Rebound Increase in Thrombin Generation and Activity After Cessation of Intravenous Heparin in Patients With Acute Coronary Syndromes
Background The abrupt cessation of heparin and other thrombin inhibitors when used to treat acute coronary syndromes has been accompanied by a clustering of thrombotic events. It is unknown whether these events are the result of inadequate antithrombin therapy or whether they represent a rebound increase in thrombin activity. This study was designed to determine whether there is a true rebound, as defined by an increase followed by a subsequent decrease, in thrombin activity after discontinuation of intravenous heparin therapy.
Methods and Results Thirty-five patients with recent acute myocardial infarction or unstable angina who had received at least 48 hours of intravenous heparin were studied. Patients underwent ST-segment monitoring, and blood samples for determination of thrombin generation and activity were drawn at 0, 3, 6, 10, and 24 hours after heparin discontinuation. Median aPTT was 65 seconds before heparin discontinuation. Median fibrinopeptide A increased from 9.5 to 16.9 ng/mL at 3 hours (P<.0004) and returned to 10.5 by 24 hours. Prothrombin fragment 1.2 likewise transiently increased, from 0.34 to 0.51 nmol/L at 6 hours (P<.0002). Modified antithrombin III decreased over time (P<.002), and activated protein C increased from 2.3 to 4.5 ng/mL at 3 hours (P<.001). Although there were no clinical thrombotic events in the first 24 hours, 4 patients had evidence of ischemia by ST-segment monitoring at a median of 12 hours after heparin discontinuation. The degree of increase in fibrinopeptide A and prothrombin fragment 1.2 was not found to be associated with baseline diagnosis, duration of heparin therapy, baseline level of antithrombin III, or activated protein C.
Conclusions This study demonstrates a transient rebound increase in thrombin activity as well as in activated protein C upon abrupt discontinuation of intravenous heparin. Clinicians should be vigilant for associated thrombotic events. Further investigation of the significance, mechanism, and possible prevention of this rebound phenomenon is needed.
Heparin has become an integral part of the treatment of patients with acute coronary syndromes. This is supported by the finding that heparin results in improved clinical outcome in patients with unstable angina1 2 and acute myocardial infarction.3 The use of heparin has been accompanied by treatment failures, however, which typically have been attributed to subtherapeutic blood levels, insufficient duration of therapy, or resistance to heparin.4 5 6 Recently, investigators have identified a clustering of thrombotic events after the abrupt cessation of heparin and other thrombin inhibitors,7 8 which suggests that the benefits of antithrombin therapy may be limited by a reactivation of thrombosis after its discontinuation.
Biochemical evidence of increased thrombin activity after discontinuation of thrombin inhibitors may explain the reactivation of thrombotic events. After stopping a heparin infusion in 10 patients who had suffered recent acute myocardial infarction, Mombelli et al9 demonstrated a significant increase in fibrinopeptide A levels when measured 2 hours later. After stopping a 4-hour infusion of the direct thrombin inhibitor argatroban, there was a rebound increase in both thrombin-antithrombin complex levels and fibrinopeptide A levels in 43 patients with unstable angina.8 Nine of these patients developed recurrent unstable angina at a mean of 5.8 hours after stopping the argatroban. A possible mechanism suggested by Gold et al8 was a reduction of thrombin available to form thrombin-thrombomodulin complex, resulting in reduced activation of the natural anticoagulant protein C. No study has determined whether the increase in thrombin activity observed upon discontinuing antithrombin agents is a true “rebound” phenomenon, with a transient increase followed by a decrease, versus a removal of suppression of a continuing underlying thrombotic process.
The aim of the present study was to determine whether there is a true rebound increase in thrombin activity after discontinuing intravenous heparin in patients with recent coronary artery thrombosis. In this study, rebound was defined as a transient increase in thrombin activity followed by a decrease back to, near, or below the baseline value. A second goal of the study was to more fully characterize the balance of thrombin generation, neutralization, and inhibition by examining hematologic markers.
Thrombin activity was evaluated by measuring serial concentrations of fibrinopeptide A, a marker of fibrin formation; prothrombin fragment 1.2, a marker of thrombin generation; modified antithrombin III, a measure reflecting the amount of thrombin inactivated by antithrombin III; and activated protein C, which is both an indirect marker of thrombin activity as well as an endogenous anticoagulant. Patients were monitored for clinical evidence of thrombosis or ECG evidence of recurrent ischemia.
Patients who were on the Cardiology Service at Duke University Medical Center between February 15 and May 15, 1993, were screened for entry into the study. Thirty-five patients who fulfilled the following criteria were enrolled: (1) any age, (2) within 7 days of an acute myocardial infarction or unstable angina with at least 1 mm of ST-segment deviation, and (3) 48 hours or more of continuous intravenous heparin infusion before the study. The infusion was considered to be continuous if it had not been interrupted for longer than 1 hour. The use, dose, and duration of heparin therapy were directed by the patients’ individual physicians and were not dictated by the study. To promote compliance with limited activity during continuous ST-segment monitoring, patients who had recently undergone cardiac catheterization and had a femoral artery sheath in place were included. Patients were excluded from the study for a hematocrit less than 25%, hypotension, ongoing ischemia, active venous thrombosis, or a known primary disorder of coagulation. The study was approved by the Institutional Review Board, and informed consent was obtained from each patient before enrollment.
Continuous digital 12-lead ST-segment monitoring with the ST-100 (Mortara Instruments) was performed for 12 to 24 hours in 29 patients who met entry criteria and who (1) were confined to bed rest or were willing to remain at bed rest for at least 12 hours after heparin discontinuation and (2) had no conduction abnormalities or other baseline ECG elements confounding to ST-segment analysis. For the purposes of this study, “ischemic” ST deviation was defined as >100-mV change from stable baseline levels, measured at J plus 60 milliseconds, lasting for >60 seconds. All ST analyses were performed by an experienced cardiologist in the Ischemia Monitoring Core Laboratory at Duke who was blinded to all hematologic or other clinical data other than the time of heparin discontinuation. The Core Laboratory’s methods of 12-lead ECG superimposition editing and analyses have been described previously.10
The medical records of all patients were reviewed for clinical events, which were documented for the first 24 hours after heparin cessation and during hospitalization (up to 30 days). Clinical events included thrombotic cardiovascular events of recurrent ischemia (pain with ST-segment or T-wave changes), myocardial infarction, deep venous thrombosis, pulmonary embolism, stroke, emergency coronary artery bypass graft surgery, emergency percutaneous transluminal angioplasty, and death.
Blood Sampling and Processing
Baseline blood samples were obtained at 0 to 1 hour before cessation of heparin infusion. Additional samples were obtained at 3, 6, 10, and 24 hours after discontinuation of heparin. The sampling frequency was based on previously published data that reactivation of events occurred approximately 5 to 10 hours after heparin discontinuation.7 Blood samples before the initiation of heparin were not obtained because of logistic difficulties and because many patients had heparin therapy initiated before transfer to our institution. Blood samples were collected for assay of heparin levels (anti-Xa), thrombin generation (prothrombin fragment 1.2), thrombin activity toward fibrinogen (fibrinopeptide A), thrombin inactivation (modified antithrombin III), activated protein C, and antithrombin III.
Each sample of free-flowing blood was collected through a fresh venipuncture site distal to any intravenous catheters present using a 21-gauge butterfly Vacutainer collection set (Becton Dickinson) with 12-in. tubing with multiple luer adapter. Although arterial sheaths were in place in most patients, none of the study blood samples were drawn through the sheaths. Tubes and collection sets were stored at 4°C until used. All collections were made by trained nurses or physicians. Blood collected during heparin infusion was obtained from the arm contralateral to the infusion. For all phlebotomies, the first 5 mL of blood was discarded. Three tubes were collected at all time points: a 4.5-mL sodium citrate tube; a 5-mL tube with lyophilized d-phenylalanyl-prolyl-arginyl chloromethyl ketone (2×10−5 mol/L, final concentration),11 trasylol and aprotinin; and a 5-mL Vacutainer tube containing 0.5 mL of citrate-benzamidine.12 The samples were immediately placed on ice and spun within 60 minutes of sample collection at 2000 to 3000g for 10 to 15 minutes at room temperature. The supernatant was immediately frozen at −20°C to −70°C.
Because fibrinopeptide A is known to be susceptible to in vitro sampling artifact, blood samples with a fibrinopeptide A of greater than 50 ng/mL were not used for determination of any markers for the analysis because these samples were most likely affected by poor sample acquisition, handling, or storage.13 14 Of the 141 samples drawn on patients included, 9 (6.4%) were removed from analysis for that reason.
All assays other than activated protein C were performed at the University of Vermont College of Medicine. Fibrinopeptide A levels were quantitated using the “RIA-mat” fibrinopeptide A radioimmunoassay (Mallinckrodt) using bentonite to absorb plasma fibrinogen. The fibrinopeptide A antibody used has a cross-reactivity to fibrinogen of <4% and a lower detection limit of 0.1 ng/mL. The 5th to 95th percentile range for our laboratory was 1.2 to 9.2 ng/mL, with a median of 2.4 ng/mL in 172 individuals between 65 and 75 years of age. Prothrombin fragment 1.2 was determined using a two-site, enzyme-linked immunosorbent assay (Dade, Baxter Diagnostics Inc) using monoclonal antiprothrombin fragment 1.2 directed against antigenic sites not available in the native prothrombin molecule, and the 5th to 95th percentile reference range for our laboratory was 0.15 to 0.62 nmol/L, with a median of 0.29 nmol/L in 172 individuals between 65 and 75 years of age. The coefficient of variation (CV) for the fibrinopeptide A assay was 7.2% and for the prothrombin fragment 1.2 assay was 10.0%.
An enzyme immunoassay of modified antithrombin III with the 4C9 monoclonal antibody (American Bioproducts) was used. Antithrombin III forms complexes with serine proteases XIa, IXa, Xa, and IIa, inhibiting these enzymes and at the same time modifying the structure of antithrombin III. Because the predominant enzyme complexing with antithrombin III is thrombin, modified antithrombin III is an approximation of thrombin-antithrombin complex. The 4C9 monoclonal antibody does not recognize prothrombin, thrombin, or native antithrombin III but strongly reacts with modified antithrombin III.15 The short half-life of modified antithrombin III allows it to be a rapidly responsive marker of thrombin inactivation by the formation of thrombin-antithrombin complexes. The CV for this assay in our laboratory was approximately 11%, and the reference range was <20 ng/mL.
Heparin levels were measured by anti-Xa concentration by the amidolytic method (American Bioproducts). The CV for this assay was 8.4%. Antithrombin III levels also were measured chromogenically (American Bioproducts), with a CV of 6.4%.
Activated protein C was measured using an enzyme capture assay performed at The Scripps Research Institute.12 Microwells on plastic microplates were coated with a murine monoclonal antibody directed against the light chain of human protein C. Activated protein C was absorbed from plasma samples containing citrate and benzamidine to prevent inactivation of activated protein C by its plasma inhibitors. Benzamidine and other unbound sample constituents were removed by washing, and the amidolytic activity of the captured enzyme, activated protein C, was measured using a chromogenic substrate. Activated protein C activity in normal pooled plasma was 2.26±0.2 ng/mL.
aPTTs were assayed at Duke University Medical Center using standard methods with Dade actin FS reagent.
The sample size was determined based on data from previous reports concerning thrombin inhibitors,8 9 with an expected fibrinopeptide A level during the heparin infusion of 10±10 ng/mL and an expected fibrinopeptide A level 10 hours after stopping heparin of 25±20 ng/mL. With an estimated difference of 15 ng/mL, a standard deviation of 20, α=.05, and power of .90, a sample size of approximately 30 was deemed sufficient.
Descriptive statistics of data include median values with interquartile ranges (25th and 75th percentiles) for continuous characteristics and frequencies for discrete variables. Analysis included univariate repeated measures (ANOVA) for any difference in fibrinopeptide A, prothrombin fragment 1.2, modified antithrombin III, and activated protein C over time. Baseline marker levels were compared by paired t test with those of subsequent sampling times. Global trends tests were used to evaluate changes, with stepdown tests performed to determine where the most significant differences occurred. Wilcoxon signed rank test was used to compare modified antithrombin III values (which did not follow a normal gaussian distribution). Total time of treatment with heparin, baseline levels of antithrombin III, and activated protein C were examined as possible determinants of fibrinopeptide A or prothrombin fragment 1.2 after heparin cessation.
Of the 35 patients enrolled in the study, 3 were excluded from analysis because of either a missing or an inadequate baseline blood sample for hematologic analysis. Characteristics of the remaining 32 patients are shown in Table 1⇓. All patients had unstable angina (37%), acute myocardial infarction (35%), or both (28%) within 1 week of enrollment. Patients had been on intravenous heparin for a median of 110 hours (interquartile range, 81 to 151 hours). Within 1 week, 56% of patients had undergone angioplasty. Twenty-nine patients had femoral artery sheaths in place at the time of enrollment, which were removed a median of 4.7 hours later. All patients were receiving oral aspirin, 325 mg per day.
ECG and Clinical Outcomes
Four of the 29 patients with continuous ST-segment monitoring had ischemic episodes of ST depression a median of 12 hours (range, 1.5 to 22.5 hours) after discontinuing heparin. None of the episodes of ST-segment deviation was accompanied by recognized angina. Five patients had clinical evidence of ischemia but none within 24 hours of discontinuing heparin (range, 2 to 9 days). One patient had urgent coronary artery bypass surgery 2 days after discontinuing heparin. No patient suffered myocardial infarction, stroke, or death.
Median heparin level (anti-Xa) just before heparin discontinuation was 0.14 U/mL (interquartile range, 0.06 to 0.38), median aPTT was 65 seconds, and median antithrombin III level was 66.8% (interquartile range, 63.2 to 82.4). Median fibrinopeptide A, prothrombin fragment 1.2, modified antithrombin III, and activated protein C levels are illustrated in Figs 1 through 4⇓⇓⇓⇓. Median fibrinopeptide A increased from a baseline of 9.5 ng/mL to a maximum of 16.9 at 3 hours and decreased to 10.5 by 24 hours. Prothrombin fragment 1.2 increased from 0.34 nmol/L to a maximum median level of 0.51 nmol/L at 6 hours and back to 0.43 nmol/L by 24 hours. Median activated protein C level increased from 2.3 to 4.5 ng/mL at 3 hours and fell to 2.5 ng/mL by 24 hours. Each of these assays had a statistically significant change (P<.001 by global trend testing) as well as from baseline to the peak value (P<.001 for all three assays). Modified antithrombin III had a change in the opposite direction, starting at a median of 20.3 ng/mL and decreasing by 10 hours to a median of 16.6 ng/mL (P<.002), with a further decrease to 15.6 by 24 hours. The test of the original primary hypothesis—that fibrinopeptide A and prothrombin fragment 1.2 would be higher at 10 hours than at baseline—was statistically significant for prothrombin fragment 1.2 but not for fibrinopeptide A because the difference had peaked at 3 hours and was returning toward baseline by 10 hours. From peak values to 10-hour values there was a statistically significant decrease for fibrinopeptide A (P=.003) and for activated protein C (P=.04).
Among the 4 patients who experienced silent ischemia measured by continuous ST-segment monitoring, no difference in the degree of increase in fibrinopeptide A and fragment 1.2 could be demonstrated compared with patients without silent ischemia (Fig 5⇓), and we may have failed to identify a larger increase among patients with silent ischemia because of the small sample size. Likewise, there was no difference in the change in fibrinopeptide A and fragment 1.2 in the 5 patients with clinical evidence of ischemia compared with patients without clinical ischemia.
The median increases in fibrinopeptide A and fragment 1.2 in clinical categories—enrollment diagnosis of acute myocardial infarction or unstable angina, recent treatment with thrombolytic therapy, and recent angioplasty—are shown in Fig 6⇓. The increase in fibrinopeptide A and prothrombin fragment 1.2 was a consistent finding across the subgroups.
Tests to correlate the degree of increase in thrombin activity, as measured by both increases in fibrinopeptide A from 0 to 3 hours and increases in prothrombin fragment 1.2 from 0 to 6 hours, with other prespecified parameters, are shown in Table 2⇓. No significant relation was found between duration of heparin treatment before discontinuation and increase in thrombin activity. Baseline levels of antithrombin III and activated protein C were not significantly correlated with degree of increase in thrombin activity. We identified no significant association of heparin level at the time of heparin discontinuation and degree of increase in fibrinopeptide A (r=−.03, P=.86) nor of the baseline level of fibrinopeptide A or prothrombin fragment 1.2 and the subsequent degree of rise (P=.22 and .23, respectively).
Timing of Sheath Removal
During the study period of 24 hours after heparin discontinuation, 26 of the 32 patients had femoral arterial sheaths removed a median of 4.5 hours (interquartile range, 3.5 to 5.5 hours) after heparin was stopped. Three of the 4 patients who did not have sheaths removed during the study period and who had complete samples had patterns similar to the other patients of increase and subsequent decrease of fibrinopeptide A and prothrombin fragment 1.2. We found no temporal relation between the removal of the sheath and the peak of the thrombin activity.
The major finding of this study was a rebound increase in thrombin generation and activity after discontinuation of intravenous heparin. Compared with the levels just before stopping heparin, both prothrombin fragment 1.2 and fibrinopeptide A rose significantly at 3 and 6 hours and decreased toward baseline levels by 24 hours. Measures of activated protein C, reported here for the first time from patients with acute ischemic heart disease, demonstrated a similar transient increase that paralleled prothrombin fragment 1.2 and fibrinopeptide A.
The patient population studied was typical of acute coronary syndromes, with two thirds of patients with acute myocardial infarction and slightly over half of the myocardial infarctions treated with thrombolytic therapy. Patients had been on heparin for an average of 4 to 5 days. The median aPTT was in the target range at 65 seconds, although the median heparin level (anti-Xa) was less than .2, which is lower than what is considered the optimal therapeutic plasma level.4
Hematologic Markers of Thrombin Activity
The commonly used hematologic markers of thrombin activity include fibrinopeptide A, a sensitive marker of fibrin formation.16 17 18 Prothrombin fragment 1.2 is a product and measure of thrombin generation.18 19 If thrombin were generated and subsequently inactivated, prothrombin fragment 1.2 would be increased in the absence of any increase in active thrombin. Therefore, prothrombin fragment 1.2 is a less direct measure of actual thrombin activity than fibrinopeptide A. Median values of both fibrinopeptide A and prothrombin fragment 1.2 were within the normal reference range before heparin discontinuation. Similar to the findings of Mombelli et al,9 we found an increase in fibrinopeptide A upon heparin discontinuation. By 3 hours after discontinuation, median fibrinopeptide A levels had increased by 78%, and the maximum detected rise (50%) in prothrombin fragment 1.2 occurred at 6 hours. By 24 hours, both measures of thrombin activity had decreased.
To prove that a “rebound” increase in thrombin activity exists, one must show not only that thrombin activity increases after stopping heparin but that the transient increase in thrombin activity is unrelated to underlying or suppressed thrombosis. The ideal demonstration of a rebound phenomenon would involve showing that the markers of thrombosis were completely suppressed on therapy, transiently elevated upon stopping heparin, and returned to normal levels within 24 hours.
In this study, the increase in thrombin activity upon stopping heparin was transient and was already decreasing toward normal within 3 to 6 hours, indicating that the increase in thrombin activity had not been simply because of “desuppression” of underlying thrombosis but rather appeared to be a true rebound or overshoot in thrombin activity after abrupt discontinuation of heparin. If the increase in thrombin activity was due to a pure unmasking of underlying thrombosis, one would expect that patients with incomplete suppression while on heparin might have more pronounced increased thrombin activity upon heparin discontinuation. This was not observed; there was no association between the degree of increase in thrombin activity and the degree of thrombin activity before heparin discontinuation.
It is possible that having an arterial sheath in place caused systemic activation of thrombin, as has been demonstrated by Nichols et al,14 and that the decrease in thrombin activity after 6 hours was in part due to removal of the femoral arterial sheath. This is unlikely to be the primary explanation for this finding, however, because the decrease was also observed in patients who did not have arterial sheaths, there was no temporal relation between the decrease in thrombin activity and the timing of sheath removal, and the effect on activation markers from an arterial sheath would be expected to be small.
Another limitation of this study is the lack of levels of thrombin activity before initiation of heparin, which would allow more complete characterization of the effect of heparin on thrombin activity over time.
Modified antithrombin III, which consists primarily of thrombin-antithrombin complex, has generally been considered to be a measure of the circulating thrombin level because it increases with increased available thrombin. Interestingly, the level of modified antithrombin III changed in the opposite direction from fibrinopeptide A and prothrombin fragment 1.2, with the highest level at baseline and a gradual fall after heparin discontinuation. This may be because the generation of modified antithrombin III is dependent not only on thrombin activity but also on the affinity of antithrombin III for thrombin. In the absence of the influence of heparin to increase the affinity of antithrombin III for thrombin 1000-fold, even with increased thrombin activity, the level of modified antithrombin III could decrease.
Activated Protein C
Activated protein C is a natural anticoagulant; it is an important regulator of thrombosis11 20 and has never been reported before from a group of patients with acute coronary thrombosis. Although activated protein C inhibits factors Va and VIIIa, it is also an indicator of thrombin activity, because thrombin when bound to thrombomodulin is a major activator of protein C. In this study, the change in activated protein C appeared to be more influenced by thrombin activity than vice versa. Activated protein C followed the same general pattern as fibrinopeptide A and prothrombin fragment 1.2: It was in the normal range before heparin discontinuation, it increased significantly upon discontinuation, and it returned to its prior level within 24 hours. The consistency of the changes in these levels supports the conclusion that the transient increase in thrombin activity is a true physiological event. In addition to the increase in thrombin activity leading to increased protein C activation, the transient increase in activated protein C may have been in part the result of depletion of inhibitors, resulting in decreased clearance.
ECG and Clinical Outcomes
Although no patient in this study had a clinical thrombotic event in the 24 hours after discontinuing heparin, 4 patients had evidence of silent ischemia on ST-segment monitoring during a time frame consistent with previous reports of clustering of ischemic events after discontinuation of heparin.7
We found no association between the degree of increase in thrombin activity and the presence of asymptomatic ischemia by ST-segment monitoring nor the presence of subsequent clinical ischemic events. This study, however, was not designed to compare the degree of thrombin activity among patients with and without clinical events, which would require a substantially larger sample size. For example, assuming a 10% incidence of clinical events and a 25% higher increase in fibrinopeptide A in patients with events compared with patients without events, approximately 1000 patients would be needed to have an 80% power to demonstrate a statistically significant difference. Therefore, the findings of this study should not be interpreted as inconsistent with the hypothesis that rebound in thrombin activity correlates with clinical outcomes.
Although there are only two studies that demonstrate the clinical phenomenon of reactivation of thrombosis after discontinuation of antithrombin therapy, there is considerable clinical concern regarding this issue. This concern is reflected in the National Heart, Lung, and Blood Institute–sponsored “Clinical Practice Guideline” for unstable angina,21 in which a gradual discontinuation of heparin therapy by switching from intravenous to subcutaneous heparin is suggested as a possible means to reduce the risk of rebound. The addition of aspirin appeared to prevent the clinical reactivation of thrombosis in the study by Theroux et al7 but not in the study by Gold et al.8 Although the point estimate for the incidence of reactivation events in the Theroux study with heparin and aspirin was 5%, the 95% confidence interval ranged up to 9%, consistent with a possible significant rebound effect even with aspirin. Definitive demonstration of a preventive effect of aspirin for clinical rebound would require a much larger number of patients than have been studied to date. For example, assuming the addition of aspirin to heparin compared with heparin alone reduced reactivation events by 50%, from 12% to 6%, a total of 375 patients per group would be required to have an 80% chance of demonstrating a statistically significant difference.
Several mechanisms have been proposed for a rebound increase in thrombin activity after discontinuing heparin or other antithrombin agents.8 15 22 23 The general mechanism may involve an accumulation of prothrombotic factors when thrombin is inhibited, which results in a transient hypercoagulable state upon discontinuation. Specific mechanisms may include downregulation of factors that inhibit thrombin activity—antithrombin III or activated protein C—while on heparin. Although downregulation of antithrombin III is known to occur in patients on heparin,23 and this has been hypothesized to be an explanation for hypercoagulability in the early hours after heparin discontinuation, we did not identify a correlation of decreased levels of antithrombin III with increased thrombin activity. Likewise, a decrease in protein C activation while receiving heparin could result in a hypercoagulable state after heparin discontinuation, but we did not find evidence to support this. Lower levels of activated protein C did not correlate with greater increases in thrombin activity. However, a relation between fall in activated protein C or fall in antithrombin III and rebound increase in thrombin activity may have been unrecognized because baseline levels before beginning heparin therapy were not obtained.
Although reactivation of clinical thrombosis after discontinuing heparin has been documented previously, it was not known before this study whether this was due to inadequate degree and duration of antithrombin therapy or to a true rebound increase in thrombin activity upon discontinuation. This study suggests that the increase in thrombin activity is at least in part due to a true rebound phenomenon. The clinical significance (particularly in the presence of aspirin), the mechanism, and the methods for prevention of this phenomenon are yet to be established. Gradual discontinuation of heparin might attenuate the rebound response.21 The findings of this study suggest that vigilance for thrombotic events upon rapid discontinuation of intravenous heparin is warranted, and continued investigation for increased thrombin activity upon discontinuation of new antithrombin agents is needed.
This study was supported in part by research grants HL-42490 and HL-36587 from the National Heart, Lung, and Blood Institute, Bethesda, Md; research grant HS-05636 from the Agency for Health Care Policy and Research, Rockville, Md; the Cigarette and Tobacco Surtax Fund of the State of California through the Tobacco-Related Disease Research Program of the University of California, grant 2RT0342; and the Stein Endowment Fund.
- Received September 21, 1994.
- Accepted November 6, 1994.
- Copyright © 1995 by American Heart Association
Hull RD, Raskob GE, Hirsh J, Jay RM, Leclerc JR, Geerts WH, Rosenbloom D, Sackett DL, Anderson C, Harrison L, et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximal-vein thrombosis. N Engl J Med. 1986;315:1109-1114.
Turpie AGG, Robinson JG, Doyle DJ, Mulji AS, Mishkel GJ, Sealey BJ, Cairns JA, Skingley L, Hirsh J, Gent M. Comparison of high-dose with low-dose subcutaneous heparin to prevent left ventricular mural thrombosis in patients with acute transmural anterior myocardial infarction. N Engl J Med. 1989;320:352-357.
Gold HK, Torres F, Garabedian H, Werner W, Jang I, Khan A, Hagstrom J, Yasuda T, Leinbach R, Newell J, et al. Evidence for a rebound coagulation phenomenon after cessation of a 4-hour infusion of a specific thrombin inhibitor in patients with unstable angina pectoris. J Am Coll Cardiol. 1993;21:1039-1047.
Mombelli G, Hof VI, Haeberli A, Straub PW. Effect of heparin on plasma fibrinopeptide A in patients with acute myocardial infarction. Circulation. 1984;69:684-689.
Krucoff MW, Croll MA, Pope JE, Granger CB, O’Connor CM, Sigmon KN, Wagner BL, Ryan JA, Lee KL, Kereiakes DJ, et al, for the TAMI 7 Study Group. Continuous 12-lead ST-segment recovery analysis in the TAMI 7 study: performance of a noninvasive method for real-time detection of failed myocardial reperfusion. Circulation. 1993;88:437-446.
Tracy R, Bovill E, Stump D, Lin T, Gomol T, Collen D, Mann K. Reduction of in vitro artifact during blood collection in TIMI II. Blood. 1988;72(suppl 1):376A. Abstract.
Gruber A, Griffin JH. Direct detection of activated protein C in blood from human subjects. Blood. 1992;79:2340-2348.
Nichols AB, Owen J, Grossman BA, Marcella JJ, Fleisher LN, Lee MM. Effect of heparin bonding on catheter-induced fibrin formation and platelet activation. Circulation. 1984;70:843-850.
Amiral J, Vissac AM, Mimilla F, Grosley B. Assay of blood activation by measurement of AT III-serine esterases complexes (ATM) using a specific monoclonal antibody 4C9. Thromb Haemost. 1989;62:479. Abstract.
Bilezikian SB, Nossel JL, Butler VP Jr, Canfield RE. Radioimmunoassay of human fibrinopeptide B and kinetics of fibrinopeptide cleavage by different enzymes. J Clin Invest. 1975;56:438-445.
Eisenberg PR, Sherman LA, Schectman K, Perez J, Sobel BE, Jaffe AS. Fibrinopeptide A: a marker of acute coronary thrombosis. Circulation. 1985;71:912-918.
Bauer KA, Weitz JI. Laboratory markers of coagulation and fibrinolysis. In: Colman R, ed. Hemostasis and Thrombosis. 3rd ed. New York: Lippincott; 1994:1197-1210.
Teitel JM, Bauer KA, Lau HK, Rosenberg RD. Studies of the prothrombin activation pathway utilizing radioimmunoassays for the F2/F1+2 fragment and thrombin-antithrombin complex. Blood. 1982;59:1086-1097.
Esmon CT. The regulation of natural anticoagulant pathways. Science. 1987;235:1348-1352.
Braunwald E, Mark DB, Jones RH, Cheitlin MD, Fuster V, McCauley KM, Edwards C, Green LA, Mushlin AI, Swain JA. Unstable Angina: Diagnosis and Management. Clinical Practice Guideline Number 10. AHCPR Publication No. 94-0602. Rockville, Md: Agency for Health Care Policy and Research and the National Heart, Lung, and Blood Institute, Public Health Service, US Department of Health and Human Services; 1994:65.