Acute Ischemic Strokes Improving During the First 48 Hours of Onset: Predictability, Outcome, and Possible Mechanisms
A Comparison With Early Deteriorating Strokes
Background and Purpose Our aims were to identify predictors of early neurological improvement in acute ischemic stroke patients, to evaluate its impact on clinical outcome, and to investigate possible mechanisms.
Methods A consecutive series of 152 first-ever ischemic hemispheric stroke patients hospitalized within 5 hours of onset underwent a first CT scan within 1 hour of hospitalization, and the initial subset of 80 patients also underwent angiography. During the first 48 hours of hospital stay, an increase or a decrease of 1 or more points in the admission Canadian Neurological Scale (CNS) score was defined as early improvement or early deterioration, respectively. Repeated CT scan or autopsy was performed 5 to 9 days after stroke.
Results Thirty-four patients (22%) improved, 84 (56%) remained stable, and 34 (22%) deteriorated. Logistic regression, which took into account vascular risk factors, baseline clinical and CT data, and therapies administered, selected younger age, lower admission CNS score, and absence of early hypodensity at first CT as independent predictors of early improvement. Among the patients who underwent angiography, logistic regression selected arterial patency and presence of collateral blood supply as independent predictors of early improvement. At the repeated CT scan or autopsy, improving patients presented the highest frequency of small infarcts. Thirty-day case-fatality rate and disability were lower in improving patients. Variables independently associated with outcome at logistic regression were admission CNS score, early deterioration, and early improvement.
Conclusions Early improvement can be predicted by the absence of early CT hypodensity and is highly predictive of good outcome. Presence of collateral blood supply and presumably early spontaneous recanalization are likely to be the mechanisms underlying early improvement.
Most of the studies on the early clinical course of acute ischemic stroke patients deal with neurological deterioration,1 2 3 4 5 6 7 while few studies have focused on early improvement.8 9 Biller et al,8 by studying 29 patients hospitalized from 6 to 12 hours after stroke onset, observed a major neurological improvement (which they defined as a decrease of 2 or more points in the NIH Stroke Scale score) in 21% of patients within 1 hour and in 51% within 6 hours of hospital entry. However, they did not specify whether this improvement was predictive of long-term outcome. More recently, Wityk and colleagues,9 by using a slightly modified version of the NIH scale, observed a major neurological improvement (which they defined as a decrease of 4 or more points in the NIH score) in 28% of their patients 48 hours after stroke onset. Ninety-one percent of those patients remained improved at long-term follow-up (average, 44 days), which suggests that early improvement is highly predictive of a sustained good outcome, even though outcome in that study was not defined in terms of activities of daily living.
The implications of this frequently observed spontaneous improvement for the design of pharmacological trials of acute treatment are evident. We recently addressed this issue from the point of view of early deterioration.7 In a consecutive series of 152 acute ischemic stroke patients admitted within 5 hours of onset, we found that approximately one fourth of the cases presented early deterioration within 96 hours of stroke onset, and that this event predicted a poor 30-day outcome in 90% of this subgroup of patients as opposed to 58% for the whole series. In that study, approximately one third of all nondeteriorating patients made some degree of improvement in the first 48 hours. These patients are the object of the present study, which is aimed at identifying predictors of early improvement and establishing whether it can predict good functional outcome 30 days after stroke onset. Moreover, we investigated the possible mechanisms underlying early improvement.
Subjects and Methods
The methods of patient selection and referral to our Stroke Unit have previously been described in detail elsewhere.7 All patients were admitted within 5 hours of the onset of their first-ever acute ischemic supratentorial stroke. Neurological deficit at entry and during the first days of hospitalization was quantified according to the first version of the CNS,10 which evaluates level of consciousness, orientation, speech, and facial and limb strength, with a global score ranging from 1.5 (maximum deficit) to 10 (absence of deficit). Since all patients were noncomatose and had a moderate to severe neurological deficit at hospitalization, the initial CNS score was never lower than 1.5 or higher than 8.5. An increase in the global score of 1 or more points within the first 48 hours of stroke was considered early improvement, whereas a decrease in the global CNS score of 1 or more points was considered early deterioration. The remaining patients were defined as stable (they are hereafter defined as nonimproving patients when grouped together with deteriorating subjects). Moreover, on the basis of the descriptive neurological examination and partially modifying criteria of Bamford et al,11 we classified our patients as presenting a pure motor hemiparesis or a sensorimotor stroke syndrome and a partial or total hemispheric syndrome.
At admission we registered blood pressure and performed an electrocardiogram. The patient or a proxy was interviewed regarding past medical history to establish the presence of any risk factor for stroke such as arterial hypertension, diabetes, atrial fibrillation, other potentially embolic cardiopathies,12 previous transient ischemic attacks, and cigarette smoking.
Within 1 hour of hospitalization, all patients were submitted to a plain CT scan, in which we looked for the presence of early focal hypodensity,13 involving the lentiform nucleus and/or the cortex, and of initial mass effect, defined as slight compression of ventricles. A repeated CT scan was performed in the first week after stroke, and the scans were reviewed to evaluate site and size of the qualifying infarct, presence and severity of mass effect, and possible occurrence of hemorrhagic transformation.14 According to standard templates,15 the site of the infarct was defined as (1) subcortical, when internal border zone or deep MCA branch areas were involved, with additional differentiation between nonlacunar and lacunar infarcts, the latter defined as a subcortical sharply delineated lesion with a diameter equal to or less than 15 mm16 ; (2) cortical, when MCA superficial branch territories were involved; (3) cortico-subcortical, when concomitant involvement of MCA deep and superficial territories was present; or (4) other territories, when other supratentorial non-MCA areas were involved. The size of the infarct was quantified as follows: (1) small, when the lesion involved less than one half a lobe or CT scans were permanently negative; (2) medium, when the lesion involved one half to 1 lobe; and (3) large, when the lesion involved more than 1 lobe.17 Mass effect was defined as (1) slight, when only a compression of ventricles without dislocation was present; (2) moderate, when we observed a partial ventricular shift across the midline; and (3) severe, when we observed a total ventricular shift across the midline. The site and size of lesion, mass effect, and occurrence of hemorrhagic infarction were determined by autopsy in the patients who died before the second CT examination.
Immediately after the first CT scan, the initial subset of 80 patients also underwent digital subtraction angiography by direct intracarotid injection of contrast medium. Patients with ICA occlusion underwent an additional contralateral angiography unless this procedure was contraindicated. Collateral blood supply was considered to be present when the branches of the occluded vessel showed retrograde flow through cortical anastomoses within 5 seconds of the end of the carotid injection.18
During hospitalization, the patients received the medical therapies required by the concomitant diseases. In addition, when indicated,19 we administered osmotics, antiplatelet agents, subcutaneous or intravenous heparin, and oral anticoagulants, sometimes in combination. No patient received thrombolysis.
The patients were followed up for 30 days after stroke. During this period we calculated the case-fatality rate and we differentiated cerebral from cardiac and other (pulmonary embolism, sepsis) causes of death on clinical grounds, resorting to autopsy only in doubtful cases. Activities of daily living of surviving patients were quantified by considering a Barthel Index20 score lower than 60 as indicative of a poor functional outcome.
First, χ2, χ2 for trend, Fisher's exact test, and one-way ANOVA were applied to analyze clinical characteristics on admission, frequency of risk factors for stroke in past medical history, CT scan, and angiographic findings, with both overall and two-by-two tests (when indicated). Then, to look for independent predictors of early improvement (stable or deteriorating=nonimproving=0, improving=1), the baseline clinical and CT variables, which at the univariate tests were trend-related to early improvement, were taken as independent variables in the stepwise logistic regression analysis. Intravenous heparin, subcutaneous heparin, oral anticoagulants, antiplatelet agents, and osmotics were also included in the multivariate analysis as individual variables (0=not administered; 1=administered) irrespective of any possible combination and of the strength of their association with early improvement at the univariate analysis. Finally, to evaluate whether early improvement could predict 30-day outcome (good=0, death or poor=bad=1), we performed a logistic regression analysis, which also included the early clinical course.
Of 152 patients studied, 34 (22%) improved within the first 48 hours of hospitalization; 29 (19%) of these improved within the first 24 hours and the remaining 5 (3%) improved after 24 hours. Of the nonimproving patients, 84 (56%) were stable and 34 (22%) had a deteriorating course in the first 48 hours.
Table 1⇓ shows the overall comparison of demographic data, vascular risk factors, and baseline clinical and CT characteristics in the three subgroups of early clinical course. At two-by-two tests of improving versus nonimproving patients (stable or deteriorating), the former were younger (P=.001) and showed a trend toward lower systolic blood pressure, lower serum glucose levels, more severe neurological deficit, and earlier hospitalization. There were no differences between improving and nonimproving patients regarding sex, diastolic blood pressure, and frequency of lacunar, partial anterior, and total anterior circulation syndromes on admission. Finally, except for a nonsignificant lower frequency of hypertension in their past medical history, improving patients presented a pattern of risk factors for stroke similar to that of nonimproving individuals.
At the first CT scan (Table 1⇑), there was a significantly increasing frequency of early hypodensity in improving, stable, and deteriorating patients. Initial mass effect was detected as frequently in improving as in stable patients and in both cases significantly less frequently than in deteriorating patients.
The second CT scan (144 patients) and autopsy (8 patients) (Table 2⇓) showed that site and size of the lesion and frequency of mass effect and of hemorrhagic transformation were different across the three subgroups. At two-by-two tests, cortical lesions were present with a similar frequency in the three subgroups of patients, whereas subcortical lesions, including lacunes, were equally represented in improving and stable patients and less frequently detected in deteriorating patients (P=.05). This last group showed mainly large cortico-subcortical lesions, which were least frequent among improving patients (P=.001). Finally, there was a steady increase in the frequency of medium and large infarcts, mass effect, and hemorrhagic transformation from improving to deteriorating patients.
Of the 80 patients who underwent angiography, whose baseline characteristics were similar to those of the entire population studied, 18 (22.5%) had a improving course, 44 (55%) were stable, and 18 (22.5%) had a deteriorating course. Angiographic findings are shown in Table 3⇓. Nine improving patients (50%) had normal angiograms or nonstenosing ICA plaques compared with 8 stable patients (18%) and 2 deteriorating patients (11%) (P=.02, χ2 for trend). Among the occluded patients, 6 improving patients (67%) had MCA branch occlusion compared with 15 stable patients (42%) and 6 deteriorating patients (37%) (P=.2). Collateral blood supply was present in 7 of 9 improving (78%), 18 of 36 stable (50%), and 7 of 16 deteriorating patients (44%) (P=.1).
Regarding drug therapy, early improving patients were given osmotics, antiplatelet agents, and intravenous heparin as frequently as nonimproving patients. On the other hand, 6 improving patients (18%) received subcutaneous heparin compared with 33 stable (39%) and 15 deteriorating patients (44%) (P=.04), whereas oral anticoagulants were given respectively to 16 (47%), 23 (29%), and 8 patients (21%) (P=.04).
Among the baseline variables, independent predictors of early improvement selected by stepwise logistic regression were younger age (P<.01), lower CNS score at entry (P=.02), and absence of early hypodensity at first CT (P=.01). Individual drug therapies did not appear to influence the early clinical course. In the subgroup of patients who underwent angiography, independent predictors were arterial patency (P<.01) and presence of collateral blood supply (P<.05).
During the 30-day follow-up, 1 improving patient (3%) died, 6 (18%) had a poor outcome, and 27 (79%) had a good outcome, compared with 14 (17%), 38 (45%), and 32 stable patients (38%) and with 13 (38%), 17 (50%), and 4 deteriorating patients (12%), respectively (P=.0000). The death of the improving patient was due to cerebral causes following a large new stroke of unknown pathogenesis with secondary hematoma. The causes of death were cerebral in 1 (7%), cardiac in 5 (36%), and other in 8 stable patients (57%), compared with 8 (62%), 2 (15%), and 3 deteriorating patients (23%), respectively (P=.03).
Among the aforementioned baseline clinical and CT variables and the three types of early clinical course, in addition to CNS score at entry (OR, 0.50; 95% CI, 0.38 to 0.67), logistic regression also selected early deterioration (OR, 4.54; 95% CI, 1.24 to 16.7) and early improvement (OR, 0.06; 95% CI, 0.01 to 0.21) as predictors of 30-day outcome. Again, at multivariate analysis, therapy with any of the aforementioned drugs did not influence final outcome.
Our study confirms that early spontaneous improvement is a frequent occurrence in the clinical course of ischemic stroke patients.8 9 We monitored neurological improvement, as well as deterioration, over the first 48 hours of stroke9 to provide more complete information on the early clinical course of this pathology. According to the method of Côté and colleagues,21 we chose 1 as the minimum change in the global CNS score able to reflect a modification in the neurological status not liable to interobserver variability. This means that we used a merely quantitative definition of improvement, without the qualitative implications of the concept of “major neurological improvement” adopted in previous studies.8 9 Nevertheless, the fact that as many as 79% of early improving patients had a good 30-day functional outcome suggests that early improvement of a neurological deficit quantified by a neurological scale is predictive of good functional outcome as defined by an activities of daily living scale. This hypothesis is further supported by the fact that logistic regression analysis selected early improvement as an independent predictor of good functional outcome. Moreover, early deterioration predicted a 30-day bad outcome (death and poor functional outcome) in 88% of patients, a finding also confirmed by logistic regression. Since early improving and deteriorating patients together accounted for almost half of our series, we deduce that in approximately 40% of cases the 30-day functional outcome could be predicted on the basis of the neurological course in the first 48 hours of hospitalization. Therefore, both of these early neurological changes could be adopted as very short-term end points in pharmacological trials, in which reducing the frequency of stability and early deterioration in favor of early improvement could be one of the hypotheses of efficacy of a drug under test.
On the other hand, if from the time the patient is hospitalized we had the opportunity to identify the patients with subsequent spontaneous early improving or deteriorating course, we could exploit these data as criteria for enrollment in pharmacological trials. This would mean, for instance, that the treatment of spontaneously improving patients with potentially harmful drugs, such as thrombolytics, could be avoided. In our patients, among the clinical and CT data obtained at hospital entry, independent predictors of early improvement selected by logistic regression analysis were absence of early hypodensity at the first CT scan, younger age, and severe neurological deficit at entry. This latter observation, that at hospital admission early improving patients were as severely affected as early deteriorating patients, may appear quite contradictory with commonly held beliefs. However, evaluating the neurological deficit at entry of post hoc defined subgroups is quite a new point of view, which differs from the prospective evaluation of the impact of initial neurological deficit on outcome, such as is commonly reported in literature,22 so that no comparison is possible.
In reference to early CT signs, considering that the presence of early hypodensity had already been identified as a predictor of early deterioration,7 one could infer that the early clinical course might be easily predicted by the findings at a CT scan performed within 5 hours of stroke onset. However, the predictive power of this variable is unsatisfying, since only 32% of patients without early hypodensity improved while 38% of those with early hypodensity deteriorated in the first 48 hours of hospitalization. This means that with the diagnostic tools at present routinely available in an emergency setting, the identification of patients with spontaneous early improvement or deterioration is not reliable and that their exclusion from clinical trials of acute treatment is therefore unfeasible. Additional information may perhaps be provided by angiography, which in fact cannot be considered a routine examination. In the subset of patients submitted to angiography in our study, logistic regression analysis selected arterial patency (ie, normal angiograms or nonstenosing ICA plaques) and the presence of collateral blood supply as predictors of early improvement. In our previous study on deteriorating strokes monitored up to 96 hours after stroke onset,7 we highlighted the role of arterial occlusion and the absence of collateral blood supply in determining neurological deterioration. Since patients who deteriorated between 48 and 96 hours were a minority, those pathogenetic considerations also apply to the group of 48-hour deteriorating patients reported in this study. The positive predictive value of angiographic findings, namely collateral blood supply, in regard to the initial 48-hour clinical course is better, since 49% of patients with collateral blood supply improved and 46% of those without collateral blood supply deteriorated.
The angiographic study also shed some light on the possible mechanisms underlying early improvement. Arterial patency was detected in 50% of improving patients, whereas the remaining 50% had an intracranial or/and extracranial arterial occlusion, with collateral blood supply in approximately 80% of the cases. When we consider that all but one of the nonoccluded patients presented a CT territorial infarction, which is generally thought to be caused by an embolic arterial occlusion,23 it is conceivable that these patients had had an arterial occlusion and that arterial reopening had already occurred before they underwent angiography. Moreover, two thirds of the arterial occlusions shown by angiography were located in MCA branches, which are the most liable to spontaneous24 reperfusion. These data allow us to hypothesize that both the presence of collateral blood supply and spontaneous recanalization might have contributed to the improvement by limiting the extent of the final lesion.25 In fact, improving patients had the highest frequency of small infarcts on the repeated CT scan. Our considerations are obviously based on one photogram taken a few hours after stroke onset, and we do not know what really happened subsequently. Only a serial assessment of arterial status, with angiography or Doppler sonography, is likely to provide further insights into the mechanisms underlying the early clinical course in stroke patients.
Selected Abbreviations and Acronyms
|CNS||=||Canadian Neurological Scale|
|ICA||=||internal carotid artery|
|MCA||=||middle cerebral artery|
|NIH||=||National Institutes of Health|
This study was supported in part by Consiglio Nazionale delle Ricerche, grant No. 93051602086. We are indebted to Professor Cesare Fieschi for criticism and advice. We gratefully acknowledge the resident staff for patient selection and care.
Reprint requests to Dr Danilo Toni, Department of Neurological Sciences, I Chair of Neurology, University “La Sapienza,” Viale dell'Università, 30-00185, Rome, Italy.
- Received May 30, 1996.
- Revision received September 26, 1996.
- Accepted October 10, 1996.
- Copyright © 1997 by American Heart Association
Jones HR, Millikan CH. Temporal profile (clinical course) of acute carotid system cerebral infarction. Stroke. 1976;7:64-71.
Jones HR, Millikan CH, Burton AS. Temporal profile (clinical course) of acute vertebrobasilar system cerebral infarction. Stroke. 1980;11:173-177.
Baker RN, Broward JA, Fang HC, Fisher CM, Groch SN, Heyman A, Kay GR, Mc Devitt E, Scheinberg P, Schwartz W, Toole JF. Anticoagulant therapy in cerebral infarction. Neurology (Minneap). 1962;12:823-835.
Britton MD, Rödén A. Progression of stroke after arrival at hospital. Stroke. 1985;16:629-632.
Mohr JP, Caplan LR, Melsky JW, Goldstein RJ, Duncan GW, Kistler JP, Pessin MS, Bleich HL. The Harvard Cooperative Stroke Registry: a prospective registry. Neurology. 1978;28:754-762.
Dàvalos A, Cendra E, Teruel J, Martinez M, Genis D. Deteriorating ischemic stroke: risk factors and prognosis. Neurology. 1990;40:1865-1869.
Biller J, Love BB, Marsh EE, Jones MP, Knepper LE, Jiang D, Adams HP, Gordon DL. Spontaneous improvement after acute ischemic stroke: a pilot study. Stroke. 1990;21:1008-1012.
Wityk RJ, Pessin MJ, Kaplan RF, Caplan LR. Serial assessment of acute stroke using the NIH Stroke Scale. Stroke. 1994;25:362-365.
Cotè R, Hachinski VC, Shurvell BL, Norris IW, Wolfson C. The Canadian Neurological Scale: a preliminary study in acute stroke. Stroke. 1986;17:731-737.
Kittner SJ, Sharkness CM, Price TR, Plotnick GD, Dambrosia JM, Wolf PA, Mohr JP, Hier DB, Kase CS, Tuhrim S. Infarcts with a cardiac source of embolism in the NINCDS Stroke Data Bank: historical features. Neurology. 1990;40:281-284.
Okada Y, Yamaguchi T, Minematsu K, Miyashita T, Sawada T, Sadoshima S, Fujishima M, Omae T. Hemorrhagic transformation in cerebral embolism. Stroke. 1989;20:598-603.
Damasio H. A computed tomography guide to the identification of cerebral vascular territories. Arch Neurol. 1983,40:138-142.
Toni D, Del Duca R, Fiorelli M, Sacchetti ML, Bastianello S, Giubilei F, Martinazzo C, Argentino C. Pure motor hemiparesis and sensorimotor stroke: accuracy of the very early clinical diagnosis of lacunar strokes. Stroke. 1994;25:92-96.
Brott T, Marler JR, Olinger CP, Adams HP, Tomsick T, Barsan W, Bille R, Eberle R, Hertzberg V, Walker M. Measurements of acute cerebral infarction: lesion size by computed tomography. Stroke. 1989;20:871-875.
Saito I, Segawa H, Shiokawa Y, Taniguchi M, Tsutsumi K. Middle cerebral artery occlusion: correlation of computed tomography and angiography with clinical outcome. Stroke. 1987;18:863-868.
Warlow C. Disorders of the cerebral circulation. In: Walton J, ed. Brain's Diseases of the Nervous System. Oxford, England: Oxford Medical Publications; 1993:197-268.
Côtè R, Battista RN, Wolfson C, Boucher J, Adam J, Hachinski V. The Canadian Neurological Scale: validation and reliability assessment. Neurology. 1989;39:638-643.
Minematsu K, Yamaguchi T, Omae T. Spectacular shrinking deficit: rapid recovery from a major hemispheric syndrome by migration of an embolus. Neurology. 1992;42:157-162.
Ringelstein EB, Biniek R, Weiller C, Ammeling B, Nolte PN, Thron A. Type and extent of hemispheric brain infarctions and clinical outcome in early and delayed middle cerebral artery recanalization. Neurology. 1992;42:289-298.