Author + information
- Received July 10, 2000
- Revision received November 9, 2000
- Accepted December 13, 2000
- Published online March 15, 2001.
- Maria Cecilia Bahit, MD∗,* (, )
- Eric J Topol, MD, FACC†,
- Robert M Califf, MD, FACC∗,
- Paul W Armstrong, MD, FACC‡,
- Douglas A Criger, MPH∗,
- Vic Hasselblad, PhD∗,
- Amadeo Betriu, MD§,
- Jack Hirsh, MD, FACC∥,
- Diego Ardissino, MD¶ and
- Christopher B Granger, MD, FACC∗
- ↵*Reprint requests and correspondence: Dr. Maria Cecilia Bahit, Duke Clinical Research Institute, PO Box 17969, Durham, North Carolina 27715
We sought to determine the incidence of and risk factors for thrombotic events early after discontinuing antithrombin therapy in patients with acute coronary syndromes.
Discontinuation of treatment with heparin and other thrombin inhibitors in patients with unstable coronary syndromes has related to clinical and biochemical evidence of early reactivation of thrombosis.
We studied 8,943 of the 12,142 patients with acute coronary syndromes enrolled in the Global Use of Strategies To Open occluded arteries in acute coronary syndromes trial of hirudin versus heparin. We excluded patients who received no study drug, lacked timing data, died or had myocardial (re)infarction [(re)MI] during study-drug infusion, or began heparin treatment within 2 h after treatment with the study drug was stopped. We assessed the incidence and timing of (re)MI by type and timing of antithrombin treatment.
In all, 215 patients (2.4%) suffered (re)MI, 49 within 12 h of antithrombin therapy discontinuation and 166 between hour 12 and hospital discharge. The duration of infusion did not differ between the hirudin and heparin groups. The rate of early re(MI) after drug therapy discontinuation was significantly higher in patients given heparin versus hirudin (0.8% vs. 0.3%, p = 0.002). Patients with (re)MI had higher mortality at 30 days (23.6% vs. 2.4%, p = 0.001) and 1 year (35.2% vs. 6.7%, p = 0.001) compared with patients without (re)MI.
The incidence of (re)MI was clustered within 12 h of heparin therapy discontinuation, with the greatest risk within 4 h. There was no evidence of early reactivation of thrombotic events after hirudin. Patients who had (re)infarction had worse outcomes. Better understanding of the mechanism and possible prevention of recurrent thrombosis is needed.
Acute coronary syndromes are triggered by disruption of atherosclerotic plaques with resultant activation of platelets and the coagulation system. Given the central role of platelets and thrombin in arterial thrombosis, heparin and aspirin have become cornerstones of the treatment of patients with acute coronary syndromes (1,2).
Although heparin has been widely used, it has several limitations. These limitations include the inability of heparin to inactivate thrombin bound to fibrin, the requirement of antithrombin III as a cofactor and the neutralization of heparin by circulating proteins, including platelet factor 4. Consequently, the anticoagulant response to heparin varies from patient to patient and requires careful monitoring. Direct thrombin inhibitors, such as hirudin, overcome several of these limitations of heparin. Hirudin inhibits fibrin-bound thrombin and fluid-phase thrombin. Unlike heparin, hirudin does not bind to plasma proteins and results in a more predictable anticoagulant response (3). Despite the promise of thrombin inhibitors, however, clinical trials of hirudin have failed to show a clear and durable advantage over heparin (4–6).
In addition, discontinuation of treatment with both heparin (7–9)and thrombin inhibitors (10,11)is associated with an increased short-term risk of thrombotic cardiac events. Possible mechanisms involved in the reactivation of thrombosis include a persistent active underlying prothrombotic state, transient hypercoagulability (“rebound”) as reflected by increased thrombin generation and activity after treatment with thrombin inhibitors is stopped (12–16)or both. Recently, investigators have tested different strategies of heparin therapy discontinuation to prevent or attenuate the rebound phenomenon, but without success (17).
The Global Use of Strategies To Open occluded arteries in acute coronary syndromes (GUSTO-IIb) study provides a unique opportunity to describe the clinical pattern of myocardial (re)infarction [(re)MI] after discontinuation of antithrombin therapy in a large cohort. We sought to investigate the frequency, timing and natural history of (re)MI after discontinuing intravenous heparin and hirudin therapy, and to determine whether different forms of anticoagulation influence the profile.
The GUSTO-IIb trial (5)compared the efficacy of recombinant hirudin versus heparin in patients with acute coronary syndromes. To be eligible, patients had to be within 12 h of chest pain onset and have electrocardiographic (ECG) abnormalities of either ST-segment elevation or depression of ≥0.05 mV or T-wave inversion of ≥0.1 mV. Patients with a history of active bleeding, serum creatinine >2.5 mg/dl, stroke within the past year or a contraindication to heparin were excluded. The study enrolled 12,142 patients, who were stratified by the presence of ST-segment elevation. We excluded 197 patients who did not receive study drug, 304 patients who lacked drug-timing information, 134 patients who died before or during study-drug infusion, 232 who had (re)MI during study-drug infusion and 2,668 in whom intravenous heparin therapy was restarted within 2 h of study-drug discontinuation. This left 8,943 patients for this analysis (Fig. 1).
To compare the rates of in-hospital (re)MI during and after study-drug infusion, however, we included the 122 patients who had (re)MI during infusion in a separate population, for a total of 9,065 patients. Both analysis populations were stratified by the occurrence and timing of in-hospital (re)MI: none, (re)MI within 12 h after study drug therapy was stopped, or (re)MI from 12 h after treatment with the study drug was stopped to hospital discharge.
Patients with ST-segment elevation were eligible to receive thrombolytic therapy of either accelerated alteplase or streptokinase. All patients were randomly assigned to receive blinded heparin or recombinant hirudin. Hirudin (desirudin) was given as an intravenous bolus of 0.1 mg/kg, followed by a continuous infusion of 0.1 mg/kg/h. Heparin was given as an intravenous bolus of 5,000 U, followed by a continuous infusion of 1,000 U/h. Both infusions were maintained for 76 to 120 h.
Definition of (re)infarction
A blinded, independent adjudication committee reviewed and classified all suspected (re)MIs. An MI was considered present at enrollment if the creatine kinase (CK)-MB level was above normal (and ≥3% of total CK) for samples at 0 h (baseline) or 8 h after enrollment, or both. If the CK-MB was elevated only at the 16-h sample, and no symptoms had occurred between enrollment and this time, the event also was considered an MI at enrollment. If the CK-MB was elevated only at 16 h, and symptoms consistent with MI had occurred after enrollment, the review committee coded the event according to data collected from the ECG, symptoms and enzyme measurements. If CK-MB was unavailable, then total CK was required to be >2 times the upper limit of normal. An MI also was to be classified for new, significant Q-waves in at least two contiguous leads.
For patients who had an MI before enrollment, a new MI was defined as a rise in CK-MB to above-normal or to at least twice the prior value (if it had been above the upper limit), with appropriate signs, symptoms and ECG changes.
All analyses were performed using SAS software (SAS, Cary, North Carolina). Prospectively specified baseline and in-hospital variables were compared among the analysis subgroups by means of the chi-square test for trend for categorical variables and the Wilcoxon rank-sum test for continuous variables. End points reviewed by stratum were 30-day and one-year mortality. To compare the (re)MI rates during versus after study-drug infusion, the time “on” and “off” drug therapy was divided into 4-h blocks. Each patient was observed until (re)MI or hospital discharge. Patients with missing transfer or hospital discharge data were assumed to be alive and were followed up to 30 days.
The SAS GENMOD procedure was used to estimate event rates for three drug-duration periods (0 h to <12 h on drug, 12 h to <24 h on drug, and >24 h on drug), and four such periods after drug therapy was stopped (0 to <4 h off drug, 4 to <8 h off drug, 8 to <12 h off drug and >12 h off drug). We compared the (re)MI rate during study-drug infusion with the rates during the first three 4-h blocks (that is, the first 12 h postinfusion) because, from the Montreal study (7)and Global Utilization of Streptokinase and TPA (alteplase) for Occluded coronary arteries (GUSTO-I) trials (8), this is when most reactivation events are thought to occur.
Overall, 215 patients (2.4% of the 8,943 patients included in this analysis) suffered (re)MI between study-drug discontinuation and hospital discharge (Table 1). Of these, 49 (0.5%) occurred within 12 h after treatment with study drug was stopped, and 166 (1.8%) occurred between hour 12 and hospital discharge. Advanced age, female gender, diabetes and prior nitrate use all were significantly more common (p < 0.01) among patients suffering (re)MI, and patients with early reinfarction tended to have greater excess of these variables. ST segment stratum, however, did not differ significantly by the occurrence of (re)MI.
Although the median activated partial thromboplastin times (aPTTs) did not differ significantly between patients with and without (re)MI at 12 and 24 h, they tended to be higher among patients with late (re)MI. Likewise, no differences were seen in the median aPTTs before study-drug therapy discontinuation: 64 s for patients without (re)MI, 60 s for patients with (re)MI within 12 h, and 64 s for patients with reMI after 12 h (p = 0.11).
Incidence and timing of (re)MI by antithrombin treatment
In all, 97 patients in the hirudin group and 118 in the heparin group suffered (re)MI after treatment with study drug was stopped (2.2% vs. 2.7%, p = 0.12) (Table 2). The median time to (re)MI was 41 h (25th, 75th percentile, 18, 82 h) after hirudin therapy was stopped, and 34 h (9, 91 h) after heparin therapy was stopped (p = 0.14). The median duration of infusion did not differ between the two groups (72.6 h for hirudin vs. 72.8 h for heparin).
The occurrence and timing of (re)MI according to administration of heparin and hirudin is shown in Figure 2. During study-drug administration, there was a significantly lower incidence of (re)MI with hirudin (5). During the 12 h after discontinuation, patients who received heparin had more (re)MIs than those given hirudin (35 of 4,434 [0.8%] vs 14 of 4,509 [0.3%], p = 0.002). From 12 h after discontinuation until 120 h, there was a nonsignificant trend toward more reinfarctions in the hirudin group.
The incidence of (re)MI per 4-h time interval, according to drug treatment and time of study-drug administration, is shown in Figure 3. During the first 24 h of infusion, the incidence of (re)MI was lower with hirudin versus heparin. The incidence of (re)MI increased somewhat during the first 4 and 8 h after hirudin therapy discontinuation compared with during hirudin infusion (p = 0.14). After heparin therapy discontinuation, the incidence of (re)MI increased by 5.8-fold during the first 4 h compared with during the infusion, an increase that remained significant at 8 h (p = 0.0001). There was a 0.5% absolute excess of (re)MI in the first 8 h after discontinuing heparin therapy compared with both the 8-h period before discontinuing heparin therapy and the 8-h period after stopping hirudin therapy.
Percutaneous coronary interventions
Similar proportions of both treatment groups underwent percutaneous intervention during study-drug infusion (8.5% for heparin; 9.4% for hirudin). Patients who underwent intervention while receiving study drug had higher rates of (re)MI than did patients who did not undergo intervention then. No patient in either treatment group who underwent angioplasty during study-drug infusion had (re)MI within 12 h after treatment with study drug was stopped; 2.1% and 1.4% of patients suffered (re)MI beyond 12 h after drug therapy was stopped in the hirudin and heparin groups, respectively.
Outcome and mortality
The hospital stay was significantly longer for patients who suffered (re)MI compared with those without (re)MI: 7 days for no (re)MI, 11 days for (re)MI within 12 h and 13 days for (re)MI after 12 h (p = 0.001). Patients with (re)MI had 10 times higher mortality at 30 days compared with patients without (re)MI. Their mortality remained significantly higher at 6 months and at 1 year (Table 3).
In the present study, we identified a clustering of (re)infarction after discontinuation of heparin infusion, with the greatest risk in the first 4 to 8 h. Of every 200 patients who had heparin therapy stopped after an acute coronary syndrome, one extra (re)MI occurred during the next 8 h relative to the risk during treatment, or longer after treatment. This early reactivation of thrombosis was not noted in the hirudin-treated patients, although the overall rate of (re)MI after stopping antithrombin treatment was similar between groups. There was a nonsignificant excess of (re)MI in the 5 days after discontinuation of hirudin therapy, but whether reactivation does not occur, is blunted or is delayed with hirudin compared with heparin is unclear.
Clinical and hematologic evidence for reactivation of thrombosis
Reactivation of ischemic events in unstable coronary syndromes has been shown after discontinuation of heparin therapy. In the Montreal study (7), there was recurrence of angina and MI in patients not given aspirin, at a median 9.5 h after heparin therapy was stopped. Similarly, in the GUSTO-I study (8), reinfarction was clustered within 10 h of heparin therapy discontinuation; the greatest risk was within 4 h. These patients had higher aPTT values, however, suggesting that inadequate anticoagulation was not the explanation. A persistent proaggregatory effect of heparin on platelets may contribute to a prothrombotic state (18). Granger et al. (9)reported that in a group of patients with unstable angina receiving 325 mg of aspirin daily, both thrombin activity (by fibrinopeptide A) and thrombin generation (by prothrombin fragment 1.2 [F1.2]) increased after heparin therapy was stopped and then returned to baseline at 24 h.
Direct antithrombin agents
The use of direct antithrombin agents, which can inactivate clot-bound thrombin, does not seem to prevent recurrence of angina after treatment withdrawal. Gold et al. (15)showed that in patients with acute coronary syndromes, the discontinuation of argatroban therapy was associated with early recurrence of angina, in most cases 4 to 8 h after cessation. There also was a four-fold increase in the concentration of thrombin-antithrombin III complex (TAT) at 2 h after infusion, but no rebound in levels of fibrinopeptide A.
In the ThRombin Inhibition in Myocardial ischemia (TRIM) study (11), the rates of death, MI and refractory angina were increased during the first 24 h after 72 h of heparin treatment, and with higher inogatran doses. This reactivation seemed most pronounced in the heparin group, was associated with increases in fibrin turnover and thrombin generation and continued after discontinuation.
The Organization to Assess Strategies for Ischemic Syndromes (OASIS) pilot hemostasis substudy (19)compared coagulation markers in patients treated with heparin versus hirudin. During the infusion, no difference was seen in the effect of heparin or hirudin on thrombin formation (F1.2), but hirudin seemed to be more effective in the suppression of markers of thrombin activity (TAT and d-dimer). After treatment with study drug was stopped, heparin-treated patients showed significant increases in F1.2 and TAT concentrations, higher at 6 h. At 24 h after stopping antithrombin therapy, the F 1.2 and TAT levels had increased in the hirudin group, consistent with a delayed rebound effect. These studies in aggregate show clinical and hematologic evidence of reactivation of thrombosis after treatment with either heparin and direct thrombin inhibitors is stopped.
Reactivation in GUSTO-IIb
In the GUSTO-IIb hemostasis substudy, there was no suppression effect on thrombin formation (F1.2) for either heparin or hirudin during study drug infusion. There was evidence of increased thrombin activity, however, albeit delayed compared with heparin, after discontinuation of hirudin therapy. These observations, and the results of our study, indicate that reactivation of thrombotic events appears early in patients with acute coronary syndromes when heparin therapy is discontinued. Several important questions have arisen as a consequence. One is whether early benefits of heparin could be better sustained with long-term antithrombotic treatment. In the FRagmin during InStability in Coronary artery disease (FRISC) study of dalteparin (20), trends toward reactivation occurred after the dose was reduced at 5 to 8 days and on cessation of therapy after 41 days. In FRISC-II (21,22), the initial benefit was sustained for at least 4 to 6 weeks with dalteparin twice daily among patients treated noninvasively, but there was some loss of benefit by 90 days. Although there is some evidence of reactivation of thrombosis on stopping treatment with low-molecular-weight heparin, upstream inhibition of the coagulation cascade, with either low-molecular-weight heparins (with higher anti-Xa to anti-IIa activity) or with factor Xa inhibitors, might result in less thrombin generation and therefore less reactivation of thrombosis, upon therapy discontinuation (23). The use of potent antiplatelet therapy also might prevent reactivation after stopping heparin therapy.
Another question is whether patients undergoing early percutaneous coronary intervention would have a lower risk of reMI after treatment with antithrombin agents is stopped, as suggested in FRISC-II, in which prolonged treatment with dalteparin showed no trend toward benefit for patients who had undergone revascularization. In GUSTO-IIb, likewise, no patient in either treatment group who had undergone angioplasty during study-drug infusion had (re)MI in the 12 h after treatment with study drug was stopped. The difference in reinfarction rates after study-drug therapy discontinuation between patients who did and did not undergo intervention was not significant, however, perhaps related to the relatively small number of patients who had undergone intervention while receiving study drug.
Whether gradual reductions in the dose of thrombin inhibitor would prevent or attenuate this response also is under question. Recently, Becker et al. (17)tested three strategies for heparin therapy discontinuation. In a pilot study, 30 patients with acute coronary syndromes treated with intravenous heparin for 48 h were randomly assigned to abrupt cessation, intravenous weaning over 12 h or subcutaneous weaning over 12 h of abbreviated intravenous weaning. Thrombin generation was evident within 1 h of cessation, increased progressively (nearly doubling by 24 h) and correlated inversely with the concentration of tissue-factor pathway inhibitor (TFPI). The increase was greatest among patients randomized to abrupt cessation and least in those with intravenous weaning. The attenuated response with intravenous weaning may relate to endothelial-cell recovery and restoration of TFPI-mediated thromboresistance.
The results of our study suggest that hirudin is associated with less early reactivation of thrombosis than heparin, perhaps reflecting a different mechanism of action on thrombin, a longer half-life, or both.
Patients at heightened risk for early reinfarction after stopping heparin therapy appear to be the same as those generally at higher risk for reinfarction: the elderly, women, diabetics and patients with prior angina. These high-risk features likely reflect both more severe underlying vascular disease and perhaps relative hypercoagulability.
A final question is whether measurement of markers of inflammation during the initial evaluation could be used to identify patients at risk of early thrombotic events and thus higher mortality.
This study has several limitations. The most important are those inherent to retrospective studies; these findings thus should be considered hypothesis-generating rather than definitive. In addition, heparin use immediately after study-drug discontinuation accounted for a large proportion of the patients excluded from the original population (22%). The median age (64 vs. 65 years) and proportions of patients who were female (31% vs. 30%) or diabetic (17% vs. 19%) or had prior MI (27% vs. 26%) were similar between the hirudin and heparin groups, suggesting that this exclusion did not introduce systematic bias. Another issue is that we considered (re)infarction as the only indicator of reactivation; we did not analyze the occurrence of recurrent ischemia. Because recurrent ischemia was less precisely defined, and only the first episode was collected, our ability to define timing accurately was limited.
Although reactivation of ischemic events shortly after discontinuation of heparin or other antithrombin therapy has been recognized, to our knowledge, this study is the largest to address this important issue in patients with acute coronary syndromes treated with concomitant aspirin. There was no overall difference in reinfarction after stopping heparin therapy versus hirudin. Our findings confirm, however, that clustering of (re)MI occurs early after heparin therapy is stopped and suggest that this reactivation is associated with a particularly poor prognosis. Prevention of thrombotic reactivation may be an important target for future antithrombotic strategies for patients with acute coronary syndromes.
We thank Patricia French for her editorial assistance.
☆ This study was funded in part by grants from Guidant Corporation, Mountain View, California, and the Novartis Corporation, Summit, New Jersey.
- activated partial thromboplastin time
- creatine kinase (-MB)
- prothrombin fragment 1.2
- FRagmin during InStability in Coronary artery disease
- Global Utilization of Streptokinase and TPA (alteplase) for Occluded coronary arteries
- Global Use of Strategies To Open occluded arteries in acute coronary syndromes
- Organization to Assess Strategies for Ischemic Syndromes
- myocardial (re)infarction
- thrombin-antithrombin III complex
- tissue-factor pathway inhibitor
- ThRombin Inhibition in Myocardial ischemia
- Received July 10, 2000.
- Revision received November 9, 2000.
- Accepted December 13, 2000.
- American College of Cardiology
- Braunwald E.,
- Mark D.B.,
- Jones R.H.,
- et al.
- Braunwald E.,
- Jones R.H.,
- Mark D.B.,
- et al.
- Antman E.M.
- GUSTO-I Investigators,
- Granger C.B.,
- Hirsh J.,
- Califf R.M.,
- et al.
- Granger C.B.,
- Miller J.,
- Bovill E.G.,
- et al.
- GUSTO-I Hemostasis Substudy Group,
- Granger C.B.,
- Becker R.,
- Tracy R.,
- et al.
- TRIM Study Group
- Linder R.,
- Olgren J.,
- Egberg N.,
- Grip L.,
- Larson G.,
- Siegbahn A.
- Gold H.K.,
- Torres F.W.,
- Garabedian H.D.,
- et al.
- Merlini P.A.,
- Ardissino D.,
- Bauer K.A.,
- et al.
- Becker R.C.,
- Spencer F.A.,
- Li Y.,
- et al.
- OASIS Pilot Study Group,
- Flather M.,
- Weitz J.,
- Yusuf S.,
- et al.