|
|
| |||||||||||||||||||||||||||||||||||||||||
Background In 1995, the two-part National Institute of Neurological Disorders and Stroke (NINDS) Recombinant Tissue Plasminogen Activator Stroke Study found that patients who were treated with tissue plasminogen activator (t-PA) within three hours after the onset of symptoms of acute ischemic stroke were at least 30 percent more likely than patients given placebo to have minimal or no disability three months after the stroke. It was unknown, however, whether the benefit would be sustained for longer periods.
Methods In the NINDS trial, a total of 624 patients with stroke were randomly assigned to receive either t-PA or placebo. We collected outcome data over a period of 12 months after the occurrence of stroke. The primary outcome measure was a "favorable outcome," defined as minimal or no disability as measured by the Barthel index, the modified Rankin Scale, and the Glasgow Outcome Scale. We assessed the treatment effect using a global statistic.
Results Using an intention-to-treat analysis for the combined results of the two parts of the trial at 6 months and 12 months, we found that the global statistic favored the t-PA group (odds ratio for a favorable outcome at 6 months, 1.7; 95 percent confidence interval, 1.3 to 2.3; odds ratio at 12 months, 1.7;95 percent confidence interval, 1.2 to 2.3). The patients treated with t-PA were at least 30 percent more likely to have minimal or no disability at 12 months than were the placebo-treated patients (absolute increase in the proportion with a favorable outcome, 11 to 13 percentage points). There was no significant difference in mortality at 12 months between the t-PA group and the placebo group (24 percent vs. 28 percent, P=0.29). There was no interaction between the type of stroke identified at base line and treatment with respect to the long-term response. The rate of recurrent stroke at 12 months was similar in the two groups.
Conclusions During 12 months of follow-up, the patients with acute ischemic stroke who were treated with t-PA within three hours after the onset of symptoms were more likely to have minimal or no disability than the patients given placebo. These results indicate a sustained benefit of t-PA for such patients.
Methods
Patients with acute ischemic stroke who could be treated within three hours after the onset of symptoms and who had a measurable neurologic deficit according to the National Institutes of Health Stroke Scale (NIHSS)2 were eligible for the study. Randomization was stratified according to the time since onset of the stroke (about half the patients were enrolled within 90 minutes after the onset of symptoms of stroke and the remainder between 90 and 180 minutes thereafter) and the clinical center. The patients were followed for 12 months. Outcome data were systematically collected 24 hours and 3, 6, and 12 months after stroke. The stroke was classified as large-vessel occlusive, small-vessel occlusive, or cardioembolic solely on the basis of the clinical and diagnostic information available to the investigator at the time of randomization. The protocols for each part of the trial were approved by the human research committee at each site. Informed consent was obtained from all the patients.
Part 1 of the trial included 291 patients and was designed to test whether a greater proportion of patients treated with t-PA, as compared with those who received placebo, had early improvement, defined as complete resolution of the neurologic deficit or an improvement from base line of four or more points in the score on the NIHSS 24 hours after the onset of stroke. The primary hypothesis for part 2 of the trial, which included 333 patients, was that there would be a consistent and persuasive difference between the t-PA and placebo groups in terms of the proportion of patients with minimal or no neurologic deficit three months after stroke. Except for the difference in the primary hypotheses, the protocols for parts 1 and 2 were the same. The four outcome measures used in both parts of the study were the Barthel index, the modified Rankin scale, the Glasgow Outcome Scale, and the NIHSS. The NIHSS was not used in this long-term follow-up study, because the necessary data were not collected at 6 months and 12 months.
During the first 24 hours after randomization, patients were monitored closely in an intensive care setting. Blood pressure was managed according to an algorithm, and antithrombotic and antiplatelet agents were not allowed by the protocol.3 After this initial period, treatment was directed by the patient's physician and not dictated by the protocol.
Clinical investigators remained unaware of the results of part 1 until part 2 was completed and the data were analyzed. Although the outcome data at 6 and 12 months were prospectively collected in accordance with the protocol, these outcomes were not specified as secondary outcomes in the protocol. Thus the analyses presented here are considered post hoc and exploratory.
Certified nurse coordinators or study physicians telephoned the patients or their care givers to determine the vital status of the patients; their ability to perform daily activities (measured with the Barthel index)4; and the degree of functional disability (measured with the modified Rankin scale5 and the Glasgow Outcome Scale6). The evaluators were unaware of the treatment assignments when the 6- and 12-month outcomes were assessed, and the patients and their care givers were also unaware of the treatment assignments. Several studies have validated telephone assessment of the outcome of stroke.7,8,9 Data were also collected on the causes of death and serious medical events, including intracerebral hemorrhage and recurrent stroke.
A favorable outcome was defined as minimal or no disability, as measured by scores of 95 or 100 on the Barthel index (range of scores, 0 to 100), 0 or 1 on the modified Rankin scale (range of scores, 0 to 5), and 1 on the Glasgow Outcome Scale (range of scores, 1 to 5). Using an intention-to-treat analysis, we assigned the patients who died before the specified follow-up the most unfavorable scores for each of the outcome measures. Patients for whom data were missing on an outcome measure were given the most unfavorable score for that measure. To increase the ability of the study to detect differences in long-term outcomes, we combined data from parts 1 and 2, since the methods for both parts, including those used for recruitment, treatment, and follow-up, were identical.
In this study, as in the original study of this cohort, favorable outcomes were determined with the use of a global statistic (the Wald test) derived from a general linear model with logit-link function, computed with the use of generalized estimating equations.10 The global test incorporated simultaneously the proportion of patients with minimal or no disability on the Barthel index, the modified Rankin scale, and the Glasgow Outcome Scale at 6 and 12 months. If the global tests indicated a significant difference at the 0.05 level, a Mantel–Haenszel test was used to compare the scores of the patients treated with t-PA and those of the patients who received placebo for each of the three scales.11
To describe the treatment effects across the range of outcomes, outcomes were classified into one of four categories: minimal or no disability (score on Barthel index, 95 or 100; score on modified Rankin scale, 0 or 1; and score on Glasgow Outcome Scale, 1), moderate disability (Barthel index, 55 to 90; modified Rankin scale, 2 or 3; and Glasgow Outcome Scale, 2), severe disability (Barthel index, 0 to 50; modified Rankin scale, 4 or 5; and Glasgow Outcome Scale, 3 or 4), and death.
Log-rank tests were conducted to compare survival in the two groups 6 months and 12 months after the onset of stroke. The incidence of recurrent stroke and intracerebral hemorrhage was reported according to treatment group and time after randomization (
3 months, >3 to 6 months, and >6 to 12 months).
The cause of death was determined independently by a panel of three investigators who were unaware of the patients' treatment assignments. The patients' files and pertinent computed tomographic (CT) scans were reviewed. The causes of death were classified into one of the following five categories: cerebrovascular cause, in which death was due to intracerebral hemorrhage, severe infarction, or extensive cerebral edema; cardiovascular cause, which included congestive heart failure and pulmonary edema, acute myocardial infarction, ventricular arrhythmia, and pulmonary embolism; infection, primarily aspiration pneumonia and sepsis; cancer; and other, if the cause of death could not be classified into one of the preceding categories. All discrepancies regarding classification among the investigators were adjudicated by the panel until a consensus was reached.
After the cause of death was determined, its relation to the index stroke was assessed according to whether it was definitely related, possibly related, or not related. Although all base-line CT scans had to show no evidence of intracerebral hemorrhage for a patient to be included in this study, subsequent CT scans could show intracerebral hemorrhage or infarction without evidence of hemorrhage. For all causes of death that were judged to be definitely related to the index stroke, the underlying pathophysiologic feature of the stroke deemed to have been related to the patient's death was categorized as intracerebral hemorrhage, cerebral edema, or cerebral infarction. To allow comparisons between deaths due to severe or extensive ischemic infarction and those due to intracerebral hemorrhage, the categories of edema and infarction were combined. Deaths thought to be due to the severity of the index stroke could then be compared with deaths judged to be due to intracerebral hemorrhage.
Multivariable analyses were carried out to test the association of selected base-line variables, including the type of stroke, and favorable outcome at one year. Variables for inclusion in the models were selected with the use of procedures described previously.12 A multivariable model was also fitted to identify base-line variables associated with survival at one year with the use of Cox proportional-hazards regression analysis. An interaction with a P value of less than 0.1 was considered to indicate significance. Variables with a P value of less than 0.05 or interactions between variables with a P value of less than 0.1 were retained in the final model. The effect of treatment on outcome at 12 months was tested again, with adjustments for base-line covariates and interactions remaining in the final model.
Results
A total of 624 patients were randomly assigned to treatment. Only 15 of the surviving patients (2.4 percent of the original 624; 7 in the t-PA group and 8 in the placebo group) were not available for the 6-month follow-up assessment for all three outcome scales, and 26 (4.2 percent; 14 in the t-PA group and 12 in the placebo group) were unavailable for the follow-up assessment at 12 months for all three scales. According to the intention-to-treat analysis, these patients were considered to have an unfavorable outcome for each missing scale.
Like the data at 3 months, the 6-month and 12-month data showed that the patients treated with t-PA were more likely to have a favorable outcome than those who received placebo. The odds ratio for a favorable outcome at 6 months in the t-PA group as compared with the placebo group was 1.7 (95 percent confidence interval, 1.3 to 2.3) and at 12 months was 1.7 (95 percent confidence interval, 1.2 to 2.3) (Table 1). A similar pattern is evident in the results of the univariate tests for the Barthel index, the modified Rankin scale, and the Glasgow Outcome Scale (Table 1). At 12 months, the range of the absolute increase in the proportion of patients with a favorable outcome was 11 to 13 percentage points, and the range of the relative increase in favorable outcome was 32 to 46 percent for the three outcome scales.
|
|
Thirty-four of the 624 patients had symptomatic recurrent strokes by 12 months, including 2 patients with two recurrences. Twenty-four events occurred within 3 months after treatment (12 in the t-PA group, and 12 in the placebo group), 7 between 3 and 6 months (4 in the t-PA group, and 3 in the placebo group), and 3 between 6 months and 12 months (1 in the t-PA group and 2 in the placebo group). Of the 34 recurrent strokes, 2 (6 percent) were classified as small-vessel disease, 16 (47 percent) as cardioembolic, 9 (26 percent) as large-vessel disease, and 7 (21 percent) as of unknown cause. There was no significant difference in the incidence of recurrent stroke between patients in the t-PA group and those in the placebo group (P=0.89 at 6 months, and P=0.96 at 12 months).
The mortality rate 6 months after stroke was not significantly lower in the t-PA group than in the placebo group (21 percent vs. 23 percent, P=0.31) or at 12 months (24 percent vs. 28 percent, P=0.29) (Figure 2).
|
|
|
We previously reported that treatment with intravenous t-PA within three hours after the onset of symptoms of an acute ischemic stroke improved the clinical outcome at three months.1 We now report that the magnitude of this benefit was sustained at 6 months and 12 months. In addition, patients treated with t-PA were less likely to be severely disabled 12 months after the stroke. These findings indicate that the beneficial effect of t-PA is sustained and that the primary benefit of t-PA is seen early after treatment.
The rates of recurrent stroke were similar in the t-PA group and the placebo group during the 12 months of follow-up and were similar to those found in previous population-based studies of stroke.13,14,15 Although our patients were carefully selected, the similarity in the rates of recurrent stroke suggests that they may be representative of patients with ischemic stroke. The mortality rates were also similar between the two groups at 6 months (21 percent in the t-PA group and 23 percent in the placebo group) and at 12 months (24 percent and 28 percent, respectively).
In our earlier study we found that despite the increased risk of intracerebral hemorrhage in patients treated with t-PA, there was no difference in overall mortality between those receiving t-PA and those receiving placebo. Our current results suggest that the explanation for this finding is that patients who die from causes related to the index stroke are likely to die regardless of whether intracerebral hemorrhage occurs. This explanation reflects the often fatal consequences of severe ischemic stroke even in the absence of intracerebral hemorrhage.16
The results of multivariable analyses of variables related to a favorable outcome at 12 months were similar to those reported at 3 months.12 The absence of diabetes and the interaction between younger age and lower base-line NIHSS score remained in both models. In addition, younger age and a lower NIHSS score at base line were independently associated with a favorable outcome at 12 months. The base-line variables of younger age, lower NIHSS score, absence of diabetes, and the interaction of the absence of diabetes and younger age were also predictors of survival at 12 months. However, none of these variables associated with a favorable outcome or survival interacted with treatment. Therefore, even though patients who are older, have diabetes, or have higher base-line NIHSS scores are less likely to have a favorable outcome or survive 12 months after a stroke, they should not be selected for treatment solely on the basis of these features.12 For example, it would be fallacious to select a single subgroup with a high NIHSS score from those listed in Table 3 and draw the conclusion that t-PA is not beneficial for that subgroup.17 The broader trend of benefit with t-PA treatment that was demonstrable among all subgroups makes a single isolated aberration within one category of a subgroup more likely to be a random occurrence related to the small size of the sample than a clinically meaningful biologic trend.
In addition, after adjusting for the other base-line variables, we could not detect an association between the type of stroke identified at base line and the outcome over the long term. The type of stroke at base line may not correspond to the type of stroke that is ultimately identified, as determined by a complete diagnostic workup. However, the constraints of a very narrow therapeutic window mandate that a determination of the type of stroke be made at the time of clinical presentation. With these limitations in mind, our findings support the conclusion that the selection of patients for treatment with t-PA cannot be based solely on the mechanism of stroke, as determined at base line on the basis of clinical impression.
Other large studies of thrombolysis for the treatment of acute ischemic stroke have shown either no benefit18,19,20 or equivocal benefit21 from intravenous streptokinase or t-PA and high rates of early mortality and intracerebral hemorrhage. Reasons for the disparity between the results of these studies and ours may include differences in the time from the onset of symptoms to the initiation of treatment (zero to three hours in our study as compared with zero to four or six hours in the other trials), the choice and dose of thrombolytic agent, and the concomitant use or nonuse of heparin or aspirin.22,23,24 In addition, we used a strict algorithm for the management of blood pressure after beginning treatment with t-PA. A more recent trial of intravenous t-PA given within six hours after the onset of symptoms, but whose protocol was otherwise similar to ours, failed to show a benefit for t-PA in the primary analysis.25 This finding further supports the recommendation that treatment of acute ischemic stroke with intravenous t-PA be limited to patients who can be treated within three hours after the onset of stroke.
The results of this long-term follow-up study have potential implications for public health. For example, for every 100 patients treated with t-PA according to the selection criteria and management guidelines of this study, at least 11 additional patients will have a favorable outcome during the subsequent year. In addition, a cost-effectiveness analysis has demonstrated net cost savings to the health care system and improved quality of life for patients treated with t-PA.26
In conclusion, patients who received t-PA within three hours after the onset of symptoms were more likely to have minimal or no disability 3, 6, and 12 months later. Similar results may be attainable in everyday practice if guidelines for treatment and management are followed closely.24,27
Supported by contracts (N01-NS-02382, N01-NS-02374, N01-NS-02377, N01-NS-02381, N01-NS-02379, N01-NS-02373, N01-NS-02376, N01-NS-02378, and N01-NS-02380) with the National Institute of Neurological Disorders and Stroke.
Drs. Kwiatkowski, Libman, Frankel, Morgenstern, Broderick, Lewandowski, and Levine have been members of a speakers' bureau sponsored by Genentech. Dr. Morgenstern has also served on a Genentech advisory panel.
Source Information
From the Departments of Emergency Medicine (T.G.K.) and Neurology (R.B.L.), Long Island Jewish Medical Center, New Hyde Park, N.Y.; the Department of Neurology, Emory University School of Medicine, Atlanta (M.F.); the Department of Biometry and Epidemiology, Medical University of South Carolina, Charleston (B.C.T.); the Department of Neurology, University of Texas Medical School at Houston, Houston (L.B.M.); the Departments of Biostatistics and Research Epidemiology (M.L.) and Emergency Medicine (C.A.L.), Henry Ford Health Sciences Center, Detroit; the Department of Neurology, University of Cincinnati Medical Center, Cincinnati (J.P.B.); the Division of Stroke and Trauma, National Institute of Neurological Disorders and Stroke, Bethesda, Md. (J.R.M.); the Department of Neurology, Wayne State University School of Medicine, Detroit (S.R.L.); and the Department of Neurology, Mayo Clinic, Jacksonville, Fla. (T.B.).
Address reprint requests to Dr. Kwiatkowski at the Department of Emergency Medicine, Long Island Jewish Medical Center, 270-05 76th Ave., New Hyde Park, NY 11040, or at kwiatkow{at}lij.edu.
References
The following persons and institutions participated in the National Institute of Neurological Disorders and Stroke Recombinant Tissue Plasminogen Activator Stroke Study Group: Clinical Centers — University of Cincinnati (150 patients): T. Brott, J. Broderick, R. Kothari, M. O'Donoghue, W. Barsan, T. Tomsick, J. Spilker, R. Miller, and L. Sauerbeck; Affiliated sites: St. Elizabeth Hospital (South), J. Farrell, J. Kelly, T. Perkins, and R. Miller; University Hospital, T. McDonald; Bethesda North Hospital, M. Rorick and C. Hickey; St. Luke Hospital (East), J. Armitage and C. Perry; Providence Hospital, K. Thalinger and R. Rhude; Christ Hospital, J. Armitage and J. Schill; St. Luke Hospital (West), P.S. Becker, R.S. Heath, and D. Adams; Good Samaritan Hospital, R. Reed and M. Klei; St. Francis/St. George Hospital, A. Hughes and R. Rhude; Bethesda Oak Hospital, J. Anthony and D. Baudendistel; St. Elizabeth Hospital (North), C. Zadicoff and R. Miller; St. Luke Hospital–Kansas City, M. Rymer, I. Bettinger, and P. Laubinger; Jewish Hospital, M. Schmerler and G. Meirose; University of California, San Diego (146 patients): P. Lyden, K. Rapp, T. Babcock, P. Daum, D. Persona, M. Brody, C. Jackson, S. Lewis, J. Liss, Z. Mahdavi, J. Rothrock, T. Tom, R. Zweifler, J. Dunford, and J. Zivin; Affiliated sites: Sharp Memorial Hospital, R. Kobayashi, J. Kunin, J. Licht, R. Rowen, and D. Stein; Mercy Hospital, J. Grisolia and F. Martin; Scripps Memorial Hospital, E. Chaplin, N. Kaplitz, J. Nelson, A. Neuren, and D. Silver; Tri-City Medical Center, T. Chippendale, E. Diamond, M. Lobatz, D. Murphy, D. Rosenberg, T. Ruel, M. Sadoff, J. Schim, and J. Schleimer; Mercy General Hospital, Sacramento, R. Atkinson, D. Wentworth, R. Cummings, R. Frink, and P. Heublein; University of Texas Medical School, Houston (104 patients): J. Grotta, T. DeGraba, M. Fisher, A. Ramirez, S. Hanson, L. Morgenstern, C. Sills, W. Pasteur, F. Yatsu, K. Andrews, C. Villar-Cordova, P. Pepe, P. Bratina, L. Greenberg, S. Rozek, and K. Simmons; Affiliated sites: Houston Fire Department Emergency Medical Services; Hermann Hospital, St. Luke's Episcopal Hospital, Lyndon Baines Johnson General Hospital, Memorial Northwest Hospital, Memorial Southwest Hospital, Heights Hospital, Park Plaza Hospital, and Twelve Oaks Hospital; Long Island Jewish Medical Center (72 patients): T. Kwiatkowski, S. Horowitz, R. Libman, R. Kanner, R. Silverman, J. LaMantia, C. Mealie, R. Duarte, R. Donnarumma, M. Okola, V. Cullen, and E. Mitchell; Henry Ford Hospital (62 patients): S. Levine, C. Lewandowski, G. Tokarski, N. Ramadan, P. Mitsias, M. Gorman, B. Zarowitz, J. Kokkinos, J. Dayno, P. Verro, C. Gymnopoulos, R. Dafer, L. D'Olhaberriague, K. Sawaya, S. Daley, and M. Mitchell; Emory University School of Medicine (39 patients): M. Frankel, B. Mackay, C. Barch, J. Braimah, B. Flaherty, J. MacDonald, S. Sailor, A. Cook, H. Karp, B. Nguyen, J. Washington, J. Weissman, M. Williams, and T. Williamson; Affiliated sites: Grady Memorial Hospital, Crawford Long Hospital, Emory University Hospital, South Fulton Hospital, M. Kozinn and L. Hellwick; University of Virginia Health Sciences Center (37 patients): E. Haley, Jr., T. Bleck, W. Cail, G. Lindbeck, M. Granner, S. Wolf, M. Gwynn, R. Mettetal, Jr., C. Chang, N. Solenski, D. Brock, G. Ford, G. Kongable, K. Parks, S. Wilkinson, and M. Davis; Affiliated site: Winchester Medical Center, G. Sheppard, D. Zontine, K. Gustin, N. Crowe, and S. Massey; University of Tennessee (14 patients): M. Meyer, K. Gaines, A. Payne, C. Bales, J. Malcolm, R. Barlow, and M. Wilson; Affiliated sites: Baptist Memorial Hospital, C. Cape; Methodist Hospital Central, T. Bertorini; Jackson Madison County General Hospital, K. Misulis; University of Tennessee Medical Center, W. Paulsen and D. Shepard; Coordinating Center — Henry Ford Health Sciences Center: B. Tilley, K. Welch, S. Fagan, M. Lu, S. Patel, S. Li, J. Verter, J. Boura, J. Main, L. Gordon, N. Maddy, and T. Chociemski; Computed tomography reading centers: Part 1 — Henry Ford Health Sciences Center, J. Windham and H. Soltanian Zadeh; Part 2 — University of Virginia Medical Center, W. Alves, M. Keller, and J. Wenzel; Central Laboratory: Henry Ford Hospital, N. Raman and L. Cantwell; Drug Distribution Center: A. Warren, K. Smith, and E. Bailey; Committees — Executive: K. Welch, B. Tilley, and J. Marler; Steering: K. Welch, T. Brott, P. Lyden, J. Grotta, T. Kwiatkowski, S. Levine, M. Frankel, E. Haley, Jr., M. Meyer, B. Tilley, and J. Marler; Participants from Genentech: J. Froehlich and J. Breed; Data and Safety Monitoring Committee: J. Easton, J. Hallenbeck, G. Lan, J. Marsh, and M. Walker; Project Office — NINDS: J. Marler.
| |||||||||||||||||||||||||||||||||||||||||
Related Letters:
Tissue Plasminogen Activator for Acute Ischemic Stroke
Caplan L. R., Kwiatkowski T. K., Libman R. B., The NINDS Recombinant Tissue Plasminogen Activator Stroke Study Group
Extract |
Full Text
N Engl J Med 1999;
341:1240-1241, Oct 14, 1999.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
