Early Neurological Deterioration within 24 Hours after Intravenous rt-PA Therapy for Stroke Patients: The Stroke Acute Management with Urgent Risk Factor Assessment and Improvement rt-PA RegistryMori M.a, c · Naganuma M.a · Okada Y.c · Hasegawa Y.d · Shiokawa Y.e · Nakagawara J.f · Furui E.g · Kimura K.h · Yamagami H.i · Kario K.j · Okuda S.k · Koga M.b · Minematsu K.a · Toyoda K.a
aDepartment of Cerebrovascular Medicine and bDivision of Stroke Care Unit, National Cerebral and Cardiovascular Center, Suita, cDepartment of Cerebrovascular Medicine, Cerebrovascular Center and Clinical Research Institute, National Hospital Organization Kyushu Medical Center, Fukuoka, dDepartment of Neurology, St. Marianna University School of Medicine, Kawasaki, eDepartments of Neurosurgery and Stroke Center, Kyorin University School of Medicine, Mitaka, fDepartment of Neurosurgery and Stroke Center, Nakamura Memorial Hospital, Sapporo, gDepartment of Stroke Neurology, Kohnan Hospital, Sendai, hDepartment of Stroke Medicine, Kawasaki Medical School, Kurashiki, iStroke Center, Kobe City Medical Center General Hospital, Kobe, jDivision of Cardiovascular Medicine, Department of Medicine, Jichi Medical University School of Medicine, Shimotsuke, and kDepartment of Neurology, National Hospital Organization Nagoya Medical Center, Nagoya, Japan Corresponding Author
Background: The initial 24 h after thrombolysis are critical for patients’ conditions, and continuous neurological assessment and blood pressure measurement are required during this time. The goal of this study was to identify the clinical factors associated with early neurological deterioration (END) within 24 h of stroke patients receiving intravenous recombinant tissue plasminogen activator (rt-PA) therapy and to clarify the effect of END on 3-month outcomes. Methods: A retrospective, multicenter, observational study was conducted in 10 stroke centers in Japan. A total of 566 consecutive stroke patients [211 women, 72 ± 12 years old, the median initial NIH Stroke Scale (NIHSS) score of 13] treated with intravenous rt-PA (0.6 mg/kg alteplase) was studied. END was defined as a 4-point or greater increase in the NIHSS score at 24 h from the NIHSS score just before thrombolysis. Results: END was present in 56 patients (9.9%, 18 women, 72 ± 10 years old) and was independently associated with higher blood glucose [odds ratio (OR) 1.17, 95% confidence intervals (CI) 1.07–1.28 per 1 mmol/l increase, p < 0.001], lower initial NIHSS score (OR 0.92, 95% CI 0.87–0.97 per 1-point increase, p = 0.002), and internal carotid artery (ICA) occlusion (OR 5.36, 95% CI 2.60–11.09, p < 0.001) on multivariate analysis. Symptomatic intracranial hemorrhage within the initial 36 h from thrombolysis was more common in patients with END than in the other patients (per NINDS/Cochrane protocol, OR 10.75, 95% CI 4.33–26.85, p < 0.001, and per SITS-MOST protocol, OR 12.90, 95% CI 2.76–67.41, p = 0.002). At 3 months, no patients with END had a modified Rankin Scale (mRS) score of 0–1. END was independently associated with death and dependency (mRS 3–6, OR 20.44, 95% CI 6.96–76.93, p < 0.001), as well as death (OR 19.43, 95% CI 7.75–51.44, p < 0.001), at 3 months. Conclusions: Hyperglycemia, lower baseline NIHSS score, and ICA occlusion were independently associated with END after rt-PA therapy. END was independently associated with poor 3-month stroke outcome after rt-PA therapy.
© 2012 S. Karger AG, Basel
Neurological deterioration within the initial couple of days affects stroke patients’ long-term outcomes [1,2,3,4,5,6,7,8,9,10,11]. Neurological deterioration during acute stroke is associated with initial stroke severity , large vessel occlusion [3,4], hypodensity >33% in the middle cerebral artery territory , the hyperdense middle cerebral artery sign on brain CT [5,6], cerebral edema on early CT [5,6], diabetes mellitus [4,5,7], hyperglycemia [6,8,9], high or low blood pressure [7,8], early recurrent ischemic stroke , and symptomatic intracranial hemorrhage (sICH) [6,11].
Intravenous (IV) thrombolytic therapy with recombinant tissue plasminogen activator (rt-PA) has been shown to improve stroke outcomes [12,13]. However, about half of the patients were not independent in their activities of daily living or died at 3 months despite IV rt-PA therapy . Since the initial 24 h after thrombolysis are critical for patients’ conditions, partly because of the frequent occurrence of intracranial hemorrhage (ICH), continuous neurological assessment and blood pressure measurement are required during this period of time . Changes in the neurological status during this period may decisively affect outcomes after thrombolysis. The aims of this study were to identify the clinical factors that were associated with early neurological deterioration (END) within 24 h after IV rt-PA therapy and to clarify the effect of END on 3-month stroke outcomes.
Patients and Methods
The Stroke Acute Management with Urgent Risk Factor Assessment and Improvement (SAMURAI) rt-PA Registry has a multicenter, hospital-based, retrospective, observational cohort design . Details of this study have been described previously [16,17,18]. In brief, this study involved 600 consecutive patients with acute ischemic stroke receiving alteplase from October 2005 through July 2008. All of the patients treated during the study period were registered sequentially. Informed consent was obtained from all study participants. Of these, 34 patients whose 24-hour National Institutes of Health Stroke Scale (NIHSS) scores were not available were excluded from this study. The remaining 566 patients were included. Each local Ethics Committee approved the retrospective collection of clinical data from the database and submission of the data to our central office. Each patient received a single alteplase dose of 0.6 mg/kg (the recommended dose in the Japanese guidelines and the approved labeling) intravenously, with 10% given as a bolus within 3 h of stroke onset, followed by a continuous IV infusion of the remainder over 1 h .
The data collected from the database of the SAMURAI rt-PA registers for the present study are listed in table 1. Neurological deficits were assessed using the NIHSS score just before and 24 h after rt-PA, and at discharge [median hospital stay 27 days, interquartile range (IQR) 18–44.5 days]. END was defined as a 4-point or greater increase in the NIHSS score at 24 h from the NIHSS score just before thrombolysis. The ischemic stroke subtype was defined according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) categories . Early ischemic change on CT was quantitatively calculated using the Alberta Stroke Programme Early CT Score (ASPECTS) [21,22]. To identify arterial occlusion sites, MR angiography, CT angiography, or ultrasound was performed .
|Table 1. Baseline clinical characteristics|
The outcomes investigated were any ICH and sICH within the initial 24–36 h, NIHSS score at discharge, excellent functional outcome corresponding to modified Rankin Scale (mRS) score 0–1, death and dependency (mRS 3–6), and death at 3 months. Any ICH was defined as CT evidence of a new ICH ; it was assessed by at least two experienced vascular neurologists at each stroke center . Symptomatic ICH was defined with neurological deterioration corresponding to an increase of ≥1 point from the baseline NIHSS score according to the NINDS/Cochrane protocol . Symptomatic ICH was also defined according to the Safe Implementation of Thrombosis in Stroke Monitoring Study (SITS-MOST) protocol as parenchymal hemorrhage type II combined with an increase of ≥4 points from the baseline NIHSS score . Outcomes at 3 months were assessed by clinical examination at a hospital clinic or by telephone survey for patients whose neurological deficits were too severe to visit the clinic. Mainly study assistance nurses carried out the follow-up survey by telephone. When the patients or their families could not be reached, they called repeatedly till they were successful. Five patients were lost to follow-up at 3 months and for these 5 patients the mRS scores at discharge were used as their 3-month follow-up status.
Statistical analysis was performed using JMP 9.0 statistical software (SAS Institute Inc., Cary, N.C., USA). Patients’ baseline characteristics were compared between those with and without END using χ2 tests, unpaired t tests, and the Mann-Whitney U test, as appropriate. To identify the clinical factors associated with END, multivariate analyses were performed. Sex and age were initially entered, and the other variables listed in table 1 were chosen by a backward selection procedure using the Bayesian information criterion for exclusion. In addition, to identify the association between END and stroke outcomes, multivariate analyses with a backward selection procedure were performed. Statistical significance was established at p < 0.05.
A total of 566 consecutive stroke patients (211 women, 72.0 ± 11.6 years old) were studied. Of these, 56 patients (9.9%, 18 women, 71.5 ± 9.3 years old) had END (fig. 1).
|Fig. 1. Change in NIHSS score between baseline score before thrombolysis and 24 h after it.|
The baseline clinical characteristics of patients with and without END are presented in table 1. Patients with END more commonly had diabetes mellitus (p = 0.010), dyslipidemia (p = 0.034), prior use of hypoglycemic agents (p = 0.016), prior use of statins (p = 0.022), and internal carotid artery (ICA) occlusion (p < 0.001) than patients without END. Systolic blood pressure (p = 0.005), blood glucose (p < 0.001), and hemoglobin A1c levels (p = 0.021) were higher in patients with END than in those without. Initial NIHSS scores just before thrombolysis were not significantly different between the two groups (p = 0.076).
Multivariate regression analysis to identify the clinical factors associated with END showed that higher blood glucose [odds ratio (OR) 1.17, 95% confidence intervals (CI) 1.07–1.28 per 1 mmol/l increase, p < 0.001], lower initial NIHSS score (OR 0.92, 95% CI 0.87–0.97 per 1-point increase, p = 0.002), and ICA occlusion (OR 5.36, 95% CI 2.60–11.09, p < 0.001) were independently associated with END (table 2).
|Table 2. Multivariate logistic regression analysis for END|
Of the 566 patients, 113 (20.0%) had some ICH, including 22 (3.9%) with sICH per NINDS/Cochrane and 7 (1.2%) with sICH per SITS-MOST. Compared to patients without END, patients with END more often had ICH (42.9 vs. 17.5%, p < 0.001), sICH per NINDS/Cochrane (19.6 vs. 2.2%, p < 0.001), and sICH per SITS-MOST (7.1 vs. 0.6%, p = 0.002). On multivariate regression analysis, END was independently associated with any ICH (OR 3.38, 95% CI 1.87–6.06, p < 0.001), sICH per NINDS/ Cochrane (OR 10.75, 95% CI 4.33–26.85, p < 0.001), and sICH per SITS-MOST (OR 12.90, 95% CI 2.76–67.41, p = 0.002).
The median NIHSS score at discharge of the 566 patients was 4 (IQR 1–13). The discharge NIHSS score of patients with END was different from that of patients without END [21 (IQR 11–37.5) vs. 3 (1–11), p < 0.001] (fig. 2).
|Fig. 2. Course of the NIHSS score during hospitalization. The horizontal line in the box is the median, the bottom and top of the box are the 25th and 75th percentile, respectively, and the ends of the whiskers are the minimum and maximum values.|
Five patients were lost to follow-up at 3 months, and their mRS score at discharge was used as the 3-month mRS score. None of these 5 patients showed END; at hospital discharge, 1 had mRS of 2, 2 had mRS of 4, and 2 had mRS of 5. Of the 566 patients, 190 (33.6%) had an excellent functional outcome (mRS 0–1), 295 (52.1%) had death and dependency (mRS 3–6), and 38 (6.7%) had died by 3 months. No patients with END were independent while 37.3% of patients without END were independent (p < 0.001, fig. 3). Patients with END had poorer stroke outcomes than those without END at 3 months (death and dependency: 91.1 vs. 47.8%, p < 0.001; death: 25.0 vs. 4.7%, p < 0.001).
|Fig. 3. mRS score at baseline and at 3 months.|
Multivariate regression analysis indicated that END was independently associated with death and dependency (OR 20.44, 95% CI 6.96–76.93, p < 0.001), as well as death (OR 19.43, 95% CI 7.75–51.44, p < 0.001, table 3), at 3 months.
|Table 3. Association of END with each outcome parameter|
In this observational study, the clinical factors associated with END within 24 h after IV rt-PA therapy and the effect of END on stroke outcomes were determined. The first major finding was that END was independently associated with higher blood glucose, lower baseline NIHSS score and ICA occlusion. The second major finding was that END was independently associated with early ICH after IV rt-PA. The third major finding was that none of the patients with END had an excellent outcome (mRS 0–1) at 3-month follow-up; they were more often dependent or had a fatal outcome (mRS 3–6) at 3 months after multivariate adjustment.
END was present in one tenth of the patients in our study. Previous studies indicated that END was present in 14–38% of patients after IV rt-PA [5,6] and in 13–40% of patients who did not receive thrombolysis [2,3,4,7,8]. However, it is important to note that the definition of END differed among the studies. The time interval to assess deterioration differs greatly among studies, including the initial 24 h [5,6,9,11], 36 h , 48–72 h [1,2,4,8,10], and 7 days after stroke onset . As an indicator for neurological deterioration, the Scandinavian Neurological Stroke scale [5,7], the Canadian Neurological Scale , and different cutoff scores of the NIHSS (increase of more than 1 , 3 , and 4 points [1,6,9,10,11]) were used. Thus, a direct comparison of the results among these studies is difficult.
In the present study, higher blood glucose was associated with END. Hyperglycemia is known to be a risk factor for poor outcome of stroke patients even after early recanalization following thrombolysis [9,25,26]. Hyperglycemia is also known to be a risk factor for thrombolysis-associated sICH . Possible mechanisms of hyperglycemia-associated neurological deterioration include endothelial damage, deteriorating tissue acidosis, and worsening of blood-brain barrier breakdown [9,25,26]. ICA occlusion was inversely correlated with early improvement (≥8-point decrease in the NIHSS score) 24 h after IV rt-PA in our single-center cohort . The present results may be the reverse side of the same coin. A reduction in local cerebral perfusion pressure after ICA occlusion with poor leptomeningeal collaterals could be a reason for END . It was paradoxical that the lower baseline NIHSS score was associated with END in the present study, the opposite of what was found in previous studies [2,3,5,6,7]. These unusual results may be due to a ceiling effect, preventing the high score from increasing further. Prior use of antithrombotic therapy and history of congestive heart failure may influence early neurological states after thrombolysis via growth of ICH [27,29], although they were not associated with END in the present study.
Symptomatic ICH often occurs within several hours after thrombolysis and is the most common cause of END . In our cohort, sICH was more than 10 times as frequent in END patients as in other patients. In addition, acute arterial reocclusion , acute recurrent stroke [10,11], edema progression [1,30], and noncerebral accidents including infections and cardiovascular events are less common causes of END, although these possible mechanisms of END were not assessed in the present study. Of these, large swelling edema and edema with ICH were reported to show a strong association with a poor 3-month outcome .
The clear messages from the present study are that END within 24 h excludes independence and is associated with a very high risk of death and dependency (OR 20.44) at 3 months. These findings are similar to those from previous trials on IV rt-PA [5,6]. Thus, careful stroke care to avoid END could lead to better 3-month outcomes.
Certain limitations need to be considered prior to interpretation of the present results. First, patients who did not receive rt-PA were not included in this study. Second, early recanalization of the occluded cerebral artery, which greatly affects early neurological status, was not assessed. Third, END was evaluated at 24 h and sICH was assessed within the initial 24–36 h. Fourth, biochemical variables were modified by onset of stroke and they might not reflect patients’ characteristics before stroke onset. Finally, the present results based on low-dose rt-PA therapy (0.6 mg/kg) may not be applicable to the regular dose therapy (0.9 mg/kg).
Hyperglycemia, lower baseline NIHSS score, and ICA occlusion were independently associated with END after rt-PA therapy. The present study clearly indicates that END within 24 h after thrombolysis has an important association with poor 3-month outcomes.
This study was supported in part by Grants-in-Aid (H20-Junkanki-Ippan-019 and H23-Junkanki-Ippan-010, chief investigator: Kazunori Toyoda) from the Ministry of Health, Labor and Welfare, Japan.
Yasushi Okada received an honorarium from Mitsubishi Tanabe Pharma and a consulting fee from Lundbeck. Jyoji Nakagawara received honoraria from Mitsubishi Tanabe Pharma, Kyowa Hakko Kirin, and Lundbeck. Masatoshi Koga received research support from the Japan Cardiovascular Research Foundation (the Bayer Scholarship for Cardiovascular Research). Kazuo Minematsu received research support from the Ministry of Health, Labour and Welfare, Japan, and research grants for Cardiovascular Diseases, Grants-in-Aid, the Foundation for Biomedical Research and Innovation, Mitsubishi Tanabe Pharma Corporation, and Kyowa Hakko Kirin Pharma, Inc., and Hitachi Medical Corporation. Kazunori Toyoda received research support from Grants-in-Aid from the Ministry of Health, Labour and Welfare, Japan and Mitsubishi-Tanabe Pharma Corporation.
Kazunori Toyoda, MD, Department of Cerebrovascular Medicine
National Cerebral and Cardiovascular Center
5-7-1 Fujishiro-dai, Suita, Osaka 565-8565 (Japan)
Tel. +81 6 6833 5012, E-Mail firstname.lastname@example.org
Received: February 23, 2012
Accepted: May 29, 2012
Published online: August 1, 2012
Number of Print Pages : 7
Number of Figures : 3, Number of Tables : 3, Number of References : 30
Vol. 34, No. 2, Year 2012 (Cover Date: September 2012)
Journal Editor: Hennerici M.G. (Mannheim)
ISSN: 1015-9770 (Print), eISSN: 1421-9786 (Online)
For additional information: http://www.karger.com/CED