Waters, Jonathan H. MD; Anthony, David G. BS; Gottlieb, Alexandru MD; Sprung, Juraj MD, PhD*
Platelet receptor glycoprotein (GP) IIb/IIIa antagonists (abciximab, eptifibatide) are antithrombotic agents that provide comprehensive blockade of receptors necessary for the final common pathway of platelet aggregation. Perioperative bleeding, a concern whenever platelet function is inhibited, has been described in surgical patients after treatment with abciximab (1,2). To allow recovery of platelet function and to prevent bleeding, the infusion of abciximab should be discontinued 12–24 h before surgery (3). For eptifibatide, with an elimination half-life of 2.5 h, this interval is even shorter. We report a patient who preoperatively received several antiplatelet drugs (aspirin, clopidogrel, abciximab, and eptifibatide) and consequently experienced massive perioperative bleeding although surgery was performed at a time when, according to pharmacodynamic and pharmacokinetic properties of individual drugs, the antiplatelet action should have been terminated. This case illustrates a need to preoperatively assess resolution of platelet function after exposure to these antiplatelet drugs.
In addition, this case indicated that Sonoclot® (SNC; Sienco, Wheat Ridge, CO) might detect platelet dysfunction in the presence of an apparently normal Thrombelastograph® (TEG®; Haemoscope, Skokie, IL) tracing. This led us to conduct the in vitro study to examine the sensitivity of these two methods for detecting platelet dysfunction in the presence of the GP IIb/IIIa inhibitor, eptifibatide, which was the last antiplatelet drug our patient received before surgery.
A 70-yr-old, 75-kg woman was scheduled for repair of a 9-cm abdominal aortic aneurysm. She had a history of coronary artery disease, congestive heart failure, hypertension, and carotid artery stenosis. Eight days before the abdominal aortic aneurysm repair, she underwent internal carotid artery stenting and was treated with the GP IIb/IIIa inhibitor abciximab (ReoPro®; Eli Lilly, Indianapolis, IN) (0.125 μg · kg · min−1 for 12 h) followed by two antiplatelet medications, clopidogrel (Plavix®; Bristol-Myers Squibb, New York, NY) (75 mg daily for 2 days) and aspirin (325 mg daily for 2 days). After carotid artery stenting, her serum creatinine concentration increased from 0.9 mg/dL to 2.1 mg/dL. Five days before the abdominal aortic aneurysm surgery, the patient was treated with a second GP IIb/IIIa inhibitor, eptifibatide (Integrilin®; Key Pharmaceuticals, Kenilworth, NJ) (2 μg · kg · min−1 on the first day; then the rate was reduced to 1 μg · kg · min−1 in view of the increased serum creatinine concentration), which was discontinued 8 h before surgery. The preoperative platelet count was 1.25 × 105/μL, prothrombin time was 13.8 s, international normalized ratio was 1.21, and activated partial thromboplastin time was 25 s. Platelet function was not preoperatively assessed. During surgery, the patient lost 1500 mL of blood from “capillary bleeding” before aortic cross-clamping. To assess coagulation abnormality in the operating room we tested the patient’s blood with TEG® and SNC®. TEG® tracing appeared to be in the normal range (maximal amplitude [MA] was 62 mm) (Fig. 1A). At the same time SNC® signature was consistent with slow clot retraction of the fibrin gel, i.e., consistent with poor platelet function (time to peak was delayed to 23 min, normal is approximately 12 min) (Fig. 1B, dashed superimposed curve represents a normal SNC® signature obtained from the same patient 4 days later). Four platelet and two fresh-frozen plasma units were given without adequate hemostasis or significant change in SNC® signature. As the aortic cross-clamp had not been applied, the surgery was aborted. That night, the patient received multiple blood and platelet transfusions and was taken back to surgery for evaluation of the continuing bleeding (continuous decrease in hemoglobin levels). Before the repeat surgery, a coagulation profile was performed which showed no evidence of disseminated intravascular coagulation (prothrombin time 12.3 s, activated partial thromboplastin time 25.7 s, international normalized ratio 1.08, and fibrinogen 286 mg/dL). Exploratory surgery revealed a large retroperitoneal hematoma and no signs of active surgical or microcapillary bleeding. She was taken back to the intensive care unit where she remained for the next 24 h before being discharged to a regular nursing floor.
Assessment of Platelet-Inhibiting Effects of Eptifibatide with Thrombelastograph® and Sonoclot®
Because our bleeding patient had an apparently normal TEG® and an abnormal SNC® signature, we designed this in vitro study to quantify the effect of the GP IIb/IIIa receptor inhibitor, eptifibatide, with TEG® and SNC®. We believe that the eptifibatide was responsible for the bleeding in our patient because it was the last antiplatelet drug administered before surgery.
After IRB approval and written consent, 10 mL of blood was drawn from 12 volunteers. These volunteers had been asked to take no antiplatelet drugs and herbal supplements for 14 days before blood sampling. In addition, they had nothing by mouth for 12 h before the blood draw. The whole blood samples were collected into siliconized Vacutainer glass tubes containing 3.8% trisodium citrate with a sodium citrate/blood ratio maintained at 9:1 (vol/vol). The eptifibatide stock solution was prepared by diluting it with 0.9% normal saline. We added 10 μL of this stock solution to each blood sample (350 μL) in disposable TEG® or SNC® cuvettes. The final eptifibatide concentrations in the cuvettes was 4 μg/mL (0.5 × therapeutic clinical concentration), 8 μg/mL (1.0 × therapeutic plasma concentration), and 18.5 μg/mL (2.2 × therapeutic plasma concentration). A 0.9% saline solution (10 μL) was used as a control. Collected citrated whole blood was incubated for 5 min at 37°C before analyses. Before each test, the blood sample was recalcified with 0.013 mL of 0.2 M CaCl2. The analysis of the following TEG® ratio variables was performed: R (reaction time, time from sample placement in the cuvette until TEG® tracing amplitude reaches 2 mm; represents the rate of initial fibrin formation), K, clot formation time (measured from R time to the point when the amplitude of the tracing reaches 20 mm; represents the time for development of fixed degree of viscoelasticity during clot formation), α angle (α°) (angle formed by the slope of the initial TEG® tracing; denotes speed at which solid clot forms, MA (the greatest amplitude of the TEG® tracing and is a reflection of the absolute strength of the fibrin clot, i.e., reflects platelet function). The following SNC® variables were recorded: sonACT, Sonoclot activated clotting time (liquid phase or onset of clot formation, may be compared to the R interval in TEG®), clot rate (slope of the SNC® signature, characterizes fibrin gel formation and correlates with the TEG® α°, and Platelet Function (a number calculated from an algorithm that takes into account elements of clot retraction from the SNC® signature; it can be compared to the TEG® MA variable). We also measured SNC® time-to-peak (this variable represents the amount of time SNC® signature reaches the peak; it places great emphasis on how fast it takes to activate platelets rather than how much the platelets contribute to the clot retraction). Comparisons between TEG® and SNC® variables (only those that reflect platelet function) were made for each of the above three concentrations of eptifibatide. MA and SNC® platelet function values decreased while time-to-peak increased in the presence of eptifibatide. For statistical analysis we compared relative changes of each test regardless of the direction of that change. Statistical analysis was performed by repeated-measures analysis of variance, and differences were considered statistically significant at P < 0.05. All variables were expressed as mean and sd. A coefficient of variation for both methods was performed by repeatedly (n = 5) measuring SNC® Platelet Function and TEG® MA on a blood sample from a single patient.
Figure 2 shows the TEG® and SNC® dose-response tracings to eptifibatide. The eptifibatide effect on SNC® signature clot retraction is characteristic; in all studied concentrations eptifibatide only delayed clot retraction, but it never inhibited it completely. Figure 3 compares relative changes of TEG® MA, and SNC® Platelet Function and time-to-peak. At lower levels of platelet GP IIb/IIIa receptor blockade (4 μm/mL eptifibatide), the TEG® MA and SNC® Platelet Function were reduced 16 ± 9% and 28 ± 14% from baseline, respectively (P <0.03), while time-to-peak increased 45% above the baseline. At normal therapeutic range (8 μg/mL) and above the therapeutic range (18.5 μg/mL) SNC® Platelet Function and TEG® MA exhibited similar reductions (approximately 40% and 60%, respectively) (P = 0.9). At any concentration of eptifibatide studied, SNC® time-to-peak variable changed the most. Table 1 shows absolute values of TEG® and SNC® variables before and after addition of eptifibatide. There was an 18% to 20% variability in baseline values between the patients. Coefficient of variation for TEG® MA and SNC® Platelet Function was 1.6% and 2.6%, respectively (n = 5 each).
Antiplatelet drugs are administered to reduce myocardial infarction and mortality associated with unstable angina, percutaneous transluminal coronary angioplasty, and after placement of vascular stents. We described a patient who received several antiplatelet medications after carotid artery stenting but developed severe bleeding during consequent surgery for aortic aneurysm repair. It appears that the residual antiplatelet effects of the multiple medications this patient received may have contributed to the bleeding. Aspirin and clopidogrel taken 6 days before the surgery might have had lingering antiplatelet action (10 to 15 days). Eptifibatide has primarily renal excretion; therefore, its prolonged elimination might be seen in patients with renal insufficiency (3). Typically, eptifibatide effects are abated if the drug is stopped eight hours before surgery; however, this might not be true if the patient’s serum creatinine is increased, as in our patient.
The activated form of GP IIb/IIIa mediates the final common pathway of platelet aggregation. The reversibility of platelet inhibition and the rate of plasma clearance are largely a function of GP IIb/IIIa pharmacokinetic and pharmacodynamic properties (4). Clopidogrel is an antiplatelet drug that inhibits the binding of adenosine 5′-diphosphate to its platelet receptor, which leads to direct inhibition of the binding of fibrinogen to the GP IIb/IIIa receptor. The antiplatelet effect of clopidogrel lasts about 10 days, which corresponds to the life span of the platelet (5). Abciximab, a direct high-affinity GP IIb/IIIa receptor antagonist, has a biologic half-life of eight hours (6,7). Abciximab can be detected on the surface of circulating platelets for at least two weeks after discontinuation of the drug (8). Eptifibatide has a higher specificity and a lower affinity for GP IIb/IIIa receptors resulting in a biological half-life of 2.5 hours and a rapidly reversible antiplatelet effect (4). A substantial recovery of platelet aggregation is apparent within four hours of completion of eptifibatide infusion, whereas the bleeding time returns to baseline within one hour (9).
In our patient, a SNC® signature performed more than 12 hours after discontinuation of eptifibatide therapy clearly showed poor platelet function, suggesting that we should not assume the cessation of the effects of antiplatelet drugs based on their individual pharmacokinetic principles. At the same time, a TEG® tracing appeared within normal limits (Fig. 1), which suggests that TEG® and SNC® may have different sensitivities in detecting changes in the viscoelastic properties of blood treated with antiplatelet drugs.
One of the problems in accurately assessing the level of platelet dysfunction with TEG® and SNC® is the wide range of normal values (Table 1) that is caused by person-to-person variability; i.e., differences in platelet counts, fibrinogen concentrations, and variability in number of GP IIb/IIIa receptors and its ligand-binding function (10). We have demonstrated that in healthy individuals who were not on antiplatelet medications, normal values of TEG® and SNC® varied up to 20%. At the same time, the small dose of eptifibatide decreased platelet function (as represented by the change in test values) an average of 15% for TEG® and 28% for SNC®. Therefore, for both methods, there may be an overlap of values from either normal blood or blood treated with antiplatelet drugs. This may be more true for the TEG® MA that changes less than SNC® Platelet Function at low levels of platelet GP IIb/IIIa receptor blockade. In other words, TEG® MA may still be in a “normal” range despite the presence of moderate platelet inhibition, and in the absence of a baseline tracing it may be difficult to interpret the results. Unfortunately, there is a problem because we typically do not have a baseline TEG® or SNC® tracings performed before antiplatelet therapy was instituted.
This study showed that at smaller clinical concentrations of eptifibatide (4 μg/mL), SNC® was a more sensitive monitor of inhibited platelet activation than TEG®. At normal (8 μg/mL) and large (18.5 μg/mL) clinical concentrations, both SNC® and TEG® measured decreases in platelet activation with equal sensitivity. At any used eptifibatide concentration, SNC® time-to-peak was affected more than the TEG® MA was (Fig. 3). It is equally important to note that eptifibatide in the studied concentrations never completely inhibited SNC® time-to-peak; therefore, this variable was always available for qualitative visual assessment. Also, this indicates a very specific action of eptifibatide on platelet GP IIb/IIIa receptors: clot retraction is always present, albeit prolonged. This visual qualitative assessment of the SNC® signature might be useful because the sharp, well-defined peak indicates strong clot retraction (good platelet function) whereas a poorly defined peak and its delayed onset indicate weak and slow clot retraction (poor platelet function). Therefore, even in the absence of a baseline reference the configuration of SNC® signature will be visibly altered when platelet dysfunction is present, and we believe that exactly this characteristic of SNC® signature represents an advantage over any other numerical value.
In conclusion, patients treated with GP IIb/IIIa antagonists should be evaluated before surgery with platelet function tests, and adequate platelet function should not be assumed based on the drugs’ pharmacokinetic profiles. TEG® and SNC® can be used as bedside monitors for assessment of platelet activation; however, baseline values before an antiplatelet drug is given are necessary to quantify the extent of antiplatelet action. In the absence of a reference tracing (baseline), the changes in SNC® signature configuration might be more indicative of platelet dysfunction compared with the TEG® tracing. This might be especially true at the lower levels of platelet GP IIb/IIIa receptor inhibition.
1. Gammie JS, Zenati M, Kormos RL, et al. Abciximab and excessive bleeding in patients undergoing emergency cardiac operations. Ann Thorac Surg 1998; 65: 465–9.
2. Alvarez JM. Emergency coronary bypass grafting for failed percutaneous coronary artery stenting: increased costs and platelet transfusion requirement after the use of abciximab. J Thorac Cardiovasc Surg 1998; 115: 472–3.
3. GUSTO Investigators. An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. N Engl J Med 1993; 329: 673–82.
4. Kleiman NS. Pharmacokinetics and pharmacodynamics of glycoprotein IIb-IIIa inhibitors. Am Heart J 1999; 138: S263–75.
5. Sharis PJ, Cannon CP, Loscalzo J. The antiplatelet effects of ticlopidine and clopidogrel. Ann Intern Med 1998; 129: 394–405.
6. Tcheng JE, Ellis SG, George BS, et al. Pharmacodynamics of chimeric glycoprotein IIb/IIIa integrin antiplatelet antibody Fab 7E3 in high-risk coronary angioplasty. Circulation 1994; 90: 1757–64.
7. Simoons ML, de Boer MJ, van den Brand MJ, et al. Randomized trial of a glycoprotein IIb/IIIa platelet receptor blocker in refractory unstable angina. Circulation 1994; 89: 596–603.
8. Mascelli MA, Lance ET, Damaraju L, et al. Pharmacodynamic profile of short-term abciximab treatment demonstrates prolonged platelet inhibition with gradual recovery from glycoprotein IIb/IIIa receptor blockade. Circulation 1998; 97: 1680–8.
9. Harrington RA, Kleiman NS, Kottke-Marchant K, et al. Immediate and reversible platelet inhibition after intravenous administration of a peptide glycoprotein IIb/IIIa inhibitor during percutaneous coronary intervention. Am J Cardiol 1995; 76: 1222–7.
10. Jackson CW, Jennings LK. Heterogeneity of fibrinogen receptor expression on platelets activated in normal plasma with ADP: analysis with flow cytometry. Br J Haematol 1989; 72: 407–14.