to have factor reductions through hemodilution alone. During cardiopulmonary bypass for pediatric open heart surgery, circuit priming solutions devoid of plasma lead to significant hemodilution and as much as 50% reduction in coagulation protein levels.2° The ECMO priming solution used in our institution has evolved empirically and at the time of this study contained approximately 20% fresh-frozen plasma. Although the factors supplied in this solution were often insufficient to raise overall factor levels into the nor-

McManus et al.

903

mal range, it is likely that more precipitous falls in factor concentrations were prevented. A survey of hematologic support practices during ECMO is now in progress; preliminary data obtained from 58 centers in the United States indicate that fewer than one third routinely include any factorcontaining blood products in the circuit prime. Several centers now using fresh-frozen plasma in their prime have indicated intentions to replace this with simpler and less costly albuminated solutions. Our data call into question the advisability of such practice and represent a starting point for further investigation. We conclude that pediatric patients undergoing ECMO often have multiple severe, previously unrecognized coagulation factor deficiencies before cannulation and that these deficiencies may persist beyond or even be exacerbated by the initiation of bypass. We speculate that these coagulation factor deficiencies may contribute to the risk of hemorrhage during ECMO, and suggest that the widespread practice of using factor-deficient priming solutions should be reevaluated. We are indebted to Nicholas J. Morana and John J. Sears for preparation and performance of factor assays. REFERENCES

1. Watson JW, Brown DM, Lally KP, Null D, Clark R. Complications of extracorporeal membrane oxygenation in neonates. South Med J 1990;83:1262-5. 2. Sell LL, Cullen ML, Whittlesey GC, et al. Hemorrhagic complications during extracorporeal membrane oxygenation: prevention and treatment. J Pediatr Surg 1986;21:1087-91. 3. Bui KC, LaClair P, Vanderkerhove J, Bartlett RH. ECMO in premature infants: review of factors associated with mortality. ASAIO Trans 1991;37:54-9. 4. Cilley RE, Zwischenberger JB, Andrews AF, et al. Intracranial hemorrhage during extracorporeal membrane oxygenation in neonates. Pediatrics 1986;78:699-704. 5. Extracorporeat Life Support Organization. International ECMO registry report. June 1994. 6. Glass P, Miller M, Short B. Morbidity for survivors of extracorporeal membrane oxygenation: nenrodevelopmental outcome at 1 year of age. Pediatrics 1989;83:72-8. 7. Kanto WP. A decade of experience with neonatal extracorporeal membrane oxygenation. J PEDIATR t994;124:33547. 8. Clark RH, Yoder BA, Sell MS. Prospective, randomized comparison of high-frequency oscillation mad conventional ventilation in candidates for extracorporeal membrane oxygenation. J PEDIATR1994;124:447-54. 9. Short B, Walker L, Traysman R. Impaired cerebral autoregulation in the newborn lamb during recovery from severe, prolonged hypoxia combined with carotid artery and jugular vein ligation. Crit Care Med 1994;22:1262-8. 10. Andrew M, Paes B, Johnston M, et al. Development of the human coagulation system in the full-term infant. Blood 1987; 70:165-72. 11. Andrew M, Paes B, Milner R, et al. Development of the human

904

12.

13. 14.

15.

16.

McManus et aL

The Journal of Pediatrics June 1995

coagulation system in the healthy premature infant. Blood 1988;75:1651-7. Andrew M, Paes B, Johnston M. Development of the hemostatic system in the neonate and TOting infant. Am J Pediatr Hematol Oncol 1990;12:95-104. Thompson AR, Counts RB. Removal of heparin and protamine from plasma. J Lab Clin Med 1976;88:922-9. Chessells J, Wigglesworth J. Coagulation studies in preterm infants with respiratory distress and intracranial hemorrhage. Arch Dis Child 1972;47:564-70. Cole V, Durbin G, Olaffson A, et al. Pathogenesis of intraventricular haemorrhage in newborn infants. Arch Dis Child 1974;49:722-8. McDonald M, Johnson M, Rumack C, et al. Role of coagulop-

17. 18.

19.

20.

athy in newborn intracranial hemorrhage. Pediatrics 1984; 74:26-31. Seegers WH. Prothrombin complex. Semin Thromb Hemost 1981;7:291. Bick RL. Physiology of hemostasis. In: Johnson KD, ed. Disorders of thrombosis and hemostasis. Chicago: ASCP [American Society of Clinical Pathologists] Press, 1992:126. Montgomery R, Scott J. Hemostasis: diseases of the fluid phase. In: Nathan D, Osld F, ed. Hematology of infancy and childhood. 4th ed. Philadelphia: WB Saunders, 1993:1605. Kern FH, Morana NJ, Sears JJ, I-IickeyPR. Coagulation defects in neonates during cardiopulmonary bypass. Ann Thorac Surg 1992;54:541-6.

FELLOWSHIPS Fellowships available in pediatric subspecialties and those for general academic pediatric training are listed once a year, in January, in THE JOURNAL OF PEDIATRICS.Each June, forms for listing fellowships available for the academic year beginning 18 months after publication are sent to the Chairman of the Department of Pediatrics at major hospitals in the United States and Canada. In addition, a copy of the application form appears in the July, August, and September issues of THE JOURNAL (please use the current form). Should you desire to list fellowships, a separate application must be made each year for each position. All applications must be returned to Mosby-Year Book, Inc., by October 15 preceding the listing year to ensure publication. Additional forms will be supplied on request from the Joumal Editing Department, Mosby-Year Book, Inc., 11830 Westline Industrial Drive, St. Louis, MO 63146-3318/800-325-4177, ext. 4317, or 314-4534317.