Computed tomography (CT) is now used routinely

Size: px

Start display at page:

Download "Computed tomography (CT) is now used routinely"

Darrell Singleton
5 years ago
Views:

506 Computed Tomography in Prognostic Stroke Evaluation D. Rasmussen, MD; O. Kohler, MD; S. Worm-Petersen, MD; N. Blegvad, MD; H.L. Jacobsen, MSc; I. Bergmann, MD; M. Egeblad, MD; M. Friis, MD; and N.

1 506 Computed Tomography in Prognostic Stroke Evaluation D. Rasmussen, MD; O. Kohler, MD; S. Worm-Petersen, MD; N. Blegvad, MD; H.L. Jacobsen, MSc; I. Bergmann, MD; M. Egeblad, MD; M. Friis, MD; and N.T. Nielsen, MD Background and Purpose: Computed tomography is now routinely used in many hospitals to investigate cerebrovascular disease. The purpose of our prospective study was to determine whether cranial computed tomography in connection with neurological assessment was useful in prognostic evaluation of survival after acute stroke. Methods: Two-hundred forty-five consecutive stroke patients were included in the project during a -year period. Each had a detailed neurological assessment 4-7 hours after stroke onset and underwent cranial computed tomography without intravenous contrast injection within the first week after admission. The lesions were divided according to neuroanatomic regions. In the statistical analyses we used a multiple logistic regression model with survival/death as the binary variable. Results: Computed tomography showed 76% of the patients had infarcts, % had hemorrhages, and 3% had no acute lesion. Forty-three patients had more than one acute lesion, and 57 had one or more old infarctions. The temporal, parietal, and frontal regions and the basal ganglia were most often affected. Conclusions: We conclude that age, level of consciousness, and involvement of the temporal lobe on computed tomography were factors of prognostic significance regarding survival in the acute phase. (Stroke ;3:506-50) KEY WORDS cerebrovascular disorders prognosis tomography, x-ray computed Computed tomography (CT) is now used routinely in the investigation of cerebrovascular disease to distinguish between infarct and hemorrhage, in connection with intention to treat, and to confirm a diagnosis of stroke. The possible role of CT in the prognostic evaluation of stroke patients has not yet been established. When treating stroke patients there are two main questions to be answered. The first is whether the patient will survive; the second, what the functional outcome will be. Many studies have attempted to identify criteria for survival and functional prognosis after stroke, - focusing mainly on clinical neurological symptoms in patients in the acute phase 5 or during rehabilitation, 34 and have often included highly selected patients. The purpose of our prospective study was to analyze in all stroke patients seen in year the potential benefit of cranial CT in prognostic evaluation. In our first analyses, we focused on factors important for survival from the acute stroke and examined findings from CT scans performed systematically during the acute period on all stroke patients. Subjects and Methods All patients with stroke admitted to Bispebjerg Hospital in a -year period were included in the project. From the Departments of Neurology and Radiology, Bispebjerg Hospital, Copenhagen, and the Statistical Research Unit, the Panum Institute University of Copenhagen, Denmark. Address for correspondence: Dorte Rasmussen, Kuhlausgade 0, DK 00 Copenhagen, Denmark. Received March, ; accepted vember 6,. This general hospital in the city of Copenhagen receives patients either directly through the emergency ward or as referrals to the hospital by general practitioners. We defined stroke according to the World Health Organization definition. 3 Patients with subarachnoid hemorrhage were excluded. Patients were included regardless of age or earlier stroke and all had a detailed neurological assessment, including a bedside mental status test. 4 Examinations and registrations were performed by four neurologists who obtained information about preceding illness, including stroke and transient ischemic attack, and social conditions. Cranial CT was performed by means of a Siemens Somatom DRG scanner and analyzed by four radiologists who had no knowledge of the clinical details of the stroke. The CT was made without intravenous contrast injection. Neurological assessment was performed 4-7 hours after stroke onset. Two hundred fifty-two patients were initially included. Six patients were excluded because CT scan showed subdural hematoma in two, metastasis in three, and meningioma in one. In all six cases, the neurological symptoms were explained by the lesions found. A seventh patient was excluded because of meningitis. Thus, our final study consisted of 45 patients, 3 women and 4 men, with a mean age of 75 (range 6-6) years. Computed tomography was performed in 0 patients. Twenty-two of these underwent CT in the acute stage of stroke and 88 between days 3 and 8. Computed tomography was not performed in 35 patients: 6 died, seven were too ill, and two patients weighed more than the capacity level (0 kg) of the CT bed.

2 Rasmussen et al CT in Prognostic Stroke Evaluation 507 TABLE. Regional Localization of Acute Lesion by Computed Tomography in 8 Stroke Patients Location Frontal Parietal Temporal Occipital Basal ganglia Thalamus Brain stem Cerebellum. patients Percentage=regional load To determine factors of importance for surviving acute stroke, i.e., surviving the first 3 weeks after stroke onset, a multiple logistic regression model with survival/ death after 3 weeks as the dependent binary variable was analyzed. In the statistical analyses, we excluded 50 patients, including those comatose patients whose neurological examinations were limited and whose poor prognosis is well established, ' 5-7 ' 0 as well as patients with no acute CT. Thus, our multivariate analysis consisted of 5 patients, 6 of whom died within the first 3 weeks. Because of the relatively small number of deaths, it was not possible to analyze a model including all covariates simultaneously. Consequently, we used a kind of "forward selection" procedure in which the next most significant variable outside the model was successively taken in. Within the final model, we obtained estimated expected survival probabilities for patients with a certain combination of covariates. Furthermore, the effects of the remaining significant factors were easily interpreted using odds ratios. After the first analyses of survival from the acute stroke, we conducted a -year follow-up with neurological assessments, cranial CT scans, and Barthel scores 5 at 3 weeks and year after stroke to find factors of prognostic importance regarding functional outcome TABLE. Results of Cranial Computed Tomography in 0 Stroke Patients Variables. acute lesions > Size of acute lesion Lacunar xx cm >xxcm Depth of acute lesion Subcortical Cortical Edema/mass effect Old infarction One >One Periventricular hypodensity Atrophy Infarction (n=5) measured by Barthel score. The results of this study will be published at a later date. Results The results of the patients' first CT scans appear in Tables and. Of the 0 patients, 5 (76%) had infarction, 3 (%) had hemorrhage, and 8 (3%) had no definite acute lesion. Forty-three (0%) had more than one acute lesion; of these, three had hemorrhages. In Table, the size and location regarding depth of the most extensive lesion are listed. Approximately three fourths of the lesions were xx cm or larger (Table ). Mass effect/edema was seen in about 50% of the patients with infarction and in nearly all the patients with hemorrhage. 6 The regional distribution of the lesions is shown in Table. 7 Because of lesion size or the presence of more than one lesion, many patients are registered in several groups. The single percentage can nevertheless indicate the regional load or to what extent each region was affected. The lesions were equally distributed between the left and right hemispheres. The frontal, parietal, and temporal lobes were equally involved, as were the basal ganglias. The occipital lobe and thalamus were less often affected. Other CT findings independent of the present stroke are shown in Table. Fifty-seven patients (7%) had one or more old infarctions on CT. Only 7 of these patients had a recognized earlier stroke or TLA, which means that more than half of the old lesions resulted from silent strokes. Periventricular hypodensity was seen in 48 (3%) and atrophy in 58 (8%) of the patients. 8 Table 3 shows the results from the first neurological examination. The variables in Tables 4 and 5 were analyzed using the logistic regression model. Criteria for choosing these variables were based on results from marginal two-way tests (probability values shown in Table 4) and their anticipated clinical importance (Table 5). We made marginal two-way tables for all avail- Hemorrhage («=3) acute lesion («=8) oo to 6 0 Percentage (n=0) / * n > O4.J

3 508 Stroke Vol 3, 4 April TABLE 3. Acute Neurological Findings in 45 Stroke Patients Percentage Finding (n=45) Level of consciousness Awake Somnolent Coma.8 Mood rmal Abnormal Side of paresis Right Left 40.4 Both 7.8 Aphasia 5. Neglect 55.4 Hemianopsia Conjugate eye deviation Abnormal pupils 4.8 Hyperactive reflexes (+/- Babinski) 8. able data on noncomatose patients, which explains the total number of 6 for data about preceding illness and for data from the neurological examination. Twentyone of these patients did not undergo CT and were excluded in the further analyses. We found that age, level of consciousness, and lesions affecting the temporal lobe had significant effect on the probability of survival. The odds ratios of these variables seen in Table 6 show that the risk of death increases with rising age independently of other significant factors. For each year of age, the risk of dying is.06 times greater. Similarly, somnolent patients have a four times greater risk of dying than do patients who are awake in the acute phase. Also, patients with lesions affecting the temporal lobe have a poorer prognosis than do patients without the lesions. Prognostic values for the probability of being alive after 3 weeks for different combinations of age, level of consciousness, and temporal lobe location are shown in Figure. Discussion The patients were hospitalized in the departments of internal medicine and neurology. During the study period, treatment and rehabilitation of stroke patients continued as usual. This study is not epidemiological in nature, and the results cannot be used for incidence calculations. Our patients come primarily from the northwestern part of Copenhagen, whose population consists mainly of elderly people; this explains the relatively high mean age of 75 years and why the majority of our patients are women. An acute lesion was demonstrated in 87% of the patients in whom CT was performed within the first week after stroke onset without intravenous contrast injection. 4 The incidence of hemorrhage and infarction in our stroke patients was, as expected, approximately 0% hemorrhages The regional distribution of the lesions shows that the territories of the posterior cerebral artery, the occipital lobe, and the thalamus were TABLE 4. Univariate Results for Variables Chosen After Marginal Two-Way Tests and Considered for Use in the Logistic Regression Model Variable (n=6) Dead Age 0.0 Heart disease 8 6 (0%) 34 (%) Level of consciousness Awake 70 (6%) Somnolent 46 7(37%) < Hemianopsia 66 (8%) 6(5%) Missing Conjugate eye deviation 7 6 (35%) 4 8(5%) <0.000 Other 40 0 (5%) Missing 0 Hyperactive reflexes (+/- Babinski) Unilateral 35 (4%) Bilateral 33 8(4%) (%) (n = 5) Involvement of the temporal lobe on CTscan 63 (7%) 3 5 (4%) Involvement of the frontal lobe on CT scan 57 (6%) 38 7(5%) 0.00 Size of the biggest lesion on CT scan ne 6 0 Lacunar 46 (4%) xx cm 6 4(6%) >xxcm 6 0(6%) Edema or mass effect on CTscan 86 (4%) 0 4(4%) less affected than the territories of the middle and anterior cerebral arteries. The basal ganglia was noticeably affected in 4% of the patients. Four patients, too small a number to be useful, showed lesions of the brain stem. However, some brain stem lesions may be hidden in the group of patients with no acute lesions. Thus, after neurological examination and before the first CT, we proposed expected lesions of the brain stem in patients. We divided the lesions into functional regions 7 rather than by arterial distribution because we

4 Rasmussen et al CT in Prognostic Stroke Evaluation 50 TABLE 5. Univariate Results for Variables Chosen From a Neurological Point of View and Considered for Use in the Logistic Regression Model Variable Arterial hypertension Diabetes mellitus Abnormal pupils Missing (n=6) Atrophy (central) on CT scan 63 3 Periventricular hypodensity on CT scan 4 53 Dead 0 (6%) (%) 5 (0%) (%) (%) 4 (3%) 0. (3%) (%) (%) (0%) TABLE 6. Odds Ratios for Variables With Significant Effect on Probability of Being Alive 3 Weeks After Stroke Odds ratios Age.06 Level of consciousness 4.00 (somnolent/awake) Affection of the temporal lobe on CT scan 3. (yes/no) Confidence limits (.,.03) (7.4,.) (.3,7.7) 0.0 wanted to look at the lesions from a rehabilitation point of view. Finding a number of old infarctions on CT was expected because the patients were included in the study regardless of any history of earlier stroke. Like the Framingham study, 3 we also expected to find a number of silent strokes. They found 0%, whereas we found 4%, with the higher percentage in our study probably explained by our high mean age of 75 years. Findings related to more general arteriosclerotic changes, such as periventricular hypodensity and atrophy, were found in 3% and 8% of the cases, respectively. Six of the patients with old infarctions had both atrophy and periventricular hypodensity. It is not possible to estimate what was expected regarding atrophy in our population. A comparison with other series is complicated, primarily by different ways of measuring atrophy PDEAD FIGURE. Probabilities of death (PDEAD) within the first 3 weeks after stroke onset for different combinations of age, level of consciousness, and affection of the temporal lobe on computed tomographic scan. O, Awake with affection of the temporal lobe; *, somnolent with affection of the temporal lobe;, awake without affection of the temporal lobe; and n, somnolent without affection of the temporal lobe

5 50 Stroke Vol 3, 4 April and, secondarily, by differences in selection criteria, especially regarding age and earlier stroke. Our purpose was to find statistically significant factors for survival from stroke during the first 3 weeks after onset. In the statistical analyses, we used all available data on the patients, including information about earlier stroke. From the primary analyses, it was obvious that the variables age and level of consciousness were of importance and that involvement of the temporal lobe revealed by CT was of additional independent significant importance. The significance of the other variables was then analyzed after correction of these three factors. In our study, as in previous series, age and level of consciousness were important factors for survival. 5 ' 7-0 Other neurological signs were less important, perhaps because our patients were heterogenous and included a group with symptoms of various degrees from earlier strokes. Patients with lesions of the temporal lobe had a poorer prognosis with respect to survival. Although we are unable to give a precise explanation of this finding, we doubt that the size or cortical involvement of these lesions is of importance because these two variables were analyzed separately and found to have no prognostic significance. However, as a link in autonomic regulation, the temporal lobe might be important in maintenance of vital functions. 4 The presence of edema, atrophy, old lesions, and periventricular hypodensity on CT scan had no prognostic influence on survival within the first 3 weeks. We conclude that CT, without intravenous contrast injection and performed systematically on all stroke patients within the first week after onset, can be of benefit. An acute lesion can be demonstrated in more than three fourths of patients, which is of prognostic importance because patients whose temporal lobe is affected have a four times greater risk of dying than patients without. References. Marquardsen I: The natural history of acute cerebrovascular disease. Acta Neurol Scand 6;45(suppl 38):7. Kotila M, Waltimo O, Niemi ML, Laaksonen R, Lempinen M: The profile of recovery from stroke and factors influencing outcome. Stroke 84;5: Jongbloed L: Prediction of function after stroke: A critical review. Stroke 86;7: Miller LS, Miyamoto AT: Computed tomography: Its potential as a predictor of functional recovery following stroke. Arch Phys Med Rehabil 7;60: Bonita R, Ford MA, Stewart AW: Predicting survival after stroke: A three-year follow-up. Stroke 88;: Sotaniemi KA, Pyhtinen J, Myllyla VV: Correlation of clinical and computed tomographic findings in stroke patients. Stroke 0;: Howard G, Walker MD, Becker C, Coull B, Feibel J, McLeroy K, Toole JF, Yatsu F: Community hospital-based stroke programs: rth Carolina, Oregon, and New York: III. Factors influencing survival after stroke: Proportional hazards analysis of 4 patients. Stroke 86;7:4-8. Kunitz SC, Gross CR, Heyman A, Kase CS, Mohr JP, Price TR, Wolf PA: The Pilot Stroke Data Bank: Definition, design, and data. Stroke 84;5: Bamford J, Sandercock P, Dennis M, Burn J, Warlow C: A prospective study of acute cerebrovascular disease in the community: The Oxfordshire Community Stroke Project 8-86:. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intercerebral and subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry 0;53:6-0. Sheikh K, Brennan PJ, Meade TW, Smith DS, Goldenberg E: Predictors of mortality and disability in stroke. / Epidemiol Community Health 83;37: Sacco RL, Wolf PA, Kannel WB, McNamara PM: Survival and recurrence following stroke: The Framingham Study. Stroke 8; 3:0-5. Bamford J, Sandercock P, Dennis M, Warlow C, Jones L, McPherson K, Vessey M, Fowler G, Molyneux A, Hughes T, Burn J, Wade D: A prospective study of acute cerebrovascular disease in the community: The Oxfordshire Community Stroke Project 8-86:. Methodology, demography and incident cases of first-ever stroke. J Neurol Neurosurg Psychiatry 88;5: Hatono S: Experience from a multicentre stroke register: A preliminary report. Bull World Health Organ 76;54: Strub RL, Black FW: The Mental Status Examination in Neurology, ed. Philadelphia, Pa, FA Davis Co, Mahoney FI, Barthel DW: Functional evaluation: The Barthel index. Rehabilitation (Stuttg) 65;4: Bories J, Derhy S, Chiras J: CT in hemispheric attacks. Neuroradiology 85;7: Kretschmann H-J, Weinrich W: Neuroanatomy and cranial computed tomography. Stuttgart/New York, Georg Thieme Verlag, Meese W, Kluge W, Gramme T, Hopfenmuller W: CT-evaluation of the CSF-spaces of healthy persons. Neuroradiology 80;: Boysen G, Nyboe J, Appleyard M, S0rensen PS, Boas J, Somnier F, Jensen G, Schnohr P: Stroke incidence and risk factors for stroke in Copenhagen, Denmark. Stroke 88;: Sandercock P, Molyneux A, Warlow C: Value of computed tomography in patients with stroke: Oxfordshire Community Stroke Project. Br Med J 85;0:3-7. Wang AM, Lin JC, Rumbaugh CL: What is expected of CT in the evaluation of stroke? Neuroradiology 88;30: Brott T, Marler JR, Olinger CP, Adams HP Jr, Tomsick T, Barsan WG, Biller J, Eberle R, Hertzberg V, Walker M: Measurements of acute cerebral infarction: Lesion size by computed tomography. Stroke 8;0: Kase CS, Wolf PA, Chodosh EH, Zacker HB, Kelly-Hayes M, Kannel WB, D'Agostino RB, Scampini L: Prevalence of silent stroke in patients presenting with initial stroke: The Framingham Study. Stroke 8;0: Wannamaker BB: Autonomic nervous system and epilepsy. Epilepsia 85;6:3-3

Localizing lesion locations to predict extent of aphasia recovery. Abstract

Localizing lesion locations to predict extent of aphasia recovery Abstract Extensive research has related specific lesion locations to language impairment in aphasia. However, far less work has focused