Heparin-induced anticoagulation is routinely used in procedures requiring cardiopulmonary bypass (CPB) and in major vascular procedures. Although this has been the case for more than 40 yr, the characteristics of heparin onset after IV bolus injection remain incompletely defined. Consequently, in clinical practice the interval between heparin administration and interventions such as vascular cannulation and arterial cross-clamping varies. Precisely defining the minimum safe interval for these interventions would facilitate efficient clinical practice, especially in emergency situations.
The activated clotting time (ACT) is widely used to establish heparin-induced anticoagulation at the bedside. This test was introduced by Hattersley in 1966 (1) as a more practical bedside method imitating the lengthier, more cumbersome Lee-White whole blood clotting time. The work of Bull et al. (2,3) in 1975 popularized this technique for monitoring heparin-induced anticoagulation in the cardiac surgical setting. Previous studies have used the ACT to assess the onset of heparin action, but these studies have limited their observations to two or more minutes after the completion of heparin injection (4,5). Clinical recommendations for measuring the ACT after an IV heparin bolus have varied but have most often suggested an interval of 5 min or longer after dosing. This study differs from previous ones by investigating exclusively the onset characteristics of four different doses of heparin very early after its injection into the central venous circulation.
After receiving informed consent, as approved by each institution’s human research review board, we studied 40 patients undergoing coronary artery bypass grafting (CABG) or abdominal aortic aneurysmectomy (AAA). None of the patients received drugs that affect ACT measurement or plasma coagulation before surgery. All patients were anesthetized with fentanyl, a short-acting sedative-hypnotic drug (thiopental or etomidate), and succinylcholine followed by maintenance of anesthesia with fentanyl, isoflurane, midazolam, and pancuronium. All had two peripheral, large-bore IV catheters and a 7.5F indwelling, heparin-coated multilumen pulmonary artery (PA) catheter introduced via the right internal jugular vein through an 8.5F side-port introducer. Ten patients undergoing CABG at Wake Forest University (Group A) received bovine lung heparin (Upjohn, Kalamazoo, MI) 350 U/kg and had arterial and venous blood samples withdrawn simultaneously at predetermined intervals. In this group, the heparin was injected into a right ventricular port of a multilumen PA catheter. The arterial samples were taken from a 5-cm 20-gauge catheter in the left radial artery, and the venous samples were taken from the side port introducer at 30, 60, 90, 120, 180, and 600 s after completing heparin injection and a 20-mL flush of lactated Ringer’s solution. The flush required 15–20 s to complete in all groups.
At Allegheny General Hospital (AGH), 20 CABG patients had an indwelling 20-gauge 15-cm femoral artery catheter, and 10 AAA patients had an indwelling 20-gauge 5-cm radial artery catheter. The AGH CABG patients were alternately assigned to Group B (10 patients), who were given 200 U/kg of heparin, or to Group C (10 patients), who were given 300 U/kg of heparin. After surgical incision and baseline ACT determination, 10 AAA patients (Group D) were given 70 U/kg of bovine lung heparin. All AGH patients received bovine lung heparin (1000 USP U/mL) (Fujisawa USA, Deerfield, IL). After completion of the heparin injection and a lactated Ringer’s flush as described for Group A, blood samples were taken from the arterial catheter at 30, 60, 90, 120, 180, and 300 s in Groups B, C, and D.
For Group A, the arterial sampling dead space consisted of the catheter, a 15-cm length of high-pressure tubing, and the sampling port, which was a three-way stopcock. For Groups B, C, and D, the arterial sampling dead space consisted of the catheter itself, a 198-cm length of high-pressure tubing, and a three-way stopcock. For the side-port introducer in all groups, the sampling dead space consisted of the introducer lumen surrounding the indwelling PA catheter, the integral side-port transparent tubing, and the sampling three-way stopcock connected to the end of the sidearm tubing. The sampling dead space of each system was measured at <4.5 mL (<2 mL in Group A), and 10 mL was prewithdrawn from each sampling site before withdrawing the 1-mL volume used for each ACT measurement (6). The dead space measurements were made ex vivo by slowly injecting an empty example of each system with clear fluid from the sampling site and observing when fluid emerged from the intravascular end of the system.
At each sampling interval, the 1-mL blood sample was injected into celite-activated, high-range ACT cartridges (Medtronic HemoTec, Inc., Englewood, CO) for duplicate ACT determination, for which the two values were averaged. ACTs were discontinued if they reached 999 s, and any such values were recorded and analyzed as 999 s.
Data for each of the four groups were analyzed with analysis of variance. Mean values were used to identify significant differences among groups. For each patient, the time to maximum ACT was determined and compared with the time after injection of heparin in seconds. A one-tailed t-test was performed with a P < 0.05 significance level.
Table 1 provides the mean ACT values, and Table 2 lists the range of values at each measurement period for all groups. The time course for arterial and venous ACTs after heparin injection in Group A is shown in Figure 1. This plot shows significant arterial and venous ACT prolongation from baseline beginning at 30 s after injection and continuing throughout the measurement period. Venous ACTs peaked later, and arterial and venous ACTs never differed significantly from one another (Fig. 1). All but one of the venous ACTs peaked at 30 or 60 s, and all but one of the arterial ACTs peaked within 30 s of injection. Postheparin arterial ACTs differed significantly from the end-point (600 s) arterial ACT at 30 and 120 s. None of the postheparin venous ACTs differed significantly from the end-point venous ACT.
The time course for arterial ACTs after heparin injection for Groups B, C, and D is shown in Figure 2. The time course for arterial ACTs for each patient in Groups B, C, and D is shown in Figures 3, 4, and 5, respectively. Table 2 lists the range of values at each measurement period for those three groups. As in Group A, the arterial ACTs at all time periods after heparin injection differed significantly from baseline in Groups B, C, and D. In Groups B and C, ACT at 30 s after injection differed significantly from the end-point ACT (300 s), but this was not the case in Group D. Significant ACT prolongation occurred within 30 s in all 10 patients who received heparin 70 U/kg (Group B), with a minimum ACT value of 174 s at that time. If the ACT coefficient of variation is assumed to be 5% in this ACT range (7), then all but two Group B patients were effectively at peak ACT prolongation by 30 s, and the remaining two patients reached peak ACT prolongation at 60 s. In all patients receiving 200, 300, and 350 U/kg of heparin, anticoagulation suitable for CPB (ACT > 300 s) occurred within 30 s after heparin administration (P < 0.05). In Group C (200 U/kg), however, some ACTs decreased to <300 s at intervals between 60 and 300 s.
Early studies assessing the onset of heparin suggested that peak anticoagulation might occur between 5 and 25 minutes after IV heparin administration (4,8,9). In particular, the study of Effeney et al. (4) suggested that maximum ACT occurred 20–30 minutes after heparin administration in cardiac surgical patients, whereas they found an earlier peak (10 minutes) in vascular surgical patients. It seems likely that their findings were influenced by CPB-induced prolongation of the ACT (10,11). All ACT samples for Groups A, B, and C in this study were taken before CPB; therefore, it seems unlikely that any ACT prolongations resulted from factors other than the administration of the heparin bolus. By using injection conditions and heparin doses similar to those reported in the present Groups A and C, Gravlee et al. (5) measured ACTs at periods from 2 to 20 minutes after heparin administration. They found that ACT peaked at two minutes after injection, which raised the possibility that the actual peak occurred even sooner. They also found a slight redistribution effect, as judged by small but statistically significant decreases in ACT between two minutes after injection and subsequent ACTs. In this study, a similar redistribution effect was seen in Group B (200 U/kg) and Group A (350 U/kg), as evidenced by statistically significant decreases in ACT between the peak and end-point ACTs. It is unclear why this was not also observed in Groups C and D; however, the small group sizes and the variability of the ACT measurements at this level of anticoagulation (11) may preclude reliable detection of a relatively small ACT difference between the early peak and end-point values.
ACT rather imprecisely measures heparin activity. By using different instrumentation for determining ACT, Gravlee et al. (7) found striking decreases in the precision of ACT measurements after administering large boluses of heparin. Stenbjerg et al. (12) have also suggested that ACTs of more than 600 seconds correlate poorly with blood heparin concentration (12). The routine use of duplicate ACT determination in this study should reduce this imprecision somewhat, but imprecision of the test itself discourages overinterpretation of the observed ACT trends between 30 and 300 seconds after injection. ACT, however, reflects clinical practice in most centers. The activated partial thromboplastin time is more precise but becomes infinitely prolonged at the heparin concentrations used for CPB. The traditional thrombin time suffers the same fate, and the precision of the large-dose thrombin time has not been adequately validated. Assays of heparin concentration such as anti-Xa and anti-IIa levels are not readily available as a point-of-care test. The automated protamine titration technique (Medtronic) provides a discontinuous measurement and does not reliably assess clinical anticoagulation.
The venous ACT measurements were taken because we considered it clinically relevant to demonstrate that heparin-induced anticoagulation had passed through the systemic circulation and returned to the central venous blood, particularly in the context of anticipated vascular clamping. Because heparin was injected into the right ventricle in Group D and the tips of the PA catheter introducers were probably located in the proximal superior vena cava, it appears unlikely that the venous ACTs would reflect retrograde flow of the heparin bolus from the site of injection. Although Figure 1 suggests a lower peak ACT and a slightly delayed time course for central venous heparin action, simultaneously measured arterial and venous ACTs did not significantly differ at any mea-surement period. This finding is in contrast with previous work that demonstrated that venous ACT mea-surements, when obtained from an introducer sheath containing a heparin-coated pulmonary catheter, are increased compared with arterial ACTs (13). Although endothelial cells absorb heparin (14–17), our findings suggest that this effect is minor with respect to the immediate onset of heparin action in systemic arteries and veins.
We cannot exclude the possibility that arterial heparin onset would be more delayed if heparin were administered via a peripheral vein or in the presence of a low cardiac output. A smaller heparin bolus might also exhibit different onset characteristics, possibly as a result of saturable first-pass pulmonary vascular endothelial heparin uptake. This phenomenon has been demonstrated on cultured endothelium, so we speculate that it could occur in the pulmonary circulation (14,15,17). However, it seems that bolus doses smaller than 70 U/kg would be fairly uncommon, and thus this potential effect might not be clinically important.
In summary, we have shown that, when a large bolus of heparin is administered into the central venous circulation, the onset of peak ACT prolongation in systemic arteries is almost immediate (<30 seconds), and that it is nearly that rapid in systemic veins. Consequently, in the presence of circulatory stability, it appears acceptable to initiate such procedures as vascular cross-clamping and cannulation of the central circulation within 60 seconds of heparin administration. If ACT sampling is desired to confirm the adequacy of heparin-induced anticoagulation, this sampling could safely occur within 60 seconds of heparin administration into the central circulation, which could expedite the onset of CPB. We recommend waiting for the ACT result before initiating CPB except under emergency conditions.
1. Hattersley PG. Activated coagulation time of whole blood. JAMA 1966; 196: 430–6.
2. Bull BS, Korpman RA, Huse WM, Briggs BD. Heparin therapy during extracorporeal circulation. I. Problems inherent in existing heparin protocols. J Thorac Cardiovasc Surg 1975; 69: 674–84.
3. Bull BS, Huse WM, Brauer FS, Korpman RA. Heparin therapy during extracorporeal circulation. II. The use of a dose-response curve to individualize heparin and protamine dosage. J Thorac Cardiovasc Surg 1975; 69: 685–9.
4. Effeney DJ, Goldstone J, Chin D, et al. Intraoperative anticoagulation in cardiovascular surgery. Surgery 1981; 90: 1068–74.
5. Gravlee GP, Angert KC, Tucker WY, et al. Early anticoagulation peak and rapid distribution after intravenous heparin. Anesthesiology 1988; 68: 126–9.
6. Clapham MC, Willis N, Mapleson WW. Minimum volume of discard for valid blood sampling from indwelling arterial cannulae. Br J Anaesth 1987; 59: 232–5.
7. Gravlee GP, Case LD, Angert KC, et al. Variability of the activated clotting time. Anesth Analg 1988; 67: 469–72.
8. Jaques LB, Ricker AG. The relationship between heparin dosage and clotting time. Blood 1948; 3: 1197–212.
9. deTakats G. Heparin tolerance: a test of the clotting mechanism. Surg Gynecol Obstet 1943; 77: 31–9.
10. Culliford AT, Gitel SN, Starr N, et al. Lack of correlation between activated clotting time and plasma heparin during cardiopulmonary bypass. Ann Surg 1981; 193: 105–11.
11. Cohen EG, Camerlengo LJ, Dearing JP. Activated clotting times and cardiopulmonary bypass. I. The effect of hemodilution and hypothermia upon activated clotting time. J Extracorporeal Technol 1980; 12: 139–41.
12. Stenbjerg S, Berg E, Albrechtsen OK. Heparin levels and activated clotting time (ACT) during open heart surgery. Scand J Haematol 1981; 26: 282–4.
13. McNulty S, Maguire D, Thomas R. Effect of heparin-bonded pulmonary artery catheters on the activated coagulation time. J Cardiothorac Vasc Anesth 1998; 12: 533–5.
14. Barzu T, Molho P, Tobelem G, et al. Binding and endocytosis of heparin by human endothelial cells in culture. Biochim Biophys Acta 1985; 845: 196–203.
15. Hiebert LM, Jaques LB. The observation of heparin on endothelium after injection. Thromb Res 1976; 8: 195–204.
16. Mahadoo J, Hievert L, Jaques LB. Vascular sequestration of heparin. Thromb Res 1977; 12: 79–90.
© 2001 International Anesthesia Research Society
17. Glimelius B, Busch C, Hook M. Binding of heparin on the surface of cultured human endothelial cells. Thromb Res 1978; 12: 773–82.