Author + information
- Received July 10, 1997
- Revision received November 12, 1997
- Accepted December 22, 1997
- Published online March 15, 1998.
- P.Gabriel Steg, MD, FESC, FACCA,* (, )
- Thierry Laperche, MDA,
- Jean-Louis Golmard, MD, PhDB,
- Jean-Michel Juliard, MDA,
- Hakim Benamer, MDA,
- Dominique Himbert, MDA,
- Pierre Aubry, MDA,
- for the PERM Study Group1
- ↵*Dr. P. Gabriel Steg, Cardiologie, Hôpital Bichat, 46 rue Henri Huchard, 75877 Paris Cedex 18, France.
Objectives. We sought to examine the relation between time to treatment and 90-min patency rates in patients receiving intravenous streptokinase (SK) or accelerated tissue-type plasminogen activator (t-PA).
Background. Early patency of the infarct-related artery is a major determinant of survival after thrombolysis for acute myocardial infarction. Some data suggest that time to treatment may influence the efficacy of nonfibrin-specific thrombolytic agents in restoring early patency of the infarct-related vessel.
Methods. We performed a retrospective analysis of a cohort of 481 patients receiving thrombolytic therapy for acute myocardial infarction <6 h after pain onset, all of whom underwent 90-min coronary angiography. Patency of the infarct-related artery was graded by two observers who had no knowledge of the treatment received or the time between pain and therapy.
Results. There was no difference in baseline clinical or angiographic characteristics according to the timing or nature of treatment. Thrombolysis in Myocardial Infarction (TIMI) flow grade 2 or 3 patency rate after SK correlated negatively with the time between onset of pain and thrombolysis (r = 0.8, p = 0.05), whereas the 90-min patency rate after t-PA appeared stable as a function of time to treatment. When patients were categorized as having received treatment <3 or ≥3 h after pain onset, the patency rate was similar with t-PA, but significantly higher when SK was administered early rather than late, regardless of whether TIMI flow grades 2 and 3 were pooled (86.9% vs. 59.4%, p = 0.0001) or TIMI flow grade 3 alone was considered to indicate patency (81.7% vs. 53.6%, p = 0.0001). Multivariate logistic regression analysis showed a negative effect of time to treatment on the patency probability for SK (p = 0.0001) but not for t-PA.
Conclusions. The efficacy of streptokinase but not t-PA in restoring early coronary patency after intravenous thrombolysis is markedly lower when patients are treated later after onset of pain.
Early (60 to 90 min) and complete (i.e., Thrombolysis in Myocardial Infarction [TIMI] flow grade 3) patency of the infarct-related artery is a major determinant of survival after thrombolysis for acute myocardial infarction [1–3]. Several studies have suggested that, at least with nonfibrin-specific thrombolytic agents, time to treatment may influence the patency rates achieved at 90 min [4–6]. We examined differences between thrombolysis with streptokinase (SK) and tissue-type plasminogen activator (t-PA) with regard to 90-min patency, according to the timing of treatment after pain onset.
This retrospective analysis was based on data from two recent prospective French studies of coronary patency after thrombolysis [7, 8], involving two cohorts of consecutive patients receiving intravenous thrombolysis for acute myocardial infarction <6 h after symptom onset and who underwent routine 90-min coronary angiography, regardless of clinical and electrocardiographic signs of reperfusion. Patients were enrolled if they met the following criteria: nitrate-resistant chest pain of ≤6 h but >30 min duration; ST segment elevation or giant T waves or left bundle branch block; and no contraindication to thrombolysis. All the patients received at least 250 mg of aspirin followed by either SK (1.5 million U intravenously over 60 min) or accelerated t-PA (15-mg bolus followed by 0.75 mg/kg body weight in 30 min [maximum 50 mg] and 0.50 mg/kg over 60 min [maximum 35 mg]). The choice of the agent was left to the discretion of the attending physician. Seventy-two additional patients who had received other thrombolytic regimens were excluded from the analysis. All the patients underwent 90-min coronary angiography, with a bolus of at least 5,000 IU heparin intravenously at the beginning of the procedure. All angiograms were recorded on 35-mm cine film and were analyzed off-line by two independent observers from a group of five experienced angiographers who had no knowledge of the treatment received nor the time between pain onset and therapy. The infarct-related artery could not be ascertained in eight patients, so these patients were excluded from the analysis. In all other cases, flow in the infarct-related vessel was graded according to the TIMI score, with discrepancies being resolved by consensus. Patency was graded after at least four injections of dye through the infarct-related vessel and, in patients with systolic blood pressure ≥100 mm Hg, after intracoronary injection of nitrates. Multivessel disease was defined as a lumen diameter reduction of at least 50% in at least two major epicardial arteries.
1.1 Statistical Analysis
All continuous variables are presented as the mean value ± SD. Comparisons of baseline characteristics were based on the Student ttest for continuous variables and the chi-square test for categoric variables. A p value <0.05 was used to denote statistical significance. To investigate the relation between early patency, treatment and time to thrombolysis, multivariate analysis was performed using the SAS statistical package (SAS Institute). This analysis was based on the logistic model and involved two steps: in the first step, there was a model expressing the probability of early patency as a logistic function of three potential factors—namely, the thrombolytic agent used, the time to thrombolysis and an interaction term. The values of the variables for each factor were compared with zero using the Wald test [9, 10]. In the second step, a final model was derived from the initial model by discarding the factors that were not linked to patency.
Four hundred eighty-one patients were analyzed, 237 of whom received t-PA and 244 of whom received SK. Their baseline characteristics were similar (Table 1). Coronary angiography was performed 90 min after the start of thrombolysis in every patient and found single-vessel disease in 53.8% and 56.2% of patients receiving t-PA and SK, respectively (p = NS). In-hospital deaths occurred in 5.4% and 2.3% of patients receiving t-PA and SK, respectively (p = NS). TIMI flow grade 2 or 3 patency was achieved in 72.6% and 77.4% of patients treated with t-PA and SK, respectively (p = NS). The TIMI flow grade 2 or 3 patency rate after SK correlated negatively with the time between the onset of pain and the beginning of thrombolysis (r = 0.8, p = 0.05), whereas the 90-min patency rate after t-PA was stable (Fig. 1).
Fig. 1suggests that the effect of time on probability of patency is different for the two thrombolytic agents. This hypothesis was confirmed by multivariate logistic regression analysis. The variables estimated from the initial model and the p values derived from the Wald test are displayed in Table 2A. This table shows a significant effect of the type of treatment (SK or t-PA) and of the treatment–time interaction, but the effect of time was not significant. Therefore, this last variable was discarded from the final model displayed in Table 2B: for patients receiving SK (coded 1), the probability of patency is a logistic function of a term expressed as: (intercept variable + treatment variable + interaction variable × time to treatment). Because the latter is significantly lower than zero, this means that for these patients, the probability of patency decreases when time to treatment increases. In contrast, for patients receiving t-PA, the probability of patency is only a function of the intercept variable (because t-PA was coded 0, it nullifies the treatment and the interaction terms). Therefore, at least within 6 h of pain onset, the effect of time to treatment on the probability of patency can be considered as constant for t-PA. This multivariate analysis gave similar results regardless of whether patency was defined as TIMI flow grade 2 and 3 (Table 2) or as TIMI flow grade 3 alone (data not shown).
When patients receiving t-PA were categorized as having received treatment <3 or ≥3 h after pain onset, the patency rate was similar regardless of whether TIMI flow grades 2 and 3 were pooled (71.6% vs. 71.9%, p = NS) or whether TIMI flow grade 3 alone was considered to indicate patency (61.9% vs. 57.3%, p = NS). In contrast, the efficacy of SK was significantly higher among patients treated <3 versus ≥3 h after pain onset (86.9% vs. 59.4% for pooled TIMI flow grades 2 and 3, p = 0.0001 and 81.7% vs. 53.6% for TIMI flow grade 3 alone, p = 0.0001) (Fig. 2). No baseline characteristics other than time to treatment were significantly associated with 90-min patency. Among patients receiving t-PA, there was no difference in the average time to treatment between patients with patent (TIMI flow grade 2 or 3) infarct-related arteries and patients with occluded infarct-related arteries at 90 min (172 ± 79 vs. 169 ± 79 min, respectively, p = NS). In contrast, among patients receiving SK, those who had a patent infarct-related artery at 90 min had been treated, on average, earlier than patients with occluded arteries (163 ± 54 vs. 218 ± 90 min, respectively, p = 0.0001).
The results of this analysis suggest that the efficacy of SK but not t-PA in restoring early coronary patency after intravenous thrombolysis is markedly reduced when patients are treated >3 h after pain onset. There were no differences in baseline characteristics that could have accounted for the different behavior of t-PA and SK. This confirms a preliminary observation in the TIMI-1 study . A similar negative correlation between the time from symptom onset to treatment and the patency rate has been reported with anistreplase [5, 6]and urokinase , suggesting that this phenomenon is common to “nonfibrin-specific” thrombolytic agents. In the Thrombolytic trial of Eminase in Acute Myocardial infarction (TEAM-1) study , the 90-min reperfusion rate dropped from 60% if anistreplase was given <4 h after onset to 33% if it was given later after 4 h. These observations are consistent with the notion that more recent clots may be more susceptible to thrombolysis with SK and related fibrinolytic agents, possibly owing to lesser fibrin cross-linking. In contrast, older clots may be more resistant to thrombolytic agents through continuing fibrin polymerization and plasminogen activator inhibitor release by platelets. In contrast, t-PA appears to lyse fresh and older clots equally efficiently. This is consistent with in vitro observations and may be related to the much higher fibrin affinity with t-PA than with other plasminogen activators, as well as to the differences in the spatial distribution of fibrinolytic components during thrombolysis. Using confocal microscopy, Sakharov et al. have shown that the depth of the prelysis zone of a plasma clot is dependent on the fibrin-binding properties of the plasminogen activator used and corresponds to the depth of penetration of the activator into the clot. Given its high affinity for fibrin, t-PA penetrates less deeply than nonfibrin-specific agents and remains confined to a very sharp and thin front of lysis . In the case of older, retracted thrombi, these differences in spatial penetration by plasminogen activators may translate into substantial differences in infarct-related artery patency. This indicates that the time between pain onset and therapy should be taken into account when comparing patency rates between thrombolytic regimens.
This finding has several potential implications. First, because t-PA has no clear superiority over SK in terms of survival when patients are treated >6 h after pain onset , this would tend to confirm the hypothesis that SK may be beneficial in ways other than restoring early patency (e.g., decreased blood viscosity). Second, the strong relation between time to treatment and survival after thrombolysis may not be related solely to earlier and therefore more effective myocardial salvage when patency is reestablished, but may also be related, at least in the case of SK, to higher patency rates when treatment is given early.
3.1 Study Limitations
The main limitation of this study is that it is based on a retrospective analysis of data accumulated for other purposes (validation of noninvasive markers of reperfusion and patency studies). However, there was no bias in recruitment toward one treatment or the other and there was no bias in the selection of patients for angiography, as all patients who entered both studies underwent 90-min angiography. Angiographic patency has been shown to be poorly reproducible, with a notable overlap between patients with TIMI flow grades 2 and 3 . However, separation of TIMI flow grades 0 to 1 and 2 to 3 is more reliable, showing excellent interobserver agreement in the analysis by Gibson et al. , with a kappa value of 0.84 ± 0.05. In addition, the difference between the patency rates obtained when SK was given <3 and >3 h after symptom onset held, regardless of the criterion used to define angiographic patency. Finally, our blinded grading of angiograms guarantees unbiased analysis.
In acute myocardial infarction, the time between the onset of pain and the beginning of thrombolytic therapy strongly influences the patency rates obtained with SK but not with accelerated t-PA.
We are grateful to D. D. Young for help with preparation of the manuscript and to J. Ph. Collet and T. Lecompte for helpful discussions.
- Prospective Evaluation of Reperfusion Markers study
- Thrombolysis in Myocardial Infarction trial
- tissue-type plasminogen activator
- Received July 10, 1997.
- Revision received November 12, 1997.
- Accepted December 22, 1997.
- The American College of Cardiology
- Stadius ML
- Kennedy JW
- Chesebro JH,
- Knatterud G,
- Roberts R,
- et al.
- ↵Anderson JL, Rothbard RL, Hackworthy RA, et al., for the APSAC Multicenter Investigators. Multicenter reperfusion trial of intravenous anisoylated plasminogen streptokinase activator complex (APSAC) in acute myocardial infarction: controlled comparison with intracoronary streptokinase. J Am Coll Cardiol 1988;11:1153–63.
- Leizorovicz A,
- Boissel JP,
- Robert F
- ↵Laperche T, Steg PG, Dehoux M, et al., for the PERM Study Group. A study of biochemical markers of reperfusion early after thrombolysis for acute myocardial infarction. Circulation 1995;92:2079–86.
- Juliard JM,
- Himbert D,
- Golmard JL,
- et al.
- Cox DR,
- Kinkley DV
- Agresti A
- ↵Zeymer U, Tebbe U, von Essen R, Vogt A, Neuhaus KL, for the ALLK Study Group. Influence of time to treatment on early infarct-related artery patency after different thrombolytic regimens [abstract]. Circulation 1996;94 Suppl I:I-440.
- Collen D,
- Stassen JM,
- Verstraete M
- Sakharov AV,
- Nagelkerke F,
- Rijken DC
- Gibson CM,
- Cannon CP,
- Daley WL,
- et al.