Anticoagulation for Cardiopulmonary Bypass

The safe conduct of CPB could not be accomplished without anticoagulation of blood in preparation for its contact with the extracorporeal circuit. An ideal anticoagulant should be easy to administer, rapid in onset, titratable, predictable, measurable in a timely fashion, and reversible. Heparin use during CPB has continued until the present time, most likely because of its rapid onset, ease of measurement, and ease of reversibility. Decades of the use of heparin are either a testimonial to its effectiveness or demonstrative of the inability to find a more suitable alternative.

Heparin acts as an antithrombin III (AT III) agonist and accelerates AT III binding to thrombin.^3–5 In the absence of AT III, heparin is clinically ineffective as an anticoagulant; adequate AT III activity is necessary in patients about to undergo heparinization for cardiac surgical procedures.^6–9

Activated Coagulation Time

The ACT was first introduced by Hattersley in 1966 and is still the most widely used monitor of heparin effect during cardiac surgery. Whole blood is added to a test tube containing an activator, diatomaceous earth (celite) or kaolin. The presence of activator augments the contact activation phase of coagulation, which stimulates the intrinsic coagulation pathway. ACT can be performed manually, whereby the operator measures the time interval from when blood is injected into the test tube to when clot is seen along the sides of the tube. More commonly, the ACT is automated as it is in the Hemochron and Hemotec systems. In the automated system, the test tube is placed in a device that warms the sample to 37°C. The Hemochron device (International Technidyne, Edison, NJ) rotates the test tube, which contains celite activator and a small iron cylinder, to which 2 mL whole blood is added. Before clot forms, the cylinder rolls along the bottom of the rotating test tube. When clot forms, the cylinder is pulled away from a magnetic detector, interrupts a magnetic field, and signals the end of the clotting time. Normal ACT values range from 80 to 120 seconds. The Hemochron ACT also can be performed using kaolin as the activator in a similar manner (Figure 17-1).

Figure 17-1 The Hemochron Response is a dual-chamber point-of-care coagulation monitor that is capable of measuring clotting times that are compatible with Hemochron technology. This system has software capability for calculation, data management, and storage of results.

(Courtesy of International Technidyne, Edison, NJ.)

The Hemotec ACT device (Medtronic Hemotec, Parker, CO) is a cartridge with two chambers that contain kaolin activator and is housed in a heat block. Blood (0.4 mL) is placed into each chamber, and a daisy-shaped plunger is raised and passively falls into the chamber. The formation of clot will slow the rate of descent of the plunger, and this decrease in velocity of the plunger is detected by a photo-optical system that signals the end of the ACT test. The Hemochron and Hemotec ACTs have been compared in a number of investigations and have been found to differ significantly at low heparin concentrations.¹² However, differences in heparin concentration, activator concentration, and the measurement technique make comparison of these tests difficult and have led to the realization that the Hemochron ACT result and the Hemotec ACT result are not interchangeable. In adult patients given 300 U/kg heparin for CPB, the Hemochron and Hemotec (Hepcon) ACTs were both therapeutic at all time points; however, at two points, the Hemochron ACT was statistically longer¹³ (Figure 17-2). This difference was even more pronounced in pediatric patients, who have greater heparin consumption rates (Figure 17-3). The apparent “overestimation” of ACT by the Hemochron device during hypothermic CPB may be because of the different volumes of blood that each assay warms to 37°C.

Figure 17-2 The Hemochron (circles) and Hemotec (Hepcon; squares) activated coagulation time (ACT) values in 20 adults at five time points during cardiopulmonary bypass (CPB). At 40 and 80 minutes on CPB, the Hemochron ACT was significantly greater. *P < 0.01.

(From Horkay F, Martin P, Rajah SM, Walker DR: Response to heparinization in adults and children undergoing cardiac operations. Ann Thorac Surg 53:822–826, 1992, by permission of Society of Thoracic Surgeons.)

Figure 17-3 The Hemochron (circles) and Hemotec (Hepcon; squares) activated coagulation time (ACT) values in 22 pediatric patients at six time points during cardiopulmonary bypass (CPB). Hemochron ACT was significantly greater than Hemotec ACT at five time points. *P < 0.01.

(From Horkay F, Martin P, Rajah SM, Walker DR: Response to heparinization in adults and children undergoing cardiac operations. Ann Thorac Surg 53:822–826, 1992, by permission of Society of Thoracic Surgeons.)

The ACT test can be modified by the addition of heparinase. With this modification, the coagulation status of the patient can be monitored during CPB while the anticoagulant effects of heparin are eliminated. Because this test is a side-by-side comparison of the untreated ACT to the heparinase ACT, it also has the advantage of being a rapid test for the assessment of a circulating heparin-like substance or for residual heparinization after CPB.¹⁴

With the introduction of ACT monitoring into the cardiac surgical arena, clinicians have been able to more accurately titrate heparin and protamine dosages.^11,15 As a result, many investigators report reductions in blood loss and transfusion requirements, although many of these studies used retrospective analyses.¹⁶ The improvements in postoperative hemostasis documented with ACT monitoring are potentially attributable to better intraoperative suppression of microvascular coagulation and improved monitoring of heparin reversal with protamine.¹⁷

ACT monitoring of heparinization is not without pitfalls, and its use has been criticized because of the extreme variability of the ACT and the absence of a correlation with plasma heparin levels (Figure 17-4). Many factors have been suggested to alter the ACT, and these factors are prevalent during cardiac surgical procedures. When the extracorporeal circuit prime is added to the patient’s blood volume, hemodilution occurs and may theoretically increase ACT. Evidence suggests that this degree of hemodilution alone is not enough to actually alter ACT. Hypothermia increases ACT in a “dose-related” fashion. It has been shown by Culliford et al¹⁸ that, although hemodilution and hypothermia significantly increase the ACT of a heparinized blood sample, similar increases do not occur in the absence of added heparin. The effects of platelet alterations are a bit more problematic. At mild-to-moderate degrees of thrombocytopenia, the baseline and heparinized ACT are not affected. It is not until platelet counts are reduced to less than 30,000 to 50,000/μL that ACT may be prolonged.¹⁹ Patients treated with platelet inhibitors such as prostacyclin, aspirin, or platelet-membrane-receptor antagonists have a prolonged heparinized ACT compared with patients not treated with platelet inhibitors.²⁰ This ACT prolongation is not related exclusively to decreased levels of platelet factor 4 (PF4) (PF4 is a heparin-neutralizing substance) because it also occurs when blood is anticoagulated with substances that are not neutralized by PF4. Platelet lysis, however, significantly shortens the ACT because of the release of PF4 and other platelet membrane components, which may have heparin-neutralizing activities.²¹ Gravlee et al²² showed that anesthesia and surgery decrease the ACT and create a hypercoagulable state, possibly by creating a thromboplastic response or through activation of platelets.

Figure 17-4 Anticoagulation measured at baseline (−60 minutes), heparinization (−30 minutes), and six time points after institution of cardiopulmonary bypass (CPB). Note the close correlation between the anti–factor Xa (Xa; triangles) activity and whole blood heparin concentration (WBHC; squares), which does not parallel the change in Hemochron (HC ACT; circles) or Hemotec activated coagulation time (HT ACT; diamonds).

(Modified from Despotis GJ, Summerfield AL, Joist JH: Comparison of activated coagulation time and whole blood heparin measurements with laboratory plasma anti-Xa heparin concentration in patients having cardiac operations. J Thorac Cardiovasc Surg 108:1076–1082, 1994.)

During CPB, heparin decay varies substantially and its measurement is problematic because hemodilution and hypothermia alter the metabolism of heparin. In a CPB study, Mabry et al²³ found that the consumption of heparin varied from 0.01 to 3.86 U/kg/min and there was no correlation between the initial sensitivity to heparin and the rate of heparin decay.²³ In the pediatric population, the consumption of heparin is increased to more than that of adult levels. The heparin administration protocol for pediatric patients undergoing CPB should account for a large volume of distribution, increased consumption, and a shorter elimination half-life. In monitoring the effects of heparin in pediatric patients, the minimum acceptable ACT value should be increased or an additional monitor should be used. The discrepancy between the Hemochron ACT and the Hemotec ACT that is demonstrated in Figure 17-2 is even more pronounced in pediatric patients (see Figure 17-3). Some investigators recommend the maintenance of heparin concentrations in addition to ACT during pediatric congenital heart surgery to ensure that optimal anticoagulation is being achieved.^24,25

Cascade POC (Point of Care) System

A completely different technology for measuring the effect of heparin is used by the Cascade POC analyzer (Helena, Beaumont, TX; Formerly Rapid Point Coagulation Analyzer Bayer Diagnostics, Tarrytown, NY). This test system contains disposable cards with celite activator for the measurement of heparin activity. This variation of the ACT is called the “heparin management test” (HMT). This card contains paramagnetic iron oxide particles that move in response to an oscillating magnetic field within the device. When clot formation occurs, movement of the iron oxide particles is decreased and the end of the test is signaled. This system is capable of measuring prothrombin time (PT) and activated partial thromboplastin time (aPTT), which is discussed later. The suitability of this platform for the monitoring of ACT during cardiac surgery has been demonstrated in a variety of clinical studies.^26,27 Suitability for monitoring heparinization in the interventional cardiology laboratory also has been reported. HMT correlates well with anti-Xa heparin activity in CPB patients and is less variable than standard ACT measures. In a comparison with ACT, the coefficients of variation were similar between the tests at baseline but were three times greater for the ACT during heparinization. This degree of agreement with plasma anti-Xa measurements has not been demonstrated universally when studying blood from patients undergoing CPB.²⁸

Heparin Resistance

Heparin resistance is documented by an inability to increase the ACT of blood to expected levels despite an adequate dose and plasma concentration of heparin. In many clinical situations, especially when heparin desensitization or a heparin inhibitor is suspected, heparin resistance can be treated by administering increased doses of heparin in a competitive fashion. If an adequately prolonged clotting time is ultimately achieved using greater than expected doses of heparin, a better term than heparin resistance would be heparin tachyphylaxis of “altered heparin responsiveness.” During cardiac surgical procedures, the belief that a safe minimum ACT value of 300 to 400 seconds is required for CPB is based on a few clinical studies and a relative paucity of scientific data. However, inability to attain this degree of anticoagulation in the heparin-resistant patient engenders the fear among cardiac surgical providers that the patient will experience a microvascular consumptive coagulopathy or that clots will form in the extracorporeal circuit. This potential for fibrin formation in the extracorporeal circuit was reported by Young et al,²⁹ who found increased production of fibrin monomer and consumption of fibrinogen and platelets in six of nine rhesus monkeys when the ACT declined to less than 400 seconds. However, Metz and Keats³⁰ reported no adverse effects of thrombosis or excessive bleeding in 51 patients undergoing CPB whose ACT was less than 400 seconds. In a porcine model, the group whose ACT was maintained between 250 and 300 seconds did not have excessive consumption of coagulation factors, increases in fibrin monomer formation, or changes in oxygenator performance compared with the group whose ACT was maintained at greater than 450 seconds.³¹

Many clinical conditions are associated with heparin resistance.³² Sepsis, liver disease, and pharmacologic agents represent just a few^33,34 (Table 17-1). Many investigators have documented decreased levels of AT III secondary to heparin pretreatment,³⁵ whereas others have not found decreased AT III levels.³⁶ Esposito et al³⁷ measured coagulation factor levels in patients receiving preoperative heparin infusions and found that a lower baseline ACT was the only risk factor for predicting heparin resistance compared with patients not receiving preoperative heparin.

TABLE 17-1 Disease States Associated with Heparin Resistance

ATIII, antithrombin III; DIC, disseminated intravascular coagulation.

Patients receiving preoperative heparin therapy traditionally require larger heparin doses to achieve a given level of anticoagulation when that anticoagulation is measured by the ACT. Presumably, this “heparin resistance” is due to deficiencies in the level or activity of AT III.^34,37–40 Other possible causes include enhanced factor VIII activity and platelet dysfunction causing a decrease in ACT response to heparin. Levy et al⁴⁰ have shown that the in vitro addition of AT III enhances the ACT response to heparin. Lemmer and Despotis³⁹ demonstrated that this heparin resistance, as measured by the ACT, does not correlate with preoperative AT III levels. It is unclear that these patients have increased heparin requirements during CPB because the ideal ACT and monitoring techniques have yet to be elucidated. Nicholson et al⁴¹ demonstrated that the temporal courses of ACT and AT III concentration do not parallel each other, further suggesting that AT III depletion is not the sole cause of heparin resistance during CPB. Lower ACTs (lower than 480s) were tolerated in this study with no adverse outcome. AT III concentrate is available as a heat-treated human product or in recombinant form and represents a reasonable method of treating patients with documented AT III deficiency (Box 17-1). Heparin responsiveness in the form of increased ACT levels is documented both in vitro and in vivo when heparin-resistant patients are treated with AT III.

In a multicenter, randomized, placebo-controlled trial using 75 U/kg recombinant AT III, AT III–treated patients received less FFP to augment ACT levels and also had evidence of less hemostatic activation while on CPB.⁴² There was a trend toward increased blood loss in the AT III group, which is potentially a dose effect that requires further investigation.^43,44

Heparin-Induced Thrombocytopenia

The syndrome known as heparin-induced thrombocytopenia (HIT) develops in anywhere from 5% to 28% of patients receiving heparin. HIT is commonly categorized into two subtypes. Type I is characterized by a mild decrease in platelet count and is the result of the proaggregatory effects of heparin on platelets. Type II (hereafter referred to as HIT) is considerably more severe, most often occurs after more than 5 days of heparin administration (average onset time, 9 days), and is mediated by antibody binding to the complex formed between heparin and PF4. Associated immune-mediated endothelial injury and complement activation cause platelets to adhere, aggregate, and form platelet clots, or “white clots.” Among patients developing HIT, the incidence of thrombotic complications approximates 20%, which, in turn, may carry a mortality rate as high as 40%. Demonstration of heparin-induced proaggregation of platelets confirms the diagnosis of HIT. This can be accomplished with a heparin-induced serotonin release assay or a specific heparin-induced platelet activation assay. A highly specific enzyme-linked immunosorbent assay for the heparin/PF4 complex has been developed and has been used to delineate the course of IgG and IgM antibody responses in patients exposed to unfractionated heparin during cardiac surgery. Bedside antibody tests are being developed that may speed the diagnosis of this condition.

The risks and appropriate courses of action in patients with HIT are unclear because the antibodies associated with HIT often become undetectable several weeks after discontinuing heparin. Also, the clinical syndrome does not always recur on re-exposure to heparin and sometimes resolves despite continued drug therapy. Many patients never develop thrombosis and disseminated intravascular coagulation despite positive laboratory testing. HIT possibly should be considered in the differential diagnosis of intraoperative heparin resistance in patients receiving preoperative heparin therapy.

The options for treating these patients are few. If the clinician has the luxury of being able to discontinue heparin for a few weeks, often the antibody disappears and allows a brief period of heparinization for CPB without complication.^45,46 Changing the tissue source of heparin was an option when bovine heparin was predominantly in use. Some types of low-molecular-weight heparin have been administered to patients with HIT, but reactivity of the particular low-molecular-weight heparin with the patient’s platelets should be confirmed in vitro. Supplementing heparin administration with pharmacologic platelet inhibition using prostacyclin, iloprost, aspirin, or aspirin and dipyridamole have been reported, all with favorable outcomes. The use of tirofiban with unfractionated heparin has been used in this clinical circumstance. Plasmapheresis may be used to reduce antibody levels. The use of heparin could be avoided altogether through anticoagulation with direct thrombin inhibitors such as argatroban, hirudin, or bivalirudin. These thrombin inhibitors have become standard of care in the management of the patient with HIT (Box 17-2). Bivalirudin has been studied in multicenter trials in HIT patients who must undergo CPB.⁴⁷ Monitoring the direct thrombin inhibitors during CPB is discussed later in this chapter.

Measurement of Heparin Sensitivity

Even in the absence of heparin resistance, patient response to an intravenous bolus of heparin is extremely variable.⁴⁸ The variability stems from different concentrations of various endogenous heparin-binding proteins such as vitronectin and PF4. This variability exists whether measuring heparin concentration or the ACT; however, variability seems to be greater when measuring the ACT. Because of the large interpatient variation in heparin responsiveness and the potential for heparin resistance, it is critical that a functional monitor of heparin anticoagulation (with or without a measure of heparin concentration) be used in the cardiac surgical patient. Bull et al⁴⁹ documented a threefold range of ACT response to a 200-U/kg heparin dose and similar discrepancy in heparin decay rates and, thus, recommended the use of individual patient dose–response curves to determine the optimal heparin dose. This is the concept on which point-of-care individual heparin dose–response (HDR) tests are based.

An HDR curve can be generated manually using the baseline ACT and the ACT response to an in vivo or in vitro dose of heparin. Extrapolation to the desired ACT provides the additional heparin dose required for that ACT. Once the actual ACT response to the heparin dose is plotted, further dose–response calculations are made based on the average of the target ACT and the actual ACT (Figure 17-5). This methodology was first described by Bull et al⁴⁹ and forms the scientific basis for the automated dose–response systems manufactured by Hemochron and Hemotec. The Hemochron RxDx system uses the heparin-response test, which is an ACT with a known quantity of in vitro heparin (3 IU/mL). A dose–response curve is generated that enables calculation of the heparin dose required to attain the target ACT using an algorithm that incorporates the patient’s baseline ACT, estimated blood volume, and heparin-response test. The patient’s heparin sensitivity can be calculated in seconds per international units per milliliter (sec/IU/mL) by dividing the heparin-response test by 3 IU/mL.

Figure 17-5 Construction of a dose–response curve for heparin. ACT, activated coagulation time.

(From Bull BS, Huse WM, Brauer FS, et al: Heparin therapy during extracorporeal circulation: II. The use of a dose-response curve to individualize heparin and protamine dosage. J Thorac Cardiovasc Surg 69:685–689, 1975.)

The RxDx system also provides an individualized protamine dose using the protamine-response test (PRT). This is an ACT with one of two specific quantities of protamine, depending on the amount of circulating heparin suspected (2 or 3 IU/mL). The protamine dose needed to return the ACT to baseline can be calculated on the basis of a protamine–response curve using the patient’s heparinized ACT, the PRT, and an estimate of the patient’s blood volume. Jobes et al⁵⁰ reported that the heparin dose directed by the RxDx system resulted in ACT values far greater than the target ACT. In their patients, in vivo heparin sensitivity was higher than in vitro sensitivity. RxDx also resulted in lower protamine doses, lower postoperative mediastinal tube losses, and reduced transfusion requirements compared with a ratio-based system of heparin/protamine administration. In a larger study that standardized the treatment of heparin rebound, the reduced protamine dose was confirmed; however, the reductions in bleeding were not substantiated.⁵¹ The use of a protamine dose–response curve has been shown to successfully reduce the protamine dose in vascular surgery compared with standard weight-based protamine dosing.⁵²

The Hepcon HMS system uses the HDR cartridge in the Hepcon instrument (Figure 17-6). Each cartridge houses six chambers. Chambers 1 and 2 contain heparin at a concentration of 2.5 IU/mL, chambers 3 and 4 contain heparin at a concentration of 1.5 IU/mL, and chambers 5 and 6 do not contain heparin. Once information regarding patient weight, height, and CPB prime volume is entered, the information that can be obtained from this test includes the baseline ACT (chambers 5 and 6) and an HDR slope. The dose–response slope, which is the increase in ACT from 1.5 to 2.5 IU/mL heparin, is extrapolated to the desired target ACT or target heparin concentration and the heparin dose is calculated.^32,53

Figure 17-6 The HMS heparin management system has an automated dispenser that places the appropriate volume of whole blood into each chamber of the test cartridge. A variety of assays can be performed in this instrument, depending on the cartridge used.

(Courtesy of Medtronic, Parker, CO.)

Heparin Concentration

Proponents of ACT measurement to guide anticoagulation for CPB argue that a functional assessment of the anticoagulant effect of heparin is mandatory and that the variability in ACT represents a true variability in the coagulation status of the patient. Opponents argue that during CPB, the sensitivity of the ACT to heparin is altered and ACT does not correlate with heparin concentration or with anti–factor Xa activity measurement. Heparin concentration can be measured using the Hepcon HMS system (Medtronic Hemotec), which uses an automated protamine titration technique. With a cartridge with four or six chambers containing tissue thromboplastin and a series of known protamine concentrations, 0.2 mL whole blood is automatically dispensed into the chambers. The first channel to clot is the channel whose protamine concentration most accurately neutralizes the heparin without a heparin or a protamine excess. Because protamine neutralizes heparin in the ratio of 1 mg protamine per 100 units heparin, the concentration of heparin in the blood sample can be calculated. A cartridge that monitors heparin concentration over a wide range can be used first, followed by another cartridge that can measure heparin concentrations within a more narrow range. The maintenance of a stable heparin concentration rather than a specific ACT level usually results in greater doses of heparin being administered because the hemodilution and hypothermia on CPB increase the sensitivity of the ACT to heparin. The measure of heparin concentration has been shown to correlate more closely with anti–factor Xa activity measurements than the ACT during CPB⁵⁴ (Figure 17-7), although the precision and bias of the test may not prove to be acceptable for exclusive use clinically (Figure 17-8).

Figure 17-7 There is a strong linear relation between whole-blood heparin concentration (WB heparin conc) and the anti-Xa plasma heparin concentration (Xa heparin conc). WB heparin conc was measured using the Hepcon protamine titration assay and was corrected for hematocrit value. The Xa heparin conc was measured in plasma using a substrate assay.

(From Despotis GJ, Summerfield AL, Joist JH: Comparison of activated coagulation time and whole blood heparin measurements with laboratory plasma anti-Xa heparin concentration in patients having cardiac operations. J Thorac Cardiovasc Surg 108:1076–1082, 1994.)

Figure 17-8 A Bland–Altman plot analyzing the limits of agreement between the whole-blood heparin concentration and the plasma anti-Xa heparin concentration. The bias (1.45) and two standard deviations are determined to be the limits of agreement. Note that the limits of agreement do not lie within the predetermined acceptable difference between the tests.

(From Hardy J-F, Belisle S, Robitaille D, et al: Measurement of heparin concentration in whole blood with the Hepcon/HMS device does not agree with laboratory determination of plasma heparin concentration using a chromogenic substrate for activated factor X. J Thorac Cardiovasc Surg 112:154–161, 1996.)

In a small, prospective, randomized study comparing ACT and heparin concentration monitoring, Gravlee et al⁵⁵ demonstrated increased mediastinal tube drainage after surgery in the heparin concentration group, a finding that was initially attributed to greater total heparin doses. However, heparin rebound was not systematically assessed. In a follow-up study, the authors found a greater incidence of heparin rebound in the heparin concentration group, which, when treated, resulted in no difference in bleeding between the ACT and heparin concentration groups.⁵⁶ In a prospective, randomized trial, Despotis et al⁵⁷ demonstrated that by using a transfusion algorithm in association with Hepcon-based heparin management, chest tube drainage was minimally reduced and transfusion of non–red blood cell products could be significantly reduced relative to a group of patients who had ACT-based heparin management. They attributed their results to better preservation of the coagulation system by high heparin doses because the doses of heparin administered in the Hepcon group were nearly twice the doses used in the ACT management group. However, Gravlee et al⁵⁵ were unable to confirm suppression of ongoing coagulation using Hepcon CPB management. With the exception of during cold CPB, fibrinopeptide A (FPA) levels in patients who had heparin concentration monitoring were virtually indistinguishable from those in patients who had ACT monitoring (Figure 17-9). The Hepcon, however, remains one of the more sensitive tests for detecting residual heparinization after protamine reversal because the heparin concentration can be measured by protamine titration to levels as low as 0.4 IU/mL (see Monitoring for Heparin Rebound section later in this chapter).

Figure 17-9 Fibrinopeptide A (FPA) levels during cardiopulmonary bypass (CPB) were measured in three groups of patients before anesthetic induction (Control), after heparin administration (Heparin), at lowest temperature on CPB (Cold), at esophageal temperature > 36°C (Warm), and 5 minutes after completion of the protamine dose (Protamine). Group 1, patients receiving heparin 300 IU/kg with activated coagulation time (ACT) heparin management; group 2, patients receiving heparin 250 IU/kg with ACT management; group 3, patients receiving 350 to 400 IU/kg with Hepcon heparin management. There were no differences in FPA levels among groups except during cold CPB group 3 versus group 1, P < 0.05.

(From Gravlee GP, Haddon WS, Rothberger HK, et al: Heparin dosing and monitoring for cardiopulmonary bypass. A comparison of techniques with measurement of subclinical plasma coagulation. J Thorac Cardiovasc Surg 99:518–527, 1990.)

Other tests that have been used to measure the heparin concentration include polybrene titration (functions similarly to protamine titration) and factor Xa inhibition. The latter requires plasma separation and is not practical for intraoperative monitoring. A modification of the thrombin time (TT) can be useful in monitoring heparin levels. One application would be an assay in which a known quantity of thrombin is added to patient blood or plasma. When mixed with a fibrin product, cleavage of the fibrin product can be measured fluorometrically. Only thrombin not bound by the heparin–AT III complex is available to cleave fibrin, thus yielding an indirect measure of heparin concentration.

High-Dose Thrombin Time

A functional test of heparin-induced anticoagulation that correlates well with heparin levels is the high-dose thrombin time (HiTT; International Technidyne, Edison, NJ). The TT is a clotting time that measures the conversion of fibrinogen to fibrin by thrombin. The TT is prolonged by the presence of heparin and by hypofibrinogenemias or dysfibrinogenemias. Because the TT is sensitive to very low levels of heparin, a high dose of thrombin is necessary in the TT to accurately assay the high doses of heparin used for CPB. The HiTT is performed by adding whole blood to a prewarmed, prehydrated test tube that contains a lyophilized thrombin preparation. After the addition of 1.5 mL blood, the tube is inserted into a Hemochron well and the time to clot formation is measured. In vitro assays indicate that HiTT is equivalent to the ACT in evaluation of the anticoagulant effects of heparin at heparin concentrations in the range of 0 to 4.8 IU/mL (Figure 17-10). Unlike ACT, HiTT is not altered by hemodilution and hypothermia, and has been shown to correlate better with heparin concentration than the ACT during CPB.⁵⁸ While on CPB, heparin concentration and HiTT decrease, whereas the Hemochron and the Hepcon ACT increase (Figure 17-11). Another potential advantage of HiTT monitoring is for patients receiving aprotinin therapy. In the presence of heparin, aprotinin augments the celite ACT,⁵⁹ possibly because its kallikrein-inhibiting capacity prolongs activation of the intrinsic coagulation pathway by XIIa. This should not be interpreted to represent enhanced anticoagulation. The kaolin ACT is less affected by aprotinin therapy than the celite ACT, perhaps because kaolin, unlike celite, activates the intrinsic pathway by stimulation of factor XI directly.⁶⁰ Others have suggested that kaolin binds to aprotinin and reduces the anticoagulant effect of aprotinin in vitro. However, the heparinized kaolin ACT is still somewhat prolonged in the presence of aprotinin. HiTT is not affected by aprotinin therapy and can be used as a measure of heparinization for CPB patients receiving aprotinin therapy.⁵⁹ The high-dose thromboplastin time is another measure of anticoagulation that is not affected by aprotinin therapy.⁶¹ The high-dose thromboplastin time is a WBCT in which celite is replaced by 0.3 mL of rabbit brain thromboplastin to which 1.2 mL blood is added. This test measures the time to coagulation via activation of the extrinsic pathway. This pathway of coagulation is also stimulated during pericardiotomy because of the rich thromboplastin environment of the pericardial cavity.

Figure 17-10 Hemochron ACT (HC-ACT; squares), Hepcon ACT (HR-ACT; diamonds), and high-dose thrombin time (HiTT; circles) at different in vitro heparin concentrations. The relationship is linear in each group.

(From Wang J-S, Lin C-Y, Karp RB: Comparison of high-dose thrombin time with activated clotting time for monitoring of anticoagulant effects of heparin in cardiac surgical patients. Anesth Analg 79: 9–13, 1994.)

Figure 17-11 Changes over time in the Hemochron ACT (HC-ACT; light circles), Hepcon ACT (HR-ACT; squares), and high-dose thrombin time (HiTT; upward triangles) in cardiac surgical patients. HiTT is unaffected by the changes in temperature (dark circles) and hematocrit (downward triangles) during cardiopulmonary bypass (CPB). The HC-ACT and HR-ACT increase with the initiation of CPB and the heparin concentration and HiTT decrease. *Significant difference from previous time point (P < 0.05).

(From Wang J-S, Lin C-Y, Karp RB: Comparison of high-dose thrombin time with activated clotting time for monitoring of anticoagulant effects of heparin in cardiac surgical patients. Anesth Analg 79:9–13, 1994.)

Protamine Effects on Coagulation Monitoring

Reversal of heparin-induced anticoagulation is most frequently performed with protamine. Biologically, protamine binds to positively charged groups such as phosphate groups and may have important properties in angiogenesis and immune function. Different successful dosing plans have been proposed.^62,63 The recommended dose of protamine for heparin reversal is 1 to 1.3 mg protamine per 100 units heparin; however, this dose often results in a protamine excess.

Protamine injection causes adverse hemodynamic effects.⁶⁴ Protamine reactions have been classified into three types.⁶⁵ The most common of these reactions is the type I reaction characterized by hypotension. Type II (immunologic) reactions are categorized as IIA (anaphylaxis), IIB (anaphylactoid), and IIC (noncardiogenic pulmonary edema). Type III reactions are heralded by hypotension and catastrophic pulmonary hypertension leading to right-heart failure.^33,52,66

In addition to hemodynamic sequelae, protamine has adverse effects on coagulation.⁶⁷ Large doses prolong the WBCT and the ACT, possibly via thrombin inhibition.⁶⁸ In animals and in humans, protamine has been associated with thrombocytopenia, likely because of activation of the complement cascade.⁶⁴ The anticoagulant effect of protamine also may be caused by inhibition of platelet aggregation, alteration in the platelet surface membrane, or depression of the platelet response to various agonists.^68–70 These alterations in platelet function result from the presence of the heparin–protamine complex, not protamine alone. Protamine–heparin complexes activate AT III in vitro and result in complement activation. The anticoagulant effects of free protamine occur when protamine is given in doses in excess of those used clinically; however, the risk of free protamine being the cause of a hemostatic defect is small, given the rapid clearance of protamine relative to heparin.

Monitoring for Heparin Rebound

The phenomenon referred to as heparin rebound describes the re-establishment of a heparinized state after heparin has been neutralized with protamine. Various explanations for heparin rebound have been proposed.^9,56,71,72 The most commonly postulated is that rapid distribution and clearance of protamine occur shortly after protamine administration, leaving unbound heparin remaining after protamine clearance. Furthermore, endogenous heparin antagonists have an even shorter life span than protamine and are eliminated rapidly, resulting in free heparin concentrations. Also possible is the release of heparin from tissues considered heparin storage sites (endothelium, connective tissues). Endothelial cells bind and depolymerize heparin via PF4. Uptake into the cells of the reticuloendothelial system, vascular smooth muscle, and extracellular fluid may account for the storage of heparin that contributes to reactivation of heparin anticoagulation, referred to as heparin rebound.⁹

Residual low levels of heparin can be detected by sensitive heparin concentration monitoring in the first hour after protamine reversal and can be present for up to 6 hours after surgery. Gravlee et al’s⁵⁶ study suggests that without careful monitoring for heparin rebound in the postoperative period, increased bleeding as a result of heparin rebound may occur, specifically when greater doses of heparin have been administered. Monitoring for heparin rebound can be accomplished using tests that are sensitive to low levels of circulating heparin.^52,57,73,74 These tests are also useful monitors for confirmation of heparin neutralization at the conclusion of CPB (see later).

Heparin Neutralization Monitors

To administer the appropriate dose of protamine at the conclusion of CPB, it would be ideal to measure the concentration of heparin present and give the dose of protamine necessary to neutralize only the circulating heparin. As a result of heparin metabolism and elimination, which vary considerably among individuals, the dose of protamine required to reverse a given dose of heparin decreases over time. Furthermore, protamine antagonizes the anti-IIa effects of heparin more effectively than the anti-Xa effects and, thus, varies in its potency depending on the source of heparin and its anti-IIa properties. Administration of a large fixed dose of protamine or a dose based on the total heparin dose given is no longer the standard of care and may result in an increased incidence of protamine-related adverse effects. An optimal dose of protamine is desired because unneutralized heparin results in clinical bleeding and an excess of protamine may produce an undesired coagulopathy. The use of individualized protamine dose–response curves uniformly results in a reduced protamine dose and has been shown to reduce postoperative bleeding.^49,73 One such dose–response test, the Hemochron PRT test, is an ACT performed on a heparinized blood sample that contains a known quantity of protamine. With knowledge of the ACT, PRT, and the estimated blood volume of the patient, the protamine dose needed to neutralize the existing heparin level can be extrapolated. The Hepcon instrument also has a PRT, which is the protamine titration assay. The chamber that clots first contains the dose of protamine that most closely approximates the circulating dose of heparin. The protamine dose required for its neutralization is calculated on the basis of a specified heparin/protamine dose ratio by measuring the circulating heparin level.

At the levels of heparinization needed for cardiac surgery, tests that are sensitive to heparin become unclottable. ACT is relatively insensitive to heparin and is ideal for monitoring anticoagulation at high heparin levels but is too insensitive to accurately diagnose incomplete heparin neutralization. Reiner et al⁷⁵ showed that ACT had a high predictive value for adequate anticoagulation (confirmed by laboratory aPTT) when longer than 225 seconds but was poorly predictive for inadequate anticoagulation when shorter than 225 seconds. The low levels of heparin present when heparin is incompletely neutralized are best measured by other more sensitive tests of heparin-induced anticoagulation, such as heparin concentration, aPTT, and TT. Thus, after CPB, confirmation of return to the unanticoagulated state should be performed with a sensitive test for heparin anticoagulation^76–79 (Box 17-3).

Platelet Factor 4

A component of the alpha granule of platelets, PF4 binds to and inactivates heparin. The physiologic role of PF4 is that at the site of vascular injury, PF4 is released from platelets, binds heparin (or heparin-like compounds), and promotes thrombin and clot formation. PF4 adequately neutralizes heparin inhibition of factors Xa and IIa and may be superior to protamine in neutralizing anti-Xa effects. Animal data suggest that PF4 is devoid of the adverse hemodynamic effects seen with protamine and that it may be able to be infused more rapidly.^81,82 Heparin reversal has been documented by WBCT, ACT, and heparin concentration at a PF4 concentration of 40 U/mL, approximately twice the reversal dose of protamine.⁷⁴ Levy et al⁸³ subsequently found the reversal dose of PF4 to be approximately 60 U/mL and documented similar ACT and viscoelastic measurements of clot formation compared with protamine.

Heparinase

Heparinase (Neutralase I) is an enzyme that specifically degrades heparin by catalyzing cleavage of the saccharide bonds found in the heparin molecule. As demonstrated by the ACT, heparinase in a dose of 5 mg/kg has been shown to successfully neutralize heparin effects in healthy volunteers and in patients who have undergone CPB. A dose of 7 mg/kg has been demonstrated to be even more efficacious in returning ACT to baseline values. Doses sufficient to neutralize a dose of 300 IU/kg of heparin had no significant hemodynamic effects in a canine model.⁸⁴ Investigators have not found any platelet-depressive effects of heparinase in contrast with the well-documented platelet dysfunction associated with protamine therapy.⁶⁹ Return to the unanticoagulated state after the use of heparinase has been confirmed using ACT monitoring or heparin concentration monitoring.

Bedside Tests of Coagulation

PT and aPTT tests performed on whole blood are available for use in the operating room or at the bedside. The Hemochron PT test tube contains acetone-dried rabbit brain thromboplastin to which 2 mL whole blood is added and the tube is inserted into a standard Hemochron machine. Normal values range from 50 to 72 seconds and are automatically converted by a computer to the plasma-equivalent PT and INR. Hemochron aPTT contains kaolin activator and a platelet factor substitute and is performed similarly to the PT. The aPTT is sensitive to heparin concentrations as low as 0.2 U/mL and displays a linear relation with heparin concentration up to 1.5 U/mL.

The former Thrombolytic Assessment System (TAS; Pharmanetics, Raleigh, NC), now the Cascade POC (Helena, Beaumont, TX), which was previously discussed for its ability to measure heparin via the HMT, also measures PT and aPTT. The sample is added to a cartridge containing paramagnetic iron oxide particles, which oscillate in a magnetic field as described. Specific activating reagents are used for each analyte. The analytes used include rabbit brain thromboplastin for the PT, aluminum magnesium silicate for aPTT, and celite for HMT. The blood moves by capillary action and mixes with paramagnetic iron oxide particles and reagent within the testing chamber. The decreased movement of the particles is detected optically as the sample clots, and the resultant time is displayed in seconds and as INRs for PT.

The CoaguChekProDM (Roche Diagnostics, Mannheim, Germany), former CoaguChek-plus, and formerly Ciba Corning Biotrack 512 coagulation monitor for evaluating bedside PT and aPTT, uses 0.1 mL whole blood placed into a disposable plastic cartridge for either PT or aPTT. The sample is drawn by capillary action into a heated chamber where exposure to reagents occurs. The PT uses rabbit brain thromboplastin. The aPTT uses soybean phosphatide as the platelet substitute and bovine brain sulfatide as the activator. From the reaction chamber, blood traverses a reaction path where clot formation is detected by a laser optical system. The resulting time to clot formation is converted to a ratio of the control value by a microprocessor that has control values encoded.

Many investigators have studied the former Ciba Corning Biotrack system for monitoring anticoagulation in different clinical scenarios. For patients receiving oral anticoagulant therapy, the Biotrack 512 monitor has been found to be suitable for monitoring PT and INR. Reiner et al⁷⁵ compared the bedside Biotrack aPTT with the laboratory aPTT and heparin level in patients receiving therapeutic heparinization after interventional cardiac catheterization. The authors found a strong correlation (r = 0.89) between the Biotrack aPTT and the aPTT from the hospital laboratory. The correlation between Biotrack aPTT and heparin level was not strong, probably because of the many other factors such as heparin neutralization and clearance that affect the heparin concentration in vivo.⁸⁵ Another study in patients receiving heparin compared the Ciba Corning Biotrack aPTT assay with standard laboratory aPTT and documented that Biotrack was less sensitive to heparin than the laboratory aPTT; however, the correlation coefficient of these two tests was r = 0.82.⁸⁶ In patients on warfarin therapy, the Biotrack aPTT was more sensitive than the laboratory aPTT and yielded consistently greater results for aPTT value. In another study in patients being anticoagulated for nonsurgical applications, the bedside aPTT was similar to the standard aPTT in its prediction of treatment in simple therapeutic algorithms. However, in more complex clinical situations, there was less agreement between the bedside aPTT and laboratory aPTT.⁸⁷

In a comparison of bedside coagulation monitors after cardiac surgical procedures, Reich et al⁸⁰ documented acceptable accuracy and precision levels for Hemochron and Ciba Corning Biotrack PT in comparison with standard laboratory plasma PT, making them potentially valuable for use in the perioperative period. Neither Hemochron nor Ciba Corning aPTT reached this level of clinical competence compared with standard laboratory tests. Others have documented that this monitor seems to be more precise for PT than for aPTT.⁸⁸ Because of rapid turnaround times, these point-of-care coagulation monitors may be useful in predicting patients who will bleed after cardiac surgery⁸⁹ and have also been used successfully in transfusion algorithms to decrease the number of allogeneic blood products given to cardiac surgical patients.^90,91

Measures of Fibrin Formation

The “Tenase complex” is the group of factors and cofactors that includes Xa, platelet-bound factor Va, platelet factor 3, and Ca²⁺. All adhere on the platelet surface and catalyze the cleavage of prothrombin (factor II) to thrombin (factor IIa). Thrombin then catalyzes the cleavage of fibrinogen to form fibrin monomer and fibrinopeptide A and fibrinopeptide B. These end products of fibrinogen cleavage are commonly measured serum markers that help to quantify the degree of coagulation that occurs in certain experimental or clinical situations.

One such experimental situation is the use of heparin-bonded extracorporeal circuits during CPB with the expectation that thrombin activation and fibrin formation will be minimized. Coating of the extracorporeal circuit with the heparin ligand makes the circuit more biocompatible such that the inflammatory response elicited is diminished or nonexistent. Heparin-bonded circuits have been extensively studied and considered advantageous because of their ability to reduce the inflammatory response to CPB. Human studies reveal decreases in enzymes that mark leukocyte activation, thus showing a reduction in the whole-body inflammatory response similar to that seen with leukocyte-depletion techniques.^92–94

Further enhancements in biocompatibility include less leukocyte activation and preserved platelet function. Reductions in thrombin generation have been difficult to document.⁹⁵ The increases in the fibrinogen fragment F1.2 and in D-dimer levels when heparin-coated circuits are used are similar to those seen when uncoated circuits are used.^96,97 Human studies have documented less bleeding and reduced transfusion requirements with the use of a heparin-coated extracorporeal circulation when these circuits are used in conjunction with a reduced systemic heparin dose.^98,99 In this circumstance, markers of thrombin generation are increased even greater than those in patients in whom full-dose heparin and uncoated circuits are used. In fact, increases are seen in fibrinopeptide A, prothrombin fragment F1.2, thrombin–antithrombin III complexes, D-dimers, and plasminogen activators during CPB and after protamine administration regardless of whether heparin-coated circuits are used and regardless of heparin dose.¹⁰⁰ Despite this, the use of reduced heparin doses and coated circuitry has resulted in diminished transfusion requirements and reduced chest tube drainage volumes without evidence of complications.^97,101 Because microvascular coagulation is not fully inhibited, the use of a reduced dose of heparin cannot be systematically advocated and should be implemented with caution.

Fibrinogen Level

Fibrinogen concentration is traditionally measured using either clottable protein methods, end-point detection techniques, or immunochemical tests. Of the former, the most commonly used fibrinogen assay relies on the method of Clauss. This method involves a 10-fold dilution of plasma, which ensures that fibrinogen is the rate-limiting step in clot formation. Subsequently, an excess of thrombin is added to the sample and the time to clot formation is measured. The clotting time is inversely related to the fibrinogen concentration. Because this assay relies on detection of actual clot, it can be affected by fibrin degradation products, polymerization inhibitors, or other inhibitors of fibrin formation. Because of the thrombin excess, small clinical concentrations of heparin do not affect fibrinogen determination according to the Clauss technique.

A whole-blood point-of-care fibrinogen assay is available using the Hemochron system. The specific test tube contains a lyophilized preparation of human thrombin, snake venom extract, protamine, buffers, and calcium stabilizers. The test tube is incubated with 1.5 mL distilled water and heated in the Hemochron instrument for 3 minutes. Whole blood is placed into a diluent vial, where it is 50% diluted, and from this vial, 0.5 mL diluted whole blood is placed into the specific fibrinogen test tube. The clotting time is measured using standard Hemochron technology as described previously. The fibrinogen concentration is determined by comparison with a standard curve for this test. Normal fibrinogen concentration of 180 to 220 mg/dL correlates with a clotting time of 54 ± 2.5 seconds. Fibrinogen deficiency of 50 to 75 mg/dL correlates with a clotting time of 150 ± 9.0 seconds.

Unlike the method of Clauss, the end-point detection assays rely on the detection of changes in turbidity of plasma when clot is formed. This technique does not require the maintenance of a stable cross-linked fibrin product and, therefore, does not report underestimated fibrinogen measurements because of the presence of inhibitors. Immunochemical measures of fibrinogen concentration are a direct and accurate measurement technique; however, they are expensive and time consuming and require specialized laboratory facilities.

Viscoelastic Tests

Viscoelastic tests measure the unique properties of the clot as it is forming, organizing, strengthening, and lysing. As a result, fibrinolysis determination by this methodology requires that time elapse during which clot formation is occurring. It is subsequent to clot formation and platelet-fibrin linkages that clot lysis parameters can be measured. For this reason, viscoelastic tests often require longer than 1 hour to detect the initiation of fibrinolysis; however, if fibrinolysis is enhanced, results often can be obtained in 30 minutes.

End Products of Fibrin Degradation

Other methods for quantifying fibrinolysis include measurement of the end products of fibrin degradation. Fibrin degradation products are the result of the cleavage of fibrin monomers and polymers and can be measured using a latex agglutination assay. When plasmin cleaves cross-linked fibrin, dimeric units are formed that comprise one D-domain from each of two adjacent fibrin units. These “D-dimers” are frequently measured by researchers in clinical and laboratory investigations. They are measured by either enzyme-linked immunosorbent assays or latex agglutination techniques and, thus, are not available for on-site use. Controversy still exists regarding whether D-dimer level or fibrin degradation products are the most sensitive test for detecting fibrinolysis, but most agree that the presence of D-dimers is the most specific for cross-linked fibrin degradation.¹⁰⁴

Hirudin

Hirudin, a coagulation inhibitor isolated from the salivary glands of the medicinal leech (Hirudo medicinalis), is a potent inhibitor of thrombin that, unlike heparin, acts independently of AT III and inhibits clot-bound thrombin, as well as fluid-phase thrombin. Hirudin does not require a cofactor and is not susceptible to neutralization by PF4. This would seem to be beneficial in patients in whom platelet activation and thrombosis are potential problems. Recombinant hirudin was administered as a 0.25-mg/kg bolus and an infusion to maintain the hirudin concentration at 2.5 μg/mL, as determined by the ecarin clotting time in studies by Koster et al.^109–117 The ecarin clotting time, modified for use in the Cascade POC analyzer, has been used in large series of patients with HIT.^109–114 Compared with standard treatment with heparin or low-molecular-weight heparins, recombinant hirudin–treated patients maintained platelet counts and hemoglobin levels, and had few bleeding complications, if renal function was normal.^115,116 Hirudin is a small molecule (molecular weight, 7 kDa) that is eliminated by the kidney and is easily hemofiltered at the end of CPB.¹¹⁷ In patients with abnormal renal function, bivalirudin is preferable to hirudin. An alternative treatment in this setting would be administration of unfractionated heparin with a platelet antagonist, such as tirofiban, to prevent the hyperaggregability of platelets that occurs in patients with HIT.¹¹⁶

Bivalirudin

Bivalirudin is a small 20-amino acid molecule with a plasma half-life of 24 minutes. It is a synthetic derivative of hirudin and thus acts as a direct thrombin inhibitor. Bivalirudin binds to both the catalytic binding site and the anion-binding exosite on fluid-phase and clot-bound thrombin. The part of the molecule that binds to thrombin is actually cleaved by thrombin itself, so the elimination of bivalirudin activity is independent of specific organ metabolism. Bivalirudin has been used successfully as an anticoagulant in interventional cardiology procedures as a replacement for heparin therapy. In fact, in interventional cardiology, bivalirudin has been associated with less bleeding and equivalent ischemic outcomes compared with heparin plus a platelet inhibitor.¹¹⁸ This may be the result of bivalirudin being both an antithrombin anticoagulant and an antithrombin at the level of the platelet. Merry et al¹¹⁹ showed equivalence with regard to bleeding outcomes and an improvement in graft flow after off-pump cardiac surgery when bivalirudin was used (0.75-mg/kg bolus, 1.75-mg/kg/hr infusion). Case reports confirm the safety of bivalirudin use during CPB.^120–122 Multicenter clinical trials comparing bivalirudin with heparin anticoagulation in off-pump surgery¹²³ and in CPB¹²⁴ demonstrated “noninferiority” of bivalirudin. Efficacy of anticoagulation and markers of blood loss were similar in the two groups, suggesting that bivalirudin can be a safe and effective anticoagulant in CPB. These multicenter trials used the ACT as the monitor of anticoagulant activity during surgery, but ideal monitoring is performed using the ecarin clotting time, as seen with hirudin.¹²⁵ The ecarin clotting time has a closer correlation with anti-IIa activity and plasma drug levels than does the ACT. For this reason, standard ACT monitoring during antithrombin therapy is not preferred if ecarin clotting time can be measured. A plasma-modified ACT can be used to more accurately assay the anticoagulant effects of the thrombin inhibitor drugs than ACT. This test requires the addition of exogenous plasma and, thus, is not readily available as a point-of-care assay.¹²⁶

The anticoagulant effects of the thrombin antagonists can be monitored using the ACT, aPTT, or the TT. The bleeding time also may be prolonged. In a canine CPB model, dogs receiving a synthetic thrombin inhibitor had less postoperative blood loss and greater platelet counts than those receiving heparin; however, those who received a large dose of the thrombin inhibitor still had ACT increases at 2 hours after CPB.¹²⁷ There were no differences in hemodynamics noted in the groups.

Bivalirudin has been compared favorably with heparin in patients undergoing coronary angioplasty for unstable angina.¹²⁸ The half-life of aPTT prolongation is approximately 40 minutes, and reductions in formation of fibrinopeptide A are evidence of thrombin inhibition and fibrinogen preservation. Careful monitoring should be used because there may be a rebound prothrombotic state after cessation of therapy, which could lead to recurrence of anginal symptoms (Box 17-4).

Platelet Count

Numerous events occur during cardiac surgical procedures that predispose patients to platelet-related hemostasis defects. The two major categories are thrombocytopenia and qualitative platelet defects. Thrombocytopenia commonly occurs during cardiac surgery as a result of hemodilution, sequestration, and destruction by nonendothelial surfaces. Platelet counts commonly decline to 100,000/μL or slightly less; however, the final platelet count is greatly dependent on the starting value and the duration of platelet destructive interventions (i.e., CPB).¹²⁹ Between 10,000/μL and 100,000/μL, bleeding time decreases directly; however, at platelet counts greater than 50,000/μL, neither the bleeding time nor platelet count has any correlation with postoperative bleeding in cardiac surgical patients. In contrast, platelet size or mean platelet volume does have some correlation with hemostatic function. Larger, younger platelets are more hemostatically active than smaller ones.^130,131 Mean platelet volume multiplied by the platelet count gives an estimation of overall platelet mass and is referred to as the “plateletcrit.” It is important to appreciate the inverse relation between platelet volume and platelet count when using a measure such as the plateletcrit to assess the viability of the existing platelet population. Because the mean platelet volume is dependent on the method of specimen collection, the anticoagulant used, and temperature of the storage conditions, its reproducibility is dependent on standardized laboratory procedures.

Qualitative platelet defects occur more commonly than thrombocytopenia during CPB procedures. The range of possible causes of platelet dysfunction includes traumatic extracorporeal techniques, pharmacologic therapy, hypothermia, and fibrinolysis; the hemostatic insult increases with the duration of time spent on CPB.¹³² The use of bubble oxygenators, noncoated extracorporeal circulation, and cardiotomy suctioning may cause platelets to become activated, initiate the release reaction, and partly deplete platelets of the contents of their alpha granules. Many of these changes are only transiently associated with CPB. Khuri et al¹³³ characterized the hematologic changes associated with CPB in a group of 85 patients. Whereas the platelet count declines and reaches a plateau at 2 hours after CPB, mean platelet volume reaches its nadir at 2 hours after CPB and then begins to increase during the ensuing 72 hours^131,133 (Figures 17-12 and 17-13). The relative thrombocytopenia seen up to 72 hours after cardiac surgery is not consistently associated with a bleeding diathesis. Similarly, the clotting proteins fibrinogen, factor VIII–von Willebrand factor, and factor VIII-C also increase to levels greater than baseline in the 2 to 72 hours after CPB (Figure 17-14).

Figure 17-12 Platelet count changes in patients undergoing cardiopulmonary bypass (CPB). Significant decrease in platelet count occurs on initiation of CPB and remains until at least 72 hours after surgery. *P < 0.05 change from previous value.

(From Khuri SF, Wolfe JA, Josa M, et al: Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss. J Thorac Cardiovasc Surg 104:94–107, 1992.)

Figure 17-13 Changes in mean platelet volume (MPV) in patients undergoing cardiopulmonary bypass (CPB). Note that the decrease in MPV that occurs during CPB returns to and exceeds baseline values at 24 hours after surgery. *P < 0.05 change from previous value.

(From Khuri SF, Wolfe JA, Josa M, et al: Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss. J Thorac Cardiovasc Surg 104:94–107, 1992.)

Figure 17-14 Fibrinogen values (light squares) decrease during cardiopulmonary bypass (CPB). All three clotting proteins increase to more than baseline levels in the 24 to 72 hours after CPB. *P < 0.05 change from previous value. Pre BP, prebypass; FVIIIc, factor VIIIc (dark squares); FVIII-vWF, factor VIII–von Willebrand factor (circles).

(From Khuri SF, Wolfe JA, Josa M, et al: Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss. J Thorac Cardiovasc Surg 104:94–107, 1992.)

Large doses of heparin have been shown to reduce the ability of the platelets to aggregate and to reduce clot strength.⁶⁸ This effect is not reversed when protamine is administered; however, it may be mitigated by the prophylactic administration of aprotinin.^134,135 The adverse effects of heparin on platelet function may be because of its ability to inhibit the formation of thrombin, the most potent in vivo platelet activator.¹³⁶ However, heparin also activates the fibrinolytic system, a system that, through plasmin and other activators, has the ability to depress platelet function through other mechanisms. In an extracorporeal baboon model, intravenous heparin administration resulted in increases in plasmin activity, in the quantity of immunoreactive plasmin light chain, and in immunoreactive fibrinogen fragment E.^1,137 In addition, various degrees of fibrinolysis occur after CPB. Circulating plasmin causes dissolution of the GP Ib platelet receptor and decreases the adhesiveness of platelets. Because fibrinolysis is partly responsible for the platelet dysfunction seen after heparin administration and CPB, the efficacy of antifibrinolytic agents as hemostatic drugs can be better appreciated. In addition to reducing platelet adhesiveness to von Willebrand factor, the fibrin degradation products formed depress platelet responsiveness to agonists.^138,139

Protamine–heparin complexes and protamine alone also contribute to platelet depression after CPB. Mild-to-moderate degrees of hypothermia are associated with reversible degrees of platelet activation and platelet dysfunction,¹⁴⁰ which may be partly mitigated by the use of aprotinin therapy.¹⁴¹ Overall, the potential coagulation benefits of normothermic CPB compared with hypothermic CPB require further study in well-conducted randomized trials (Box 17-5).

Bleeding Time

The bleeding time is performed by creating a skin incision and measuring the time to clot formation via the platelet plug. The Ivy bleeding time is performed on the volar surface of the forearm above which a cuff is inflated to 40 mm Hg (above venous pressure). Two parallel incisions are made using a template, and the incisions are blotted with filter paper every 30 seconds until no further bleeding occurs. The time from incision to cessation of blood seepage is the template bleeding time. The Duke bleeding time is performed on the earlobe and has advantages for cardiac surgery because the earlobe is more accessible and less likely to be subjected to the peripheral vasoconstriction seen after hypothermia. However, because neither the width/depth of the incision nor the venous pressure can be controlled in the Duke bleeding time, the Ivy bleeding time is considered the superior test. Normal bleeding time is 4 to 10 minutes.

Numerous prospective blinded investigations have confirmed that bleeding time has little or no value in predicting excessive hemorrhage after cardiac surgery.^142,143 Even in patients receiving therapeutic doses of aspirin, an increase in bleeding time does not necessarily translate to an increase in mediastinal tube drainage or transfusions if reinfusion and blood conservation techniques are used aggressively. There is substantial evidence that platelet-directed therapy in the form of platelet transfusions or desmopressin acetate shortens a prolonged bleeding time in patients with clinical hemorrhage.^144,145 In a study of 85 patients undergoing CPB, Khuri et al¹³³ demonstrated that bleeding time becomes abnormally increased during CPB and does not return to baseline even by 72 hours after surgery, whereas markers of platelet activation return to baseline by 24 hours after surgery (Figure 17-15). Because the bleeding time does not follow the temporal course of postoperative coagulopathy, the bleeding time may be a nonspecific and impractical test for detecting an existing platelet defect but may be suitable for following patient response to platelet-directed therapies.

Figure 17-15 Bleeding time (squares) increases on cardiopulmonary bypass (CPB) and remains increased until at least 72 hours after surgery. However, platelet activation is maximal during CPB. The increase in b-thromboglobulin (BTG; circles), indicating platelet activation, occurs on initiation of CPB and returns to baseline by 24 hours afterward. *P < 0.05 change from previous value.

(From Khuri SF, Wolfe JA, Josa M, et al: Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss. J Thorac Cardiovasc Surg 104:94–107, 1992.)

Aggregometry

Activated platelets undergo aggregation, which is initially a reversible process. Activation also induces the release of substances from alpha and dense platelet granules and platelet lysosomes. Because platelet granules contain many platelet agonists, the release of granular contents further stimulates platelet activation and is responsible for the secondary phase of platelet aggregation. This secondary phase of platelet aggregation is dependent on the release of thromboxane and other substances from the platelet granules, is an energy-consuming process, and is irreversible (Table 17-2).

TABLE 17-2 Platelet Adhesion and Aggregation

GP, glycoprotein; vWF, von Willebrand factor.

Aggregometry is a useful research tool for measuring platelet responsiveness to a variety of different agonists. The end result, platelet aggregation, is an objective measure of platelet activation. Platelet aggregometry uses a photo-optical instrument to measure light transmittance through a sample of whole-blood or platelet-rich plasma. Platelet-rich plasma undergoes a decrease in light transmittance on the early phase of platelet activation because of the change in platelet shape from discoid to spherical. When exposed to a platelet agonist such as thrombin, adenosine diphosphate (ADP), epinephrine, collagen, or ristocetin, the initial reversible aggregation phase results in increased light transmittance because of the platelet aggregates that decrease the turbidity of the sample. The larger the platelet aggregates, the greater is the transmittance of light. In the absence of further activation, disaggregation occurs and the plasma sample becomes turbid. However, when the platelet release reaction occurs, thromboxane and other activators are released from the platelet alpha granules and the phase of secondary, or irreversible, aggregation occurs. This results in a further increase in light transmittance.

Defects in platelet aggregation can be seen in patients with storage pool deficiency, Bernard–Soulier syndrome, or Glanzmann’s thrombasthenia and in patients taking salicylates. Impaired platelet aggregation has been demonstrated to occur after extracorporeal circulation, but investigators have had difficulty showing a correlation between impaired aggregation and clinical bleeding.^146,147 One ex vivo study demonstrated a significant correlation between platelet aggregation and 3-hour postoperative bleeding; however, correlations were greatest when preoperative aggregometry was performed using whole-blood samples and when postoperative measurements were performed using platelet-rich plasma.¹⁴⁷ The assessment of the platelet defect induced by aspirin consumption is more sensitive when whole-blood aggregometry is performed. The extreme sensitivity of this assay to minor defects in platelet function has resulted in a high negative predictive value but a rather low positive predictive value for bleeding. Its inability to be performed easily in the clinical arena has designated platelet aggregometry as strictly a research tool with occasional clinical applications.

Platelet-Mediated Force Transduction

An instrument that measures the force developed by platelets during clot retraction has been shown to be directly related to platelet concentration and function¹⁴⁸ (Hemodyne). The apparatus consists of a cup and a parallel upper plate. The cup is filled with blood or the platelet-containing solution, and the upper plate is lowered onto the clotting solution. Clot forms and adheres to the outer edges of the cup and to the plate above. A thin layer of oil is deposited onto the surfaces that are exposed to air. The upper plate is coupled to a displacement transducer that translates displacement caused by platelet retraction into a force. Normal values for platelet force development have been suggested by the investigators.⁷⁰ The antiplatelet effects of heparin have been evaluated using this force retractometer. Using this instrument, investigators have shown that high heparin concentrations completely abolish platelet force generation.⁶⁸ Furthermore, the concentration of protamine required to reverse the anticoagulant effects of heparin is not sufficient to reverse these antiplatelet effects. The antiplatelet effects of protamine alone also have been evaluated using this monitor.

Fluorescence Flow Cytometry

The introduction of fluorescence flow cytometry into the clinical laboratory has provided a sensitive and specific means for assessing causes of platelet dysfunction. Disadvantages of the in vitro assays, such as shear-induced stress and clot retraction measurements, are that they represent nonspecific markers of platelet defects. The measure of specific serum markers of platelet activation, such as b-thromboglobulin and PF4, can be performed; however, plasma collection techniques for these tests are cumbersome, and the assays are often affected by other metabolic functions. Aggregometry is only a semiquantitative process and requires a high concentration of platelets for its optimal performance.

Flow cytometry is ideal for the detection of low concentrations of specific proteins within a large population of cells. These proteins either may be static portions of the platelet surface or dynamic products of platelet activation. The platelet release reaction enables specific integrin proteins, which are a part of the platelet alpha granule membrane, to incorporate themselves into the platelet surface membrane through a mechanism analogous to exocytosis. A portion of the GP IIb/IIIa receptor is also a protein of the alpha granule membrane that becomes exposed on the surface membrane of the platelet in response to platelet activation. Flow cytometry allows for the detection and quantification of many of these surface membrane constituents as a result of immunofluorescent innovations.¹⁴⁹

Flow cytometry techniques have been enhanced by the development of specific monoclonal antibodies, which recognize antigens on the platelet (or white blood cell) surface. Antibodies developed are so specific that different ligand binding sites can be measured on the same GP IIb/IIIa molecule that characterizes different phases of receptor activation. Some of the epitopes for which monoclonal antibodies have been developed include PADGEM and GMP-140 (markers of platelet activation), the activated GP IIb/IIIa complex, and the GP Ib receptor.^150–152 A large number of monoclonal antibodies are available for identification of specific platelet ligand-binding sites (Table 17-3). Antibodies that bind specifically to activated platelets but minimally to unstimulated platelets are referred to as “activation dependent.” In utilizing activation-dependent monoclonal antibodies, flow cytometry measures the platelet reactivity or response to the addition of platelet agonists. The technique of flow cytometry can be performed using whole blood or platelet-rich plasma. The fluorescent-labeled monoclonal antibody directed against a specific platelet membrane protein is quantified by the flow cytometer, which is an instrument equipped with a laser or a light source of a specific excitation wavelength. Light scatter data are collected that help to differentiate platelets from other cellular particles. Fluorescent antibody detection is expressed as percentage of the total number of particles or as fluorescence intensity.

TABLE 17-3 Monoclonal Antibodies to Platelet Antigens

GMP, granule membrane protein; GP, glycoprotein; LAMP, lysosme membrane protein.

The ability to specifically identify platelet defects by fluorescence flow cytometry has greatly aided in the characterization of hematologic disease states such as the Bernard–Soulier syndrome and Glanzmann’s thrombasthenia.^153,154 In the cardiac surgical arena, flow cytometry has aided in diagnosing the disorders of platelet function induced by CPB and protamine administration.^{69,140,155,156} Kestin et al¹³⁶ used flow cytometric techniques to study the effects of CPB on the in vivo time-dependent upregulation of P-selectin in blood emerging from a bleeding wound. They showed that P-selectin expression is depressed after heparinization and during CPB, and recovers at approximately 2 hours after the conclusion of CPB. In contrast, in vitro activation of CPB blood with the platelet agonist phorbol myristate acetate did not reveal this depression of P-selectin expression at any time point. Using another platelet activator (thrombin-receptor agonist peptide) and flow cytometry, others did not demonstrate depression of P-selectin expression early in CPB but did so after 90 minutes of CPB and after protamine administration.¹⁵⁷

Much uncertainty still exists regarding GP Ib receptor modulation in response to CPB. Using flow cytometry, George et al¹⁴⁹ found a modest reduction in platelet surface GP Ib during CPB. However, a subsequent study by van Oeveren et al,² confirming a reduction in GP Ib, subjected the platelets to centrifugation and processing techniques that might have induced in vitro artifactual platelet activation. Kestin et al¹³⁶ used monoclonal antibodies to many epitopes expressed on GP Ib and concluded that expression of this receptor is not reduced during CPB.

As a result of the many monoclonal antibodies directed at specific epitopes on the GP IIb/IIIa receptor, investigations into the dynamics of this receptor during CPB have yielded variable results. Some studies have confirmed modest reductions in the expression of GP IIb/IIIa, although not all have used whole-blood techniques. Rinder et al,¹⁵⁵ using a whole-blood technique, demonstrated a small decrease in GP IIb/IIIa expression. Monoclonal antibodies are available that bind to the GP IIb/IIIa fibrinogen binding site, and others are available that recognize receptor-bound fibrinogen. Flow cytometric techniques also have helped to characterize the mechanisms of action of several pharmacologic agents that have shown hemostatic potential in the perioperative period.⁶⁹

Viscoelastic Tests: Thromboelastography and Sonoclot

It was the late 19th century when investigators first began to explore the possibility that viscoelastic tests of blood might yield information regarding coagulation status. The changes that occur in the viscosity of blood as it clots could be studied and measured, and this information would reflect certain aspects of coagulation function. During the early part of the 20th century, many primitive viscometers were developed that used the basic mechanisms and principles on which modern viscoelastic tests are based.

Thromboelastography

The coaguloviscometers that were developed in the 1920s formed the basis of viscoelastic coagulation testing that is now known as thromboelastography. Thromboelastography in its current form was developed by Hartert in 1948 and has been used in many different clinical scenarios to diagnose coagulation abnormalities.^158–160 Although not yet truly portable, the thromboelastograph (TEG; Haemoscope, Niles, IL [now Haemonetics, Braintree, MA]) can be performed “onsite” either in the operating room or in a laboratory and provides a rapid whole- blood analysis that yields information about clot formation and clot dissolution (Table 17-4). Within minutes, information is obtained regarding the integrity of the coagulation cascade, platelet function, platelet–fibrin interactions, and fibrinolysis. The principle is as follows: Whole blood (0.36 mL) is placed into a plastic cuvette into which a plastic pin is suspended; this plastic pin is attached to a torsion wire that is coupled to an amplifier and recorded; a thin layer of oil is added to the surface of the blood to prevent drying of the specimen; and the cuvette oscillates through an arc of 4 degrees, 45 minutes at 37°C. When the blood is liquid, movement of the cuvette does not affect the pin. However, as clot begins to form, the pin becomes coupled to the motion of the cuvette and the torsion wire generates a signal that is recorded. The recorded tracing can be stored by computer, and the parameters of interest are calculated using a simple software package. Alternatively, the tracing can be generated online with a recording speed of 2 mm/min. The tracing generated has a characteristic conformation that is the signature of the TEG (Figure 17-16).

Figure 17-16 Schematic diagram of the thromboelastograph (TEG) instrumentation (left) and a sample tracing (right).

A whole-blood sample is placed into the cup into which a plastic pin is suspended. This plastic pin is attached to a torsion wire that is coupled to an amplifier and recorder.

(From Mallett SV, Cox DJA: Thromboelastography. Br J Anaesth 69:307–313, 1992.)

The specific parameters measured by the TEG include the reaction time (R value), coagulation time (K value), “a” angle, maximal amplitude (MA), amplitude 60 minutes after the maximal amplitude (A60), and clot lysis indices at 30 and 60 minutes after MA (LY30 and LY60, respectively). The reaction time, R, represents the time for initial fibrin formation and is a measure of the intrinsic coagulation pathway, the extrinsic coagulation pathway, and the final common pathway. R is measured from the start of the bioassay until fibrin begins to form, and the amplitude of the tracing is 2 mm. Normal values vary depending on the type of activator used, range from 7 to 14 minutes using celite activator, and are as short as 1 to 3 minutes using tissue factor activator. The K value is a measure of the speed of clot formation and is measured from the end of the R time to the time that the amplitude reaches 20 mm. Normal values (3 to 6 minutes) also vary with the type of activators used. The a angle, another index of speed of clot formation, is the angle formed between the horizontal axis of the tracing and the tangent to the tracing at 20-mm amplitude. Alpha values normally range from 45 to 55 degrees. Because both the K value and the a angle are measures of the speed of clot strengthening, each is improved by high levels of functional fibrinogen. MA (normal is 50 to 60 mm) is an index of clot strength as determined by platelet function, the cross-linkage of fibrin, and the interactions of platelets with polymerizing fibrin. The peak strength of the clot, or the shear elastic modulus “G,” has a curvilinear relation with MA and is defined as G = (5000 • MA)/(96 − MA). The percentage reduction in MA after 30 minutes reflects the fibrinolytic activity present and normally is not more than 7.5%.

Characteristic TEG tracings can be recognized to be indicative of particular coagulation defects. A prolonged R value indicates a deficiency in coagulation factor activity or level and is seen typically in patients with liver disease and in patients on anticoagulants such as Coumadin or heparin. MA and a angle are reduced in states associated with platelet dysfunction or thrombocytopenia and are reduced even further in the presence of a fibrinogen defect. LY30, or the lysis index at 30 minutes after MA, is increased in conjunction with fibrinolysis. These particular signature tracings are depicted in Figure 17-17.

Figure 17-17 Signature thromboelastograph tracings.

Tracing identification from left to right: (A) normal; (B) coagulation factor deficiency; (C) platelet dysfunction or deficiency; (D) fibrinolysis; (E) hypercoagulability.

(From Mallett SV, Cox DJA: Thromboelastography. Br J Anaesth 69:307–313, 1992.)

TEG is a useful tool for diagnosing and treating perioperative coagulopathy in patients undergoing cardiac surgical procedures because of a variety of potential coagulation defects that may exist.¹⁵⁹ Within 15 to 30 minutes, on-site information is available regarding the integrity of the coagulation system, the platelet function, fibrinogen function, and fibrinolysis.^160,161 With the addition of heparinase, TEG can be performed during CPB and can provide valuable and timely information regarding coagulation status.¹⁶² Because TEG is a viscoelastic test and evaluates whole-blood hemostasis interactions, it is suggested that TEG is a more accurate predictor of postoperative hemorrhage than routine coagulation tests that analyze individual components of the hemostasis system. A number of clinical trials have confirmed that, in cardiac surgical patients, TEG has a greater predictive value and greater specificity than routine coagulation tests for diagnosing patients known as “bleeders.”¹⁶³ Tuman et al¹⁶⁰ studied 42 patients, of whom 9 were classified as bleeders. A routine coagulation screen consisting of ACT, PT, aPTT, and platelet count had only a 33% accuracy for predicting bleeding, whereas TEG and Sonoclot (Sienco, Morrison, CO; another viscoelastic test) had 88% and 74% accuracy, respectively. Mongan and Hosking¹⁶⁴ also found that TEG abnormalities predict postoperative bleeding, and using TEG parameters, they also were able to identify a population of patients who respond to therapy with desmopressin acetate. In a prospective study of 16-hour postoperative blood loss, Gravlee et al¹⁶⁵ reported on 897 cardiac surgical patients in whom routine coagulation tests were measured immediately on heparin reversal in the operating room. The weak correlations and poor predictive values of these tests confirm that these tests perform poorly as predictors of bleeding. TEG was not studied in this trial.

In a large, retrospective evaluation in more than 1000 patients, Spiess et al¹⁶⁶ found that the institution of a transfusion algorithm using TEG resulted in a significant reduction in the incidence of mediastinal exploration and in the rate of transfusion of allogeneic blood products. Because of its ease of use and application at the bedside, TEG has been used in many research settings to assess drug effects on platelet function and clot strength.^167,168 Information from the TEG also has been used to guide clinical decision making during orthotopic liver transplantation, obstetric procedures, and vascular surgical procedures.^168–171

Thromboelastography Modifications

Thromboelastography was originally performed using recalcified citrated whole blood or celite activator. The addition of recombinant human tissue factor as an activator can be used to accelerate the rate of thrombin formation and, thus, the formation of fibrin.¹⁷² This serves to shorten the time required for development of the MA. Because the MA primarily is reflective of clot strength and platelet function, this information can be obtained more quickly with tissue factor enhancement. The recombinant tissue factor is a thromboplastin agent and is available from a number of manufacturers.

An application of thromboelastography in the clinical arena is its use in monitoring GP IIb/IIIa-receptor blockade and ADP-receptor blockade in patients treated with specific antiplatelet agents. TEG with tissue factor acceleration speeds the appearance of MA and is accurate for monitoring the platelet inhibition by large concentrations of GP IIb/IIIa-receptor blockers. Using this technique with platelet-rich plasma, researchers have used the reduction of the MA as an index of platelet inhibition by GP IIb/IIIa-receptor blockers in the catheterization laboratory.¹⁷¹ Comparison with the baseline MA yields a relative measure of the degree of platelet inhibition. Thrombin-receptor agonist peptide (TRAP)–induced aggregation correlates strongly with the TEG values measured in this fashion.¹⁷¹

Because the MA is a function of the platelet–fibrinogen interaction, a reduction in the MA can be accomplished by the addition of potent GP IIb/IIIa-receptor blockade to the assay. The resultant MA, in the presence of excessively high GP IIb/IIIa-receptor blockade, primarily is due to the fibrinogen concentration and the strength of fibrin alone. This value (called Ma_f) correlates strongly with plasma fibrinogen concentration.^173,174

The thienopyridine ADP-receptor blockers, clopidogrel and ticlopidine, are widely used in cardiovascular medicine. The ability to measure the platelet defect induced by these drugs is difficult unless sophisticated laboratory techniques such as ADP-aggregometry are used. Aggregometry yields accurate results; however, it is not readily available in the perioperative period as a point-of-care test. Native TEG analysis does not measure the thienopyridine-induced platelet defect because the formation of thrombin in the assay has an overwhelming effect on the development of the TEG MA. A modification of the TEG removes thrombin from the assay and studies a nonthrombin clot, strengthened by the addition of ADP. Figure 17-18 depicts the different signature TEG tracings that are used to calculate the platelet contribution to MA when a platelet inhibitor is present. This assay was specifically created to measure the platelet inhibition by ADP antagonists such as clopidogrel and is referred to as the “platelet mapping assay.” The MA_kh is the maximal activation of platelets and fibrin, and is the largest amplitude that can be achieved. The MA_f is the maximal amplitude that is obtained when a thrombin-depleted fibrin clot is formed without a platelet contribution. The MA_pi is the MA_f contribution plus the platelet contribution. MA_pi is created by adding an activator such as ADP to the MA_f assay (for clopidogrel testing). Only platelets that can be activated by ADP contribute to the MA_pi. The following formula calculates the percentage reduction in platelet activity using this assay:

Figure 17-18 Thromboelastograph tracings using the modification to measure platelet inhibition by the thienopyridine drugs clopidogrel or ticlopidine. The measurements made are the maximal amplitude allowing thrombin activation of platelets (MA_kh; red). This is the standard kaolin-activated MA. Also measured are the maximal amplitude measuring only the fibrinogen component of MA (MA_f; blue). using a fibrinogen activator, and the maximal amplitude measuring the platelet contribution to MA by platelets able to be activated by adenosine diphosphate (ADP; MA_pi; green). This is using a fibrinogen activator plus ADP. Only platelets that are responsive to ADP will contribute to the MA_pi.

Clopidogrel, ticlopidine, and even aspirin inhibition now can be studied at the point-of-care using this modification.^175,176

Sonoclot

Another test of viscoelastic properties of blood is the Sonoclot. In 1975, von Kaulla introduced the Sonoclot (Sienco, Wheat Ridge, CO) which is a coagulation analyzer that measures the changing impedance on an ultrasonic probe that is immersed in a coagulating blood sample.¹⁷⁷ The machine consists of a plastic disposable probe mounted on an ultrasonic transducer that vibrates vertically at 200 Hz. The probe is immersed to a standard depth into 0.4 mL whole blood or plasma. The fluid exerts a force, or resistance, on the probe, which does not allow it to vibrate freely, and when the sample begins to clot, fibrin strands form on the tip of the probe and further increase the resistance. The electronic mechanism that drives the probe’s vibration also acts as a transducer, measures the impedance to vibration, and, subsequently, converts it to a signal on paper. This signature of the Sonoclot reflects coagulation in real time, from the start of fibrin formation, to fibrin cross-linkage, platelet-mediated clot strengthening, and, eventually, to clot retraction and fibrinolysis.

Immersion of the probe into the blood sample causes an initial increase in the signal because of the increased impedance to vibration by fluid relative to air. The onset time is the time for initial fibrin formation and is defined as the time taken to reach an amplitude of 1 mm. This time also has been referred to as the “SonACT” (Figure 17-19). Further increase in the signal occurs because of an increased rate of fibrinogen conversion to fibrin. The rate of rise of this first peak (R1) is expressed as the percentage of the peak amplitude per unit time (normal values, 18% to 45%). After R1, there is a shoulder or a dip before the rise in amplitude that characterizes R2. This shoulder is a result of the action of platelets and fibrin in producing clot retraction. As the clot retracts from the walls of the cuvette, the impedance to vibration briefly decreases. As fibrinogen converts to fibrin and fibrin polymerizes, the speed of clot formation and the platelet–fibrin interactions are reflected by the slope R2, the second wave. In the presence of greater concentrations of fibrinogen, a larger clot mass is represented by a greater amplitude of the R2 wave because of a greater impedance to vibration. The amplitude of the peak of R2 is therefore related to the concentration of normal functional fibrinogen. The subsequent downward slope, R3, occurs as a result of platelet-mediated clot retraction that causes plasma expulsion and clot size diminution, and thus a lower impedance. The magnitude of the R3 drop is reflective of platelet number and function. Figure 17-19 demonstrates the normal Sonoclot tracing. Figure 17-20 depicts a tracing in a patient with dysfunctional or deficient fibrinogen levels. When fibrinolysis occurs, the signal decreases even further to baseline values. The time for fibrinolysis to occur varies with each sample and usually is seen on Sonoclot analysis only if a patient has accelerated fibrinolysis.

Figure 17-19 Normal Sonoclot tracing.

(From Hett DA, Walker D, Pilkington SN, Smith DC: Sonoclot analysis. Br J Anaesth 75:771–776, 1995, by permission of BMJ Publishing Group.)

Figure 17-20 Sonoclot tracing in a patient with fibrinogen deficiency.

In the presence of lower concentrations of fibrinogen, a smaller clot mass is represented by a smaller amplitude of the R2 wave because of less impedance to vibration. The amplitude of the peak of R2 is therefore directly related to the concentration of normal functional fibrinogen.

(From Hett DA, Walker D, Pilkington SN, Smith DC: Sonoclot analysis. Br J Anaesth 75:771–776, 1995, by permission of BMJ Publishing Group.)

Sonoclot and TEG have been studied in the surgical arena because of their ability to measure viscoelastic properties of coagulation and their on-site applications. Tuman et al¹⁶⁰ found an accuracy of 74% using Sonoclot and an accuracy of 88% using TEG to predict bleeding after cardiac surgery. The reason that viscoelastic tests have been able to predict bleeding so successfully probably relates to their ability to measure platelet function, a major determinant of postoperative hemostasis.¹⁷⁸ Significant correlations have been documented between specific Sonoclot parameters and platelet count and coagulation factor assays. This reproducibility has allowed the use of Sonoclot to also predict and successfully treat coagulation abnormalities in patients undergoing liver transplantation.¹⁷⁹ Like the TEG, Sonoclot is also useful in the diagnosis of hypercoagulable states.¹⁸⁰

A direct comparison of TEG with Sonoclot is not likely to reveal a significant advantage of one test over the other because both tests measure the dynamic process of hemostasis by transducing the impedance changes that occur as blood clots. Both tests are easy to perform, can be performed on whole blood, and can be conveniently located in the operating room. Tuman et al¹⁶⁰ did not document a statistical difference between these two tests in clinical accuracy; however, they did show the viscoelastic tests to be superior to routine plasma coagulation testing. Sonoclot may provide coagulation information earlier because the initial deflection occurs when the first fibrin strands are forming. However, both TEG and Sonoclot accurately reflect the platelet–fibrin interactions required for clot formation.

ROTEM (Rotational Thrombelastometry)

The ROTEM gives a viscoelastic measurement of clot strength in whole blood (Pentapharm, Munich, Germany). A small amount of blood and coagulation activators is added to a disposable cuvette that is then placed in a heated cuvette holder. A disposable pin (sensor) that is fixed on the tip of a rotating shaft is lowered into the whole-blood sample. The loss of elasticity on clotting of the sample leads to changes in the rotation of the shaft that is detected by the reflection of light on a small mirror attached to the shaft. A detector records the axis rotation over time, and this rotation is translated into a graph or thromboelastogram.^160,161

The main descriptive parameters associated with ROTEM are the following:

ROTEM is approved for use in coagulation monitoring in Europe, and its use and familiarity are highest there. Spalding et al’s¹⁶¹ recent study has shown that implementation of ROTEM-guided coagulation management is useful in the choice of an appropriate therapeutic option in the bleeding patient. This reduces costs by avoiding administration of costly component therapy such as fresh-frozen plasma, cryoprecipitate, platelet concentrates, or antifibrinolytic agents. Its use in cardiac surgery and in transfusion algorithms is likely to be similar to that of TEG. ROTEM is currently seeking approval as a coagulation monitoring device from the Food and Drug Administration in the United States.

Tests of Platelet Response to Agonists

HemoSTATUS

Despite the introduction of numerous point-of-care coagulation analyzers that allow for rapid determination of a patient’s coagulation status, the qualitative measure of platelet function, at the bedside, remains an elusive challenge. HemoSTATUS (Medtronic, Parker, CO) is a point-of-care platelet function assay that used the Hepcon HMS monitoring system to measure platelet reactivity. A six-channel cartridge measures the heparinized kaolin-activated ACT without platelet activator (channels 1 and 2) and with incrementally increasing doses of platelet-activating factor [PAF] (channels 3 to 6). The ACT of the PAF-activated channels will be shortened because of the ability of activated platelets to speed coagulation. The respective doses of PAF in channels 3 through 6 are 1.25, 6.25, 12.5, and 150 nmol/L. For each of channels 3 through 6, the degree of shortening of the ACT as a ratio to the ACT without PAF is the “clot ratio” and is calculated as 1 − (ACT_activated/ACT_control). The “maximal” clot ratio is the clot ratio for channel 6 that was derived using blood from healthy volunteers (Figure 17-21). A comparison of the patient clot ratio to the maximal clot ratio (derived from normal volunteers) yields a comparative measure of platelet function, termed the “percentage of maximal platelet function.”

Figure 17-21 Schematic diagram of the HemoSTATUS measurement of platelet function.

A specific cartridge placed into a Hepcon machine measures the reduction in the activated clotting time as a result of the addition of increasing amounts of platelet-activating factor (PAF). ACT, activated coagulation time; ACT_n, ACT of selected channel number; ACT₁₊₂, ACT average of channels 1 and 2.

The potential ability to measure the qualitative function of platelets using a point-of-care assay provides innumerable advantages for clinicians caring for cardiac surgical patients. Platelet dysfunction is one of the more common hemostasis defects incurred during CPB, yet it is difficult to specifically measure platelet function rapidly and at the bedside. Viscoelastic tests conveniently measure platelet function, but their use in transfusion algorithms is limited by a lack of specificity to the measure of platelet dysfunction. Transfusion algorithms like the one published by Despotis et al,⁹¹ which have been suggested to result in reduced transfusions in cardiac surgical patients, have incorporated only the measure of platelet number because the on-site ability to measure platelet function has been so elusive. Inclusion of a measure of platelet function into a transfusion algorithm potentially would reduce allogeneic transfusions even further.

An initial investigation of HemoSTATUS in cardiac surgical patients was performed by Despotis et al.¹⁴⁴ The authors studied 150 patients and conducted multivariate analyses to evaluate the relation between postoperative blood loss and multiple demographic, operative, and hemostatic measurements. They demonstrated a significant correlation between HemoSTATUS measurements on arrival in the intensive care unit and 4-hour postoperative mediastinal tube drainage (r = −0.85, channel 5; r = −0.82, channel 6). The accuracy of a number of hemostasis assays was measured using receiver operating characteristic curves for the detection of excessive mediastinal tube drainage. The highest predictability for bleeding was found in both the channel 5 clot ratio and the bleeding time. The PT, aPTT, and platelet count had much lower predictive value. HemoSTATUS-derived clot ratios also had the capability to detect enhanced platelet function after the administration of pharmacologic platelet therapy (desmopressin acetate) and after the transfusion of platelet concentrates. Subsequent investigations in cardiac surgical patients have confirmed a significant yet weak correlation of HemoSTATUS with postoperative bleeding, but have not found this test to be superior to TEG or routine coagulation tests in its predictive value.¹⁶³

Validation studies have been performed to compare the assessment of platelet reactivity using HemoSTATUS with that measured by fluorescence flow cytometry. The percentage reduction in platelet function at multiple time points during cardiac surgery was compared using the two methodologies using each patient’s baseline platelet function as the control. Both tests reflected a similar degree of platelet dysfunction at the time points 90 minutes into CPB, after protamine, and at intensive care unit arrival. Differences in the type of assay and the type of platelet activators used (thrombin peptide vs. PAF) may have accounted for the inability of the tests to correlate with each other at all of the time points studied.¹⁵⁷ This POC platelet function assay is no longer supported commercially, nor available.

Multiplate Analyzer (Dynabyte Medical, Munich, Germany)

Multiplate analyzer is a test of platelet function in whole blood using impedance aggregometry.^205,206 First introduced in 2005, it is one of the most widely applied platelet aggregometers in Europe today. The analysis is performed in a single-use test cell, which incorporates a magnetic stirrer, as well as two independent impedance sensors. Activated platelets adhere and aggregate on the electrodes and, thus, enhance the electrical resistance between them. Typically, 300 μL buffered citrated blood is added to 300 μL isotonic saline and then analyzed. The device provides five channels for parallel determinations, as well as automatic analysis, calibration, and documentation using an integrated computer system. Several specific test reagents are available for stimulation of different receptors or activation of signal transduction pathways of platelets to detect changes induced by drugs, as well as by acquired or hereditary platelet disorders. After an incubation time of 3 minutes, the selected agonist solution is added and the increase in electrical impedance is recorded continuously for 6 minutes. The resistance change is transformed to arbitrary aggregation units (AUs) and plotted against time. The area under the aggregation curve is used to quantify the aggregation response and is expressed in units (1 unit corresponds with 10 AU/min). The mean values of the two independent determinations are expressed as the area under the curve of the aggregation tracing (Figure 17-22). The system has a high sensitivity for antiplatelet drugs (aspirin, clopidogrel, prasugrel, IIbIIIa-antagonists).²⁰⁷ Studies also have shown that Multiplate analysis is predictive for transfusion requirements in cardiac surgery^{205,206,208–210} and can be predictive of thromboembolism in stent patients who are nonresponsive to platelet inhibitors.²¹¹ This device is not currently available in the United States.

Figure 17-22 Multiplate tracing: The attachment of platelets onto the Multiplate sensors generates an increase in impedance that is transformed into arbitrary aggregation units (AU) and plotted against time. During a measurement period of 6 minutes, the parameters calculated by the software are the mean values of the two curves.

(From http://www.multiplate.net)