Share this article on:

An Evaluation of the Effects of a Standard Heparin Dose on Thrombin Inhibition During Cardiopulmonary Bypass in Neonates

Guzzetta, Nina A. MD*; Miller, Bruce E. MD*; Todd, Kathy RN, BA, CCRC; Szlam, Fania MMSc*; Moore, Renee H. BS; Tosone, Steven R. MD*

doi: 10.1213/01.ANE.0000149590.59294.3A
Pediatric Anesthesia: Research Report

We compared the adequacy of heparinization in neonates and older children undergoing cardiopulmonary bypass (CPB) by measuring heparin activity, thrombin formation, and thrombin activity. Ten neonates and 10 older children were administered 400 U/kg of heparin before CPB. Heparin anti-Xa activity, prothrombin fragment 1.2 (F1.2), and fibrinopeptide A (FPA) were measured at baseline, after 30 min on CPB, immediately post-CPB, and 3 and 24 h post-CPB. Heparin anti-Xa activity was significantly decreased during and immediately post-CPB in the neonatal group. F1.2 and FPA levels in neonates were significantly higher at baseline, decreased with the commencement of CPB, and increased to levels higher than those in older children after CPB. Our data show that with standard heparin doses, neonates exhibit less heparin anti-Xa activity during CPB. Higher baseline levels of F1.2 and FPA present in neonates indicate preoperative activation of their coagulation systems as compared with older children. Although F1.2 and FPA levels initially decrease with the commencement of CPB, probably representing hemodilution, the subsequent increase in these markers indicates significantly more thrombin formation and activity during and after CPB. These results raise the concern that 400 U/kg of heparin may not adequately suppress thrombin formation and activity in neonates undergoing CPB.

IMPLICATIONS: Neonates experience preoperative activation of their coagulation systems. Thrombin generation and activity is poorly inhibited in neonates by 400 U/kg of heparin compared with what is seen in older children. These data argue for more investigations into appropriate anticoagulation management for neonates undergoing cardiopulmonary bypass.

*Department of Anesthesiology, Emory University School of Medicine; †Cardiac Research Department, Children’s Healthcare of Atlanta at Egleston; ‡Department of Biostatistics, Emory University, Atlanta, Georgia

Accepted for publication October 19, 2004.

Address correspondence to Nina A. Guzzetta, MD, Department of Anesthesiology, Children’s Healthcare of Atlanta at Egleston, 1405 Clifton Road NE, Atlanta, GA 30322. Address e-mail to

Contact of blood with the artificial surfaces of the cardiopulmonary bypass (CPB) circuit initiates the intrinsic system of coagulation and exposure of tissue factor initiates the extrinsic system of coagulation. Activation of these systems ultimately results in the conversion of prothrombin to thrombin. As a major regulator of hemostasis, thrombin plays several roles: it modulates the cleavage of fibrinogen to fibrin, amplifies the activation of other coagulation factors (including itself), activates platelets and the process of fibrinolysis, and stimulates the vascular endothelium to release vasoactive substances and inflammatory mediators (1,2). Indeed, some of the morbidity that follows cardiac surgery may be the consequence of mediators released because of the incomplete inhibition of thrombin during CPB (1).

Anticoagulation during CPB is necessary to prevent thrombin formation and subsequent clotting as blood comes into contact with the unphysiologic surfaces of the extracorporeal circuit. This is achieved by the administration of unfractionated heparin to inhibit not only the formation of thrombin but also the activity of circulating thrombin. This occurs by several different mechanisms. Heparin inhibits the formation of thrombin by preventing the activation of prothrombin by factor Xa via an antithrombin III (ATIII)-independent mechanism (3). Heparin’s primary anticoagulant action, however, is its ability to inhibit the activity of circulating thrombin (3). This action of heparin is dependent on its binding to ATIII and causing a conformational change in the molecular structure of ATIII, thus transforming it from a poor to a rapid inhibitor of circulating thrombin. In fact, in adults undergoing CPB, heparin effect has been shown to be dependent on coupling with ATIII (4,5), and, as ATIII concentrations decrease to less than 80 U/dL, administration of supplemental ATIII is necessary to preserve heparin’s effectiveness (6).

Because ATIII levels in infants do not reach 80 U/dL until 3 to 6 months of age (7), one could infer that neonates are potentially vulnerable to inadequate anticoagulation during CPB because of inadequate inhibition of thrombin activity. Complicating this concern is the knowledge that heparin-activated ATIII can only block circulating thrombin. Thrombin bound to fibrin or other surfaces (i.e., an injured vessel wall or a plasma protein adherent to the CPB circuit) is not available for inhibition by the ATIII-heparin complex (8). Consequently, bound thrombin remains active and is able to facilitate further thrombin generation, which, in turn, increases heparin requirements. These concerns raise important questions regarding the adequacy of heparinization in infants. The aim of this study, therefore, was to determine the effectiveness of our routine heparin dosing protocol in suppressing thrombin generation and activity in neonates during the peri-bypass period by measuring thrombin formation via prothrombin fragment 1.2 (F1.2) and thrombin activity via fibrinopeptide A (FPA).

Back to Top | Article Outline


With IRB approval and written informed consent, 20 children undergoing elective cardiac surgical procedures were enrolled in this prospective, observational study. Patients were categorized into 2 groups, a neonatal group consisting of 10 patients <1 mo of age and an older group of 10 children 10 yr of age or older. Exclusion criteria included patients presenting for emergency surgery, preoperative use of anticoagulant therapy, and patients requiring 4 h or more of CPB.

Baseline celite and kaolin activated clotting times (ACT) were documented before administering a bolus of 400 U/kg of porcine heparin. After 3 min, repeat ACT values exceeding 480 s were confirmed before initiation of CPB. Aprotinin was administered according to institutional protocol, which includes its use for repeat sternotomies and complex neonatal procedures. Nonpulsatile hypothermic CPB was performed using a nonheparin-coated system and a membrane oxygenator. Neonatal circuits contained a 350 mL priming volume with an additional 1000 U of heparin. Either a pediatric circuit with a 950 mL prime and 3000 U of heparin or an adult circuit with a 1200 mL prime and 6000 U of heparin was used for the older children depending on their size. Packed red blood cells were added as needed to achieve a hematocrit deemed acceptable for the planned surgical procedure. Additional heparin was administered as necessary to maintain an ACT >480 s during CPB and, for patients receiving aprotinin, an additional 100 U/kg of heparin were administered every hour during CPB as per institutional protocol. Termination of CPB was accomplished after normalization of body temperature and ionized calcium levels. After weaning from CPB, 4 mg/kg of protamine was used to neutralize heparin. After confirmation of heparin neutralization by ACT, persistent bleeding was treated with transfusion of platelets followed by cryoprecipitate if needed.

All blood samples were obtained from an indwelling arterial catheter after aspirating 5 mL of blood to insure that no heparin from the flush solutions was present in the collection sample. Blood was placed into the appropriate pre-chilled tubes and immediately centrifuged for 30 min. The resultant plasma was pipetted into microtubes for storage at −86°C until assayed within batches. Enzyme-linked immunosorbent assays were used to measure F1.2 (Enzygnost F1 + 2 micro; Dade Behring Inc., Deerfield, IL) and FPA (Zymutest FPA; DiaPharma Group, Inc., West Chester, OH). Chromogenic assays were used to measure ATIII values (Coamatic antithrombin; DiaPharma Group, Inc.) and heparin anti-activated factor X (heparin anti-Xa) activity (Stachrom Heparin, Diagnostica Stago, Parsippany, NJ). To determine plasma heparin levels by measurement of heparin anti-Xa activity, exogenous ATIII followed by a known excess of factor Xa is added to a plasma sample containing an undetermined amount of heparin. ATIII-heparin complexes are formed and inactivate their limit of factor Xa. The amount of factor Xa remaining in the sample is chromogenically measured and is inversely proportional to the amount of heparin in the original plasma sample. Chromogenic assay of heparin anti-Xa activity is one of the most sensitive and specific means to measure plasma heparin concentration (9).

Before surgical incision baseline values of ATIII, F1.2, FPA, and heparin anti-Xa activity were obtained. F1.2, FPA, and heparin anti-Xa activity were obtained again at the following predetermined intervals: 30 min after the initiation of CPB, immediately post-CPB but before protamine administration, 3 h post-CPB, and 24 h post-CPB.

Two-sample, two-sided Student’s t-tests were used to compare the means of the AT III levels, minutes on CPB, lowest temperature during CPB, and total heparin dose between the two groups. Comparisons of F1.2, FPA, celite ACT, and kaolin ACT by time and age group were conducted by repeated-measures analyses using a means model with the SAS mixed procedure and Bonferroni adjustments where indicated (SAS software; SAS, Cary, NC). Estimates of the standard errors were used to conduct tests and construct 95% confidence intervals. Using the SAS Genmod procedure, a repeated-measures Poisson generalized estimating equation model was fit to provide separate estimates of the means for heparin anti-Xa activity at the five time periods.

Back to Top | Article Outline


Demographic and CPB data are presented in Table 1. Neonates had significantly lower mean baseline ATIII levels (P = 0.001). Mean time on CPB was longer for the neonates versus the older children (P = 0.003) and neonates reached lower mean temperatures during CPB (P < 0.0001). Neonates also received a larger mean total heparin dose than the older children (P = 0.001). Celite and kaolin ACT values did not differ between the neonates and older children at baseline or at 3 minutes after heparin administration (Table 2).

Table 1

Table 1

Table 2

Table 2

Although heparin anti-Xa activity in the two groups followed a similar curve over the 5 time intervals, mean heparin anti-Xa activity was significantly less in neonates versus older children while on CPB (2.6 U/mL, 95% confidence interval, 2.03–3.31 versus 3.96 U/mL, 95% confidence interval, 3.31–4.73; P = 0.01) and immediately post-CPB (2.52 U/mL, 95% confidence interval, 1.99–3.18 versus 4.30 U/mL, 95% confidence interval, 3.45–5.36; P = 0.001) (Fig. 1).

Figure 1

Figure 1

F1.2 levels changed in significantly different ways between the 2 age groups over the 5 intervals (Fig. 2). At baseline, neonates had a significantly higher mean F1.2 level than the older children. With commencement of CPB, no significant difference was noted between mean F1.2 levels of the two groups. At all time intervals after CPB, mean F1.2 levels for neonates were higher than those of the older children, with statistical significance reached at 24 h post-CPB.

Figure 2

Figure 2

Within the neonatal group, there were significant differences in mean F1.2 levels for the “CPB on” measurement versus every other interval (P = 0.0002, P = 0.0003, P < 0.0001, P = 0.0004 for baseline, CPB off, 3 h post-CPB, and 24 h post-CPB, respectively, Table 3). Within the older group, the F1.2 “CPB on” measurement significantly differed from the baseline and 3-h post-CPB measurements (P < 0.0001 and P < 0.003, respectively). Other significant differences within the older group were found between baseline versus CPB off and 3-h post-CPB values (P < 0.0001), but not between baseline and 24-h post-CPB values (P = 0.02) (Table 3).

Table 3

Table 3

FPA measurements between the two age groups also changed in different ways over time. At every interval the mean FPA level in neonates was higher than that of the older children, and these differences reached statistical significance at all time intervals except “CPB on” (Fig. 2). Within the neonatal group, the “CPB off’ value was significantly different from the baseline and 3-h post-CPB values (P < 0.0001 and P = 0.0002, respectively) (Table 3). There were no significant intragroup differences for the older children (Table 3).

Back to Top | Article Outline


In this study we assessed the degree to which heparin suppresses the formation and activity of thrombin in neonates during CPB as compared with an older population having a mature hemostatic system. In summary, our data show that neonates have significantly higher baseline levels of thrombin generation (F1.2) and activity (FPA) on arrival to the operating room (OR). Administration of 400 U/kg of heparin produced adequately prolonged (>480 seconds) and equivalent ACT responses in both groups. However, neonates demonstrated significantly less heparin anti-Xa activity throughout CPB, despite receiving larger total heparin doses. Total heparin doses in neonates were larger as a consequence of heparin added to the pump priming volume and of heparin re-dosing in accordance with our aprotinin protocol. With the commencement of CPB, F1.2 and FPA levels in neonates decrease to levels seen in the older children, presumably as a result of profound hemodilution. During CPB, F1.2 levels increased in both groups, but to a greater order of magnitude in neonates. All post-CPB F1.2 values trended higher in neonates than in older children, reaching statistical significance at 24 hours. In both groups F1.2 levels returned to baseline by 24 hours post-CPB, but in neonates these 24-hour values remained significantly increased when compared with the values at the onset of CPB. FPA levels in both groups remained constant during CPB. However, post-CPB FPA levels trended higher in neonates than in the older children, showing statistical significance at all 3 post-CPB measurements.

The rationale for using heparin during CPB is based on its ability to produce anticoagulation via suppression of thrombin activity. Although heparin will not suppress the activity of all thrombin generated during CPB as a result of failure to inhibit clot-bound or surface-bound thrombin (8), there is increasing evidence that larger concentrations of heparin during CPB can more effectively reduce hemostatic activation by thrombin (8). Larger concentrations increase heparin’s potential to inactivate clot-bound and surface-bound thrombin as well as facilitate heparin’s ATIII-independent mechanisms of suppression of thrombin generation (8,10). In adults, Koster et al. (11) showed that maintenance of larger heparin concentrations during CPB led to significant reductions in thrombin generation and consequent fibrinolysis and, therefore, concluded that “heparin concentration-based anticoagulation” should perhaps become a standard in adult heparin management strategy. The decreased heparin anti-Xa activity in the neonatal group in our study did not suppress prothrombin activation and resultant thrombin generation as effectively as the increased heparin anti-Xa activity achieved in the older group despite the fact that similar initial heparin doses in each group equally affected the ACT response.

The heparin activity achieved in our neonatal group suppressed thrombin generation as measured by F1.2 levels less effectively than in the older group. In fact, more thrombin may have been generated in neonates than our data indicate because F1.2 levels are decreased by the significant hemodilution caused by the relatively large pump prime volumes used in these small children. Thus the difference in thrombin suppression between the two groups may actually be more that our study demonstrates. The heparin activity achieved in the neonatal group does appear to be sufficient to oppose thrombin activity because neonatal FPA levels do not increase during CPB. Once again, however, significant hemodilution may falsely decrease the measured values of FPA, causing us to underestimate the real degree of thrombin activity during neonatal CPB. Furthermore, thrombin does not solely mediate the cleavage of fibrinogen to fibrin. In addition to its coagulant activity, thrombin also activates platelets, neutrophils, and monocytes (1) and stimulates the process of fibrinolysis and the release of vasoactive substances and inflammatory mediators (1,2). Our investigation did not address whether these additional activities of thrombin were likewise attenuated in the neonatal group. Overall post-CPB thrombin activity as measured by FPA was significantly greater in the neonates than in the older patients. We hypothesize that the significantly larger degree of pre-CPB thrombin activity contributed to the higher FPA values seen in the post-CPB period. Additionally, we postulate that the 3-hour and 24-hour post-CPB increases in neonatal FPA levels reflect the significant amounts of coagulation products that are administered to neonates after CPB to promote clotting.

Neonates in our patient population showed significantly increased baseline levels of F1.2 and FPA as compared with the older children on entering the operating room. Other investigators have described an increased baseline level of thrombin activation in patients presenting for cardiac surgery (12,13). Nonetheless this finding is unexpected in neonates because infants have traditionally been described as having an impaired ability to generate thrombin (14). Smaller plasma concentrations of the vitamin K-dependent clotting factors, including prothrombin, provide a relative excess of ATIII to prothrombin, thus allowing less thrombin to be generated (15,16). If heparin administration for CPB was intended only to prevent de novo thrombin generation, then neonates may well be more sensitive to heparin. However, as demonstrated by our findings, neonates present to the OR with increased circulating levels of thrombin and, thus, probably increased amounts of clot-bound thrombin. Contact activation occurs preoperatively because of indwelling umbilical catheters and central lines or because of interventional manipulations in the cardiac catheterization lab. The amounts of heparin necessary to effectively inactivate these increased levels of circulating thrombin in the presence of small ATIII concentrations, not to mention to inactivate the notoriously resistant clot-bound thrombin, would be larger than expected (17). This could explain the persistently higher levels of thrombin generation and activity seen in our neonatal group after CPB.

Despite the fact that neonates received a larger total heparin dose than the older children and that ACT values in both groups were adequately prolonged in response to heparin administration, heparin anti-Xa activity measured during CPB was significantly less in the neonatal group than in the older group. This is in agreement with others who also have reported significantly smaller plasma heparin concentrations with standardized heparin dosing in pediatric patients during CPB as compared with their adult counterparts (18,19). We can anticipate that heparin levels during CPB in neonates will be less than those in adults for several reasons. More rapid metabolic rates in infants make heparin clearance by the kidney significantly faster, so that larger doses are required to reach acceptable adult ranges. Larger blood volume to body weight ratios further increase heparin requirements in neonates. Additionally, larger pump primes produce greater hemodilution of coagulation proteins, again influencing heparin’s overall effectiveness. It is therefore reasonable to assume that the optimal weight-based heparin dose required by infants undergoing CPB to reach a specific heparin concentration would indeed be larger than that required by adults.

Dietrich et al. (20) suggested that their standard heparin dose of 375 U/kg led to heparin overdosing of infants during CPB. They concluded that infants are more sensitive to heparin than adults based on the fact that neonatal ACT values were significantly more prolonged than adult ACT values in response to heparin administration before the institution of CPB. However, they did not consider thrombin suppression as an end-point nor did they measure heparin activity once CPB was established. In contrast, our results showed that no difference existed between the neonatal and older children ACT responses to the initial heparin dose, that neonates demonstrated substantially less heparin anti-Xa activity once CPB was established, and that the decreased heparin anti-Xa activity in neonates did not limit thrombin generation to the extent that it was limited in the older group. These findings suggest that neonates are less sensitive than adults to heparin and that our current heparin dosing routine for neonates may be inadequate to achieve the goal of suppressing thrombin generation and activity during CPB.

Antifibrinolytic therapy was administered to each patient according to institutional protocol. Nineteen of the 20 study patients received some type of antifibrinolytic treatment. In the neonatal group, nine patients received aprotinin and one patient received tranexamic acid. In the older group, seven patients received aprotinin, two received tranexamic acid, and one patient received no antifibrinolytic drug. Large-dose aprotinin has been shown to reduce prothrombin and fibrinogen conversion in adult patients undergoing myocardial revascularization (21) and, in patients <10 kg, large-dose aprotinin suppressed generation of F1.2 up to 4 hours post-CPB (22). Given that our dosing protocol of aprotinin is equivalent to these large-dose regimes, the presence of aprotinin in our patients should have acted to further suppress F1.2 and FPA levels. Presumably, without aprotinin F1.2 and FPA levels in both groups would have been more pronounced, further indicating that our neonatal heparin dose may be inadequate.

Increased post-CPB F1.2 and FPA levels in our neonatal group as compared with the older group could be explained by time on CPB. Newborns tended to have more complex surgeries with longer CPB times, and the amount of thrombin generation does increase with time on CPB (23). Neonates also tended to reach colder temperatures during CPB. Although thrombin generation does not differ among adult patients receiving normothermic versus hypothermic cardioplegia (23), it is unclear what the effect of hypothermic core temperatures is on thrombin generation. The significantly colder core temperatures used during neonatal CPB could in fact lead to greater thrombin expression.

In conclusion, our results support the concern that standard heparin doses used for neonatal CPB are inadequate to appropriately suppress thrombin generation and activity. Formation of thrombin in neonates during CPB is significant in the face of decreased heparin anti-Xa activity after standard weight-based heparin doses. Although conversion of fibrinogen to fibrin appears to be limited during CPB, it is unclear how the other actions of thrombin might impact neonates. Future investigations are needed to determine if the maintenance of larger heparin concentrations in neonates during CPB would further attenuate hemostatic activation and improve clinical outcomes.

Back to Top | Article Outline


1. Dietrich W. Reducing thrombin formation during cardiopulmonary bypass: Is there a benefit of the additional anticoagulant action of aprotinin? J Cardiovasc Pharmacol 1996;27:S50–7.
2. Slaughter TF, LeBleu TH, Douglas JM, et al. Characterization of prothrombin activation during cardiac surgery by hemostatic molecular markers. Anesthesiology 1994;80:520–6.
3. Hirsh J. Mechanism of action and monitoring of anticoagulants. Semin Thromb Hemostas 1986;12:1–11.
4. Hashimoto K, Yamagishi M, Sasaki T, et al. Heparin and antithrombin III levels during cardiopulmonary bypass: correlation with subclinical plasma coagulation. Ann Thorac Surg 1994;58:799–805.
5. Lemmer JH, Despotis GJ. Antithrombin III concentrate to treat heparin resistance in patients undergoing cardiac surgery. J Thorac Cardiovasc Surg 2002;123:213–7.
6. Despotis GJ, Vladimir L, Joist JH, et al. Antithrombin III during cardiac surgery: Effect on response of activated clotting time to heparin and relationship to markers of hemostatic activation. Anesth Analg 1997;85:498–506.
7. Andrew M, Paes B, Johnston M. Development of the hemostatic system in the neonate and young infant. Am J Pediatr Hematol Oncol 1990;12:95–104.
8. Weitz JI, Hudoba M, Massel D, et al. Clot-bound thrombin is protected from inhibition by heparin-antithrombin III but is susceptible to inactivation by antithrombin III-independent inhibitors. J Clin Invest 1990;86:385–91.
9. Yin ET, Wessler S, Butler JV. Plasma heparin: A unique, practical, submicrogram-sensitive assay. J Lab Clin Med 1973;81:298–310.
10. Despotis GJ, Joist HJ, Hogue CW, et al. More effective suppression of hemostatic system activation in patients undergoing cardiac surgery by heparin dosing based on heparin blood concentrations rather than ACT. Thromb Haemost 1996;76:902–8.
11. Koster A, Fischer T, Praus M, et al. Hemostatic activation and inflammatory response during cardiopulmonary bypass: Impact of heparin management. Anesthesiology 2002;97:837–41.
12. Chan A, Leaker M, Burrows F, et al. Coagulation and fibrinolytic profile of paediatric patients undergoing cardiopulmonary bypass. Thromb Haemost 1997;77:270–7.
13. Boisclair MD, Lane DA, Philippou H, et al. Thrombin production, inactivation and expression during open heart surgery measured by assays for activation fragments including a new ELISA for prothrombin fragment F1+2. Thromb Haemost 1993;70:253–8.
14. Schmidt B, Ofosu FA, Mitchell L, et al. Anticoagulant effects of heparin in neonatal plasma. Pediatr Res 1989;25:405–8.
15. Andrew M, Schmidt B, Mitchell L, et al. Thrombin generation in newborn plasma is critically dependent on the concentration of prothrombin. Thromb Haemost 1990;63:27–30.
16. Kern FH, Morana NJ, Sears JJ, et al. Coagulation defects in neonates during cardiopulmonary bypass. Ann Thorac Surg 1992;54:541–6.
17. Vieira A, Berry L, Ofosu F. Heparin sensitivity and resistance in the neonate: an explanation. Thromb Res 1991;63:85–98.
18. Andrew M. Anticoagulation and thrombolysis in children. Tex Heart Inst J 1992;19:168–77.
19. Horkay F, Martin P, Rajah SM, et al. Response to heparinization in adults and children undergoing cardiac operations. Ann Thorac Surg 1992;53:822–6.
20. Dietrich W, Braun S, Spannagl M, et al. Low preoperative antithrombin activity causes reduced response to heparin in adults but not in infant cardiac-surgical patients. Anesth Analg 2001;92:66–71.
21. Spannagl M, Dietrich W, Beck A, et al. High dose aprotinin reduces prothrombin and fibrinogen conversion in patients undergoing extracorporeal circulation for myocardial revascularization. Thromb Haemost 1994;72:159–60.
22. Mossinger H, Dietrich W, Braun SL, et al. High-dose aprotinin reduces activation of hemostasis, allogeneic blood requirement, and duration of postoperative ventilation in pediatric cardiac surgery. Ann Thorac Surg 2003;75:430–7.
23. Brister SJ, Ofosu FA, Buchanan MR. Thrombin generation during cardiac surgery: Is heparin the ideal anticoagulant? Thromb Haemost 1993;70:259–62.
© 2005 International Anesthesia Research Society