Persistent Thrombin Generation During Heparin Therapy in Patients With Acute Coronary Syndromes
Abstract Intravenous heparin, a fundamental therapy in the treatment of patients with acute coronary syndromes, acts by inhibiting thrombin and activated factors X, IX, XI, and XII. It has also been demonstrated that heparin reduces plasma fibrinopeptide A, a marker of thrombin activity, but it is unknown whether it decreases prothrombin fragment 1+2, an indirect marker of thrombin generation. We measured the plasma levels of prothrombin fragment 1+2, fibrinopeptide A, and antithrombin III in 64 consecutive patients with unstable angina or myocardial infarction receiving intravenous heparin. Blood samples were obtained at baseline (before any treatment) and then at 90 minutes and 24 and 48 hours after the administration of an intravenous bolus of heparin (5000 IU) followed by a continuous infusion of 1000 IU per hour to maintain activated partial thromboplastin time at more than double its baseline levels. In comparison with baseline, there was a significant decrease in fibrinopeptide A at 90 minutes and at 24 and 48 hours (baseline, 2.3 nmol/L; 90 minutes, 1.15 nmol/L; 24 hours, 1.4 nmol/L; 48 hours, 1.2 nmol/L; P<.0001) but no change in prothrombin fragment 1+2 levels (baseline, 1.27 nmol/L; 90 minutes, 1.3 nmol/L; 24 hours, 1.33 nmol/L; 48 hours, 1.29 nmol/L; P=NS). Antithrombin III activity decreased at 24 and 48 hours (baseline, 108%; 24 hours, 97%; 48 hours, 95%; P<.0001). Hence, in patients with acute coronary syndromes, intravenous heparin at a dose reaching an activated partial thromboplastin time that adequately suppresses thrombin activity does not suppress increased thrombin generation.
- Received August 29, 1996.
- Accepted November 13, 1996.
Intravenous heparin has become a recommended therapy in the treatment of patients with acute coronary syndromes.1 2 This is supported by the finding that heparin leads to an improved clinical outcome in patients with unstable angina 3 4 and myocardial infarction.5 The anticoagulant action of heparin is primarily due to its ability to bind tightly to antithrombin III, thereby accelerating the rate of inhibition of the major coagulation enzymes, particularly thrombin and, to a lesser extent, activated factors X, IX, XI, and XII.6 7 Biochemical studies in patients with acute coronary syndromes have clearly demonstrated that intravenous heparin rapidly inhibits thrombin action on fibrinogen and lowers plasma fibrinopeptide A levels to within the normal range.8 9 10 However, the effect of heparin on the earlier steps of the coagulation cascade have never been studied in vivo in patients with acute coronary syndromes. Because heparin is active in inhibiting factors Xa and IXa, theoretically it should reduce not only thrombin activity but also thrombin generation. We therefore investigated the effect of intravenous heparin on both thrombin generation (as indirectly assessed by plasma prothrombin fragment 1+2) and thrombin activity (as assessed by plasma fibrinopeptide A) in patients with unstable angina and myocardial infarction.
The study population consisted of patients with acute unstable angina or acute myocardial infarction consecutively admitted to the Division of Cardiology, Ca’ Granda Niguarda Hospital, Milan, Italy, who had no contraindications to heparin therapy.
Inclusion Criteria and Patient Subgroups
Acute unstable angina was defined as chest pain of recent onset that had occurred at rest within the previous 48 hours and was accompanied by transient ECG ischemic changes (ST-segment elevation or depression ≥1 mm 0.08 second after the J-point, new negative T waves, or the pseudonormalization of previously negative T waves), with serum creatine kinase fraction levels of less than twice the upper limit of normal.
Acute myocardial infarction was defined as chest pain of recent onset (<12 hours) lasting ≥30 minutes, which failed to respond to sublingual or intravenous nitrates and was accompanied by ST-segment elevation or depression of ≥0.1 mV in at least two contiguous ECG leads that evolved into pathological Q-wave or ST-T segment changes, and the development of elevated creatine kinase and MB fraction levels of at least twice the upper normal limit.
Patients with comorbid conditions known to alter coagulation system activity or decrease the clearance of activation fragments as well as those who were taking oral anticoagulant therapy were deemed ineligible for the study. Of the eligible patients, those with the following conditions were excluded: concomitant peripheral vascular disorders or valvular heart disease (20 patients), severe heart failure (10 patients), start of heparin therapy before baseline blood sampling (24 patients), thrombolytic therapy (80 patients), or severely limited venous access (16 patients).
After their inclusion in the study, a baseline blood sample was obtained from the enrolled patients before treatment was started. Subsequently, an intravenous bolus of heparin 5000 IU was given, immediately followed by a continuous intravenous infusion of heparin 1000 IU per hour for ≥72 hours, which was adjusted to maintain the activated partial thromboplastin time (aPTT) at more than double its baseline value. The associated treatments were a combination of intravenous nitroglycerin (0.5 to 1 μg/kg per minute) and/or diltiazem (1 to 6 μg/kg per minute) or oral β-blockers (atenolol 50 to 100 mg/d). All of the patients received aspirin (165 to 325 mg ) at the start of heparin therapy. Additional blood samples for coagulation activation markers were obtained at 90 minutes and 24 and 48 hours. Continuous ECG Holter monitoring for 72 hours was started immediately after inclusion in the study, at the same time as the start of heparin therapy, and the patients were followed up for the occurrence of adverse outcome events during the 72-hour period. Twelve-lead ECGs were recorded whenever chest pain occurred. Creatine kinase levels were measured every morning and every 4 hours after any episode of chest pain.
Blood Sampling and Handling
Clean venipunctures were performed by two specially trained investigators using 19-gauge butterfly infusion sets and a two-syringe technique. Inadequate blood samples were prospectively excluded. After the first 4 mL of blood was discarded, the samples were placed directly into refrigerated vacutainers for fibrinopeptide A, prothrombin fragment 1+2, antithrombin III, and aPTT determinations, always in the same sequence. The samples for the fibrinopeptide A and prothrombin fragment 1+2 assays were collected into refrigerated vacutainers containing an anticoagulant mixture consisting of a thrombin inhibitor (PPACK), EDTA, and aprotinin (Byk-Sangtec); the ratio of anticoagulant to blood was 1:9 (vol/vol). The samples for the antithrombin III and aPTT assays were collected directly into refrigerated vacutainers containing sodium citrate (0.5 mL, 0.129 mol/L) as anticoagulant.
All of the samples were analyzed by investigators who were unaware of the clinical data. The plasma levels of prothrombin fragment 1+2 were measured with the use of a double-antibody radioimmunoassay as previously described.11 This method has an interassay coefficient of variation of ≈8%. Plasma fibrinopeptide A concentrations were determined in duplicate by means of an enzyme immunoassay in plasma extracted twice with bentonite to remove fibrinogen (Diagnostica Stago). This technique has an interassay coefficient of variation of ≈5%. Antithrombin III was assayed by means of an amidolytic method with the chromogenic substrate S2238 assembled into a kit by Chromogenix. The aPTTs were measured by an automated system.
The adverse outcome events considered were cardiac death, Q-wave or non–Q-wave myocardial (re)infarction, or recurrent ischemia at Holter monitoring. Cardiac death was defined as death as the result of cardiac causes. Q-wave myocardial (re)infarction was defined as a prolonged episode of chest pain accompanied by a subsequent rise in creatine kinase levels to more than twice the upper normal limit, with a corresponding increase in the MB fraction and the development of Q waves on the standard 12-lead ECG. The diagnosis of non–Q-wave myocardial infarction required only the first two characteristics. Recurrent ischemia was considered to have occurred in the event of at least one symptomatic or asymptomatic ischemic attack during the 72-hour Holter monitoring period.
Holter monitoring was performed with the use of a Delmar Avionics Electrocardiocorder model 445 with a frequency response of 0.05 to 100 Hz, which meets the specifications of the American Heart Association. The leads showing the most obvious ECG changes during spontaneous attacks were monitored; those with abnormal waves or significant ST-segment shifts were avoided. The system was calibrated before and after each placement. The tapes were analyzed at 60 times real-time under continuous visual inspection, and an episode of transient ischemia was defined as ≥1 mm ST-segment elevation or depression occurring 80 ms after the J-point, lasting for ≥1 minute and separated from other episodes by ≥1 minute. When a significant ST-segment change was noted on the monitor, the episode was recorded on ECG paper at 25 mm/s.
The study was approved by the Institutional Review Board of the Ca’ Granda Niguarda Hospital (Milan, Italy), and informed consent was obtained from all of the subjects. All of the clinical studies and informed consent procedures were also approved by the Committee on Clinical Investigations of the Beth Israel Hospital (Boston, Mass).
The deviations of the plasma concentrations of prothrombin fragment 1+2 and fibrinopeptide A from a normal distribution were tested by calculating the coefficients of skewness and kurtosis. Since the plasma levels of the coagulation system markers were found to be nonnormally distributed, repeated measures were compared by means of Friedman’s test and subsequent pairwise comparisons with baseline were made with the Wilcoxon signed-rank test with a downward adjustment of the α level to compensate for multiple comparisons. The upper normal limit of plasma prothrombin fragment 1+2 and fibrinopeptide A concentrations was calculated by determination of the 95th percentile of the distribution in a control group of age-matched healthy individuals and was set at 1.02 nmol/L for prothrombin fragment 1+2 and 2.2 nmol/L for fibrinopeptide A. Plasma antithrombin III levels of <80% were considered abnormal. Prevalences were compared by means of the χ2 test. Descriptive statistics include means and standard deviations or medians and interquartile ranges as appropriate. All of the tests are two tailed. Values of P<.05 were regarded as statistically significant.
We studied 44 patients with unstable angina and 20 patients with acute myocardial infarction. Four patients with unstable angina and 6 patients with myocardial infarction had a missing or inadequate blood sample and were therefore prospectively excluded from the analysis. The clinical and ECG characteristics of the study population are reported in Table 1⇓.
Coagulation Activation Markers and Hematological Parameters During Heparin Treatment
The median and 25th and 75th percentile values of plasma prothrombin fragment 1+2 at different time points are reported in Table 2⇓. There was no change from baseline in plasma prothrombin fragment 1+2 levels at 90 minutes, 24 hours, or 48 hours. There was no difference in the prevalence of abnormal plasma prothrombin fragment 1+2 levels at the different time points: 38 patients (72%) had abnormal plasma prothrombin fragment 1+2 levels at baseline, 34 (64%) at 90 minutes, 37 (70%) at 24 hours, and 36 (68%) at 48 hours.
The median and 25th and 75th percentiles of plasma fibrinopeptide A levels at the different time points in the study population are reported in Table 2⇑. Plasma fibrinopeptide A levels decreased significantly at 90 minutes and remained persistently lower than baseline at 24 hours and 48 hours. There was a significant decrease from baseline in the prevalence of abnormal plasma fibrinopeptide A levels at 90 minutes and at 24 and 48 hours (P<.0001): 27 patients (50%) had abnormal values of plasma fibrinopeptide A at baseline, 10 (19%) at 90 minutes, 14 (26%) at 24 hours, and 13 (25%) at 48 hours.
The median and 25th and 75th percentiles of antithrombin III percent activity at the different time points are reported in Table 2⇑. Antithrombin III activity did not change after 90 minutes but significantly decreased at 24 and 48 hours. Abnormal antithrombin III activity was found in 1 patient (2%) at baseline, 1 (2%) at 90 minutes, 6 (11%) at 24 hours, and 12 (22%) at 48 hours (P=.0004). The patient who had a preexisting antithrombin III deficiency was a 68-year-old man with normal baseline plasma fibrinopeptide A (0.6 nmol/L) and prothrombin fragment 1+2 (0.9 nmol/L) levels. During heparin treatment his antithrombin III levels further declined (75% at 90 minutes, 78% at 24 hours, and 68% at 48 hours), but there was no change in fibrinopeptide A (0.6, 0.8, and 0.6 nmol/L, respectively) or prothrombin fragment 1+2 (0.42, 0.24, and 0.23 nmol/L, respectively). The patient did not experience any in-hospital event.
The median and 25th and 75th percentiles of the aPTT at the different time points are reported in Table 2⇑. There was a significant increase from baseline aPTT at 90 minutes, 24 hours, and 48 hours. An aPTT ratio of more than double the baseline value was obtained in 48 patients (90%) at 90 minutes, in 36 (67%) at 24 hours, and in 34 (65%) at 48 hours.
The median and 25th and 75th percentiles of the heparin doses received per 24 hours for the first three 24-hour periods (0 to 24, 24 to 48, and 48 to 72 hours) are shown in Table 3⇓. There were no significant differences between any of the time periods.
Coagulation Activation Markers in Patients With In-Hospital Events
During the study period, 8 patients in the unstable angina group had persistent ischemia during the first 24 hours (6 symptomatic, 2 both symptomatic and asymptomatic), 6 between the 24th and 48th hours (2 symptomatic and 4 asymptomatic), and 6 between the 48th and the 72nd hours (3 symptomatic and 3 asymptomatic). In the myocardial infarction group, 2 patients had persistent ischemia during the first 24 hours (both symptomatic) and 1 developed a reinfarction between the 48th and 72nd hours.
The median prothrombin fragment 1+2 levels measured in the plasma sample drawn before the occurrence of an adverse outcome event were higher (1.62 nmol/L; interquartile range, 1.44 to 1.89) than those measured in the plasma sample drawn at the corresponding time point from the patients who did not develop an event (1.18 nmol/L; interquartile range, 0.75 to 1.75; P=.0027) (Fig 1⇓). The prevalence of abnormal prothrombin fragment 1+2 levels was significantly higher in the patients who developed an event (100%) than in those who did not develop an event (61%; P=.0003).
The median fibrinopeptide A levels measured in the plasma sample drawn before the occurrence of an adverse outcome event were higher (1.9 nmol/L; interquartile range, 1.3 to 3.6) than those measured in the plasma sample drawn at the corresponding time point from the patients who did not develop an event (1.2 nmol/L; interquartile range, 0.9 to 2; P=.0073) (Fig 2⇓). Thirty-eight percent of the patients who developed an event had abnormal fibrinopeptide A levels before the occurrence of the event in comparison with 22% of the patients who did not develop any event (P=NS).
The median aPTT and antithrombin III activity values measured in the plasma sample before the occurrence of an adverse event (aPTT, 46 seconds; interquartile range, 36 to 54; antithrombin III, 97%; interquartile range, 87 to 111 ) were no different from those measured in the plasma sample drawn at the corresponding time point from the patients who did not develop an event (aPTT, 56 seconds; interquartile range, 33 to 65; antithrombin III, 99%; interquartile range, 88 to 110; P=NS).
Heparin therapy improves clinical outcome in patients with acute coronary syndromes1 2 3 4 5 and has become part of the standard treatment for these patients. It has already been demonstrated that intravenous heparin reduces plasma fibrinopeptide A levels in the majority of patients with acute coronary syndromes.8 9 10 This effect is to be expected because heparin dramatically increases the efficiency of antithrombin III in inhibiting thrombin and therefore reduces the cleavage of fibrinogen and the formation of fibrinopeptide A.6 7 However, heparin is also able to inhibit other serine proteases such as activated factors XII, XI, IX, and, above all, activated factor X,7 and therefore should also reduce the thrombin generated by the activated factor X incorporated in the prothrombinase complex.
Although prothrombin fragment 1+2 is generally considered to be a marker of thrombin generation in vivo,11 12 13 its production is the result of a Xa-catalyzed cleavage at ARG 271-Thr, whereas thrombin comes from an additional cleavage at Arg 320-Ile. However, despite being only an indirect index of thrombin production, prothrombin fragment 1+2 has proved to be a useful marker of increased hemostatic system activity in humans. It is well established that prothrombin fragment 1+2 is elevated in the acute phase of unstable angina and myocardial infarction,14 but it is not known whether its plasma levels are affected by intravenous heparin under these conditions. Our study shows that intravenous heparin, given at doses capable of reducing plasma fibrinopeptide A levels and obtaining an aPTT within the therapeutic range in the majority of patients with acute coronary syndromes, does not reduce plasma prothrombin fragment 1+2 levels. This is the first report that addresses the relationship between heparin therapy and both thrombin generation and activity in patients with acute coronary syndromes. A previous study of porcine endotoxic shock has shown that the heparin–antithrombin III complex is capable of inhibiting thrombin activity but not thrombin generation, as evaluated by means of the consumption of prothrombin.15 Another study involving patients with venous thrombosis or pulmonary embolism already on heparin therapy has shown that thrombin generation, as measured by prothrombin fragment 1+2 and thrombin/antithrombin assays, gradually declines over the first few days of heparin administration but remains higher than in healthy control subjects, even after a week of treatment.16 Our data show that a heparin dose that adequately suppresses thrombin activity has no effect on plasma prothrombin fragment 1+2 levels either acutely or during continuous infusion over the first 48 hours of treatment. The behavior of prothrombin fragment 1+2 and fibrinopeptide A levels was similar in the patients with unstable angina and in those with myocardial infarction, thus suggesting that the different inhibition of thrombin activity and thrombin generation does not depend on the characteristics of the thrombus, which is subocclusive and platelet rich in unstable angina but occlusive and with a fibrin/red cell cap proximally or distally (or both) to the region of stasis in myocardial infarction.17
High Thrombin Generation and Activity and Clinical Events
Although this study was not designed to assess the relationship between thrombin generation or activity and prognosis, it is interesting to note that both plasma fibrinopeptide A and prothrombin fragment 1+2 level in the blood sample drawn from the patients who developed cardiac events before the occurrence of the event itself were significantly higher than those found in patients not developing any event. Moreover, the prevalence of abnormal prothrombin fragment 1+2 levels was also significantly higher in the group of patients developing an event. From these observations, we surmise that persistently increased thrombin generation and activity during heparin infusion can affect the risk of developing adverse events.
It is well known that heparin resistance is favored by the masking of antithrombin III receptors when thrombin is bound to the thrombus,18 19 by platelet factor 4 release from platelets, and by the formation of fibrin monomer II.20 The absence of any effect of heparin on thrombin generation in particular may be due to the fact that the factor Xa bound to activated platelets is protected against inactivation by the heparin-antithrombin complex21 and, since the degree of protection is known to be related to the degree of prothrombin activation, we speculate that patients with acute coronary syndromes may have a significant amount of factor Xa bound to activated platelets in the prothrombinase complex, which cannot be neutralized by heparin. In this case, factor Xa would generate thrombin that is prevented from acting on fibrinogen by the heparin-antithrombin complex. If this hypothesis is correct, the prothrombinase complex must be cleared from the blood before any treatment with heparin can be halted without leading to the risk of further formation of fibrin thrombi. Another possible explanation for the lack of inhibition of thrombin generation could be the slight but significant reduction in plasma antithrombin III levels observed in our patients during heparin therapy; however, this is an unlikely explanation because there was no change in prothrombin fragment 1+2 levels in the 90-minute sample, when antithrombin III concentrations were similar to those at baseline.
Although the different behaviors of fibrinopeptide A and prothrombin fragment 1+2 could be due to their different half-lives (3 to 5 minutes for fibrinopeptide A and 90 minutes for prothrombin fragment 1+2), the fact that persistently decreased fibrinopeptide A levels were found over a 48-hour period but no change in prothrombin fragment 1+2 was observed over the same period argues against this interpretation.
Our data show that during intravenous heparin therapy there is a persistent increase in plasma prothrombin fragment 1+2 (an indirect index of thrombin generation) in patients with acute coronary syndromes. Recent studies have shown that the infusion of hirudin, a high-affinity direct thrombin inhibitor, does not decrease thrombin generation at significant anticoagulant (aPTT, two to three times control) and antithrombotic doses in patients with stable angina.22 The results of these and our own in vivo studies are in contrast with those of in vitro studies, which have shown that hirudin and heparin suppress prothrombin activation by inhibiting prothrombinase formation,23 and these discrepancies underscore the complexity of prothrombinase regulation in vivo. Because thrombin plays a critical role in the amplification of the coagulation cascade by activating factor V24 and factor VIII,25 26 persistent thrombin generation may partially contribute to the persistent thrombotic risk during anticoagulation. Further studies are needed to test whether the suppression of thrombin generation by drugs that inhibit earlier steps in the coagulation cascade, such as drugs with a direct anti-Xa or anti–tissue factor action, would be more effective than heparin in reducing both thrombin generation and cardiac events in patients with acute coronary syndromes.
This study was supported in part by National Institutes of Health grant HL-3314. Dr Bauer is an Established Investigator of the American Heart Association.
Verstaete MJ, Chesebro JH, Fuster V. Antithrombotic therapy. In: Fuster V, Ross R, Topol EJ, eds. Atherosclerosis and Coronary Artery Disease. Philadelphia, Pa: Lippincott-Raven; 1995:979-994.
Mark DB, Braunwald E. Medical management of unstable angina. In: Fuster V, Ross R, Topol EJ, eds. Atherosclerosis and Coronary Artery Disease. Philadelphia, Pa: Lippincott-Raven; 1995:1315-1326.
Barrowcliffe TW, Thomas DP. Heparin and low molecular weight heparin. In: Bloom AL, Forbes CD, Thomas DP, Tiddenham EGC, eds. Hemostasis and Thrombosis. Edinburgh/London/Melbourne/New York: Churchill Livingstone; 1993:1417-1438.
Mombelli G, Hof I, Haeberli A, Straub PW. Effect of heparin on plasma fibrinopeptide A in patients with acute myocardial infarction. Circulation. 1984;69:684-689.
Teitel JM, Bauer KA, Lau HK, Rosenberg RD. Studies of the prothrombin activation pathway utilizing radioimmunoassays for the F2/prothrombin fragment 1+2 and thrombin/antithrombin complex. Blood. 1982;59:1086-1097.
Tripodi A, Mannucci PM. Markers of activated coagulation and their usefulness in the clinical laboratory. Clin Chem. 1996;42:664-669.
Merlini PA, Bauer KA, Oltrona L, Ardissino D, Cattaneo M, Belli C, Mannucci PM, Rosenberg RD. Persistent activation of coagulation mechanism in unstable angina and myocardial infarction. Circulation. 1994;90:61-68.
Weitz JI, Hudoba M. Mechanism by which clot-bound thrombin is protected from inactivation by fluid-phase inhibitors. Circulation. 1992;86(suppl I):I-413. Abstract.
Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound thrombin is protected from inhibition by heparin-antithrombin III but is susceptible to inactivation by antithrombin III-independent inhibitors. J Clin Invest. 1990;86:385-391.
Hogg PJ, Jackson CM. Fibrin monomer protects thrombin from inactivation by heparin-antithrombin III: implications for heparin efficacy. Proc Natl Acad Sci U S A. 1989;86:3619-3623.
Teitel JM, Rosenberg RD. Protection of factor Xa from neutralization by the heparin-antithrombin complex. J Clin Invest. 1983;71:1383-1391.
Zoldhelyi P, Bichler J, Owen WG, Grill DE, Fuster V, Mruk JS, Chesebro JH. Persistent thrombin generation in humans during specific thrombin inhibition with hirudin Circulation. 1994;90:2671-2678.
Nesheim ME, Mann KG. Thrombin catalyzed activation of single-chain bovine factor V. J Biol Chem. 1979;254:1326-1334.
Hoyer LW, Trabold NC. Effect of thrombin on human factor VIII: cleavage of the factor VIII procoagulant protein during activation. J Lab Clin Med. 1989;97:50-64.
Pieters J, Lindhout T, Hemker HC. In situ-generated thrombin is the only enzyme that effectively activates factor VIII and factor V in thromboplastin-activated plasma. Blood. 1989;74:1021-1024.