Functional Outcome for Patients With Hemiparesis, Hemihypesthesia, and Hemianopsia
Does Lesion Location Matter?
Background and Purpose Patients with hemiparesis, hemisensory loss, and hemianopsia (“HHH” deficits) due to stroke may have large cortical lesions caused by middle cerebral trunk vessel occlusion or smaller subcortical lesions due to lenticulostriate involvement. We studied the usefulness of lesion location in predicting functional recovery within this syndrome.
Methods We reviewed our records and found 41 patients who had a single ischemic hemispheric stroke, HHH deficits, and an available CT scan performed more than 24 hours after the onset of symptoms. CT scans were read independently and blindly by the authors. Lesions were initially categorized by arterial distribution on the basis of CT templates published by Kinkel. The numerous combinations of arterial branch vessel occlusions observed did not allow for statistical analyses because of the small number of subjects within each subgroup. Lesions were therefore classified as cortical (C), subcortical (S), or mixed (M).
Results There were no significant differences among the three anatomic groups for age, sex, interval after stroke, Mini-Mental Status Examination score, or admission Barthel Index score. Functional outcome measures did not differ significantly for the three groups: mean±SD discharge Barthel score (C, 64±31; S, 47±20; M, 57±21), length of stay ([days] C, 64±25; S, 77±24; M, 73±28), and frequency of nursing home placement (C, 4/8; S, 3/6; M, 2/16).
Conclusions For patients with HHH deficits, the anatomic location of the lesion (C versus S versus M) does not affect functional outcome.
Approaches to predicting functional outcome after stroke have been both varied and unsuccessful. Demographic data have been most extensively studied; a recent review of the literature led Jongbloed1 to conclude that factors such as age, sex, side of lesion, and severity of motor impairment are only weak indicators of outcome. Neuroanatomically based prognostic systems using imaging techniques2 3 4 have not been clinically useful.
Syndromic classifications based on neurological impairments may hold more promise. The lacunar syndromes are generally thought to be compatible with good recovery; unfortunately, this impression is based more on anecdotes than on critical outcome studies. Only one study of these syndromes used a standardized outcome scale.4
Another syndromic approach is that of Reding and Potes,5 who propose an outcome-prediction schema for patients with hemiparesis after an initial unilateral ischemic hemispheric stroke. On the basis of presence or absence of hemiparesis, hemihypesthesia, and hemianopsia, stroke patients can be grouped to show statistically significant differences in both their final level of function and the time needed to reach maximal recovery. Patients with all three deficits—hemiparesis, hemihypesthesia, and hemianopsia (HHH syndrome)—have little likelihood of regaining independence but are able to reach a level of assisted care at which they can be safely managed by an aged spouse in their own home. The HHH syndrome is also associated with an increased risk of other stroke-related complications,6 including incontinence,7 the shoulder-hand pain syndrome,8 and thromboembolism.9 Outcome prediction within this group is particularly important because these patients are likely to have the longest rehabilitation stay with the lowest function at discharge.
We have observed that several anatomically different unilateral ischemic hemispheric lesions may cause the HHH syndrome. The HHH syndrome can occur after partial or complete middle cerebral artery distribution infarction. It may also occur with extensive subcortical infarction in the distribution of the lateral lenticulostriate vessels or with involvement of the anterior choroidal arterial distribution affecting the globus pallidus and posterior limb of the internal capsule. It is not yet known to what extent lesion location affects functional outcome for patients with HHH syndrome.
This study is a retrospective analysis of patients with HHH syndrome due to initial unilateral ischemic hemispheric stroke. We first define the spectrum of lesions seen on radiography that cause the HHH syndrome and then ask whether different stroke locations are associated with different functional outcomes for this group of patients.
Subjects and Methods
We reviewed the records of all patients admitted to an inpatient stroke rehabilitation unit over a 3-year interval to identify patients with the HHH syndrome. Hemiparesis was defined as having a score of 4 or less as graded by the Medical Research Council scale10 in the affected arm or leg in either proximal or distal muscle groups. Hemisensory loss was defined as having greater than a 6-inch error in localizing the affected index finger in any one of the four coronal spatial quadrants with the unaffected hand with eyes occluded. Visual field deficits were scored on the basis of confrontation testing. Patients were scored as hemianopic if they failed to perceive unilateral visual stimuli in either the upper or lower visual quadrants. Visual neglect was said to be present if patients responded to unilateral stimuli but failed to perceive bilateral simultaneous stimuli.
Only patients with unilateral ischemic hemispheric infarction were included in this study. Patients without an available CT scan or with a CT scan performed less than 24 hours after stroke onset were excluded from analysis. Forty-one patients met the above criteria for study inclusion.
CT scans were read independently and blindly by the two authors, and results were compared. In cases of disagreement, the final interpretation was reached by consensus. The CT templates of Kinkel11 were used initially to define cortical arterial distributions. For outcome evaluation, the large combination of arterial branch occlusions observed made statistical analysis of small groups impractical. To define sample sizes adequate for analysis, a simplified classification schema was used. Cortical infarcts that involved subcortical white matter, but not the basal ganglia or internal capsule, were classified as “cortical.” Infarcts were defined as “subcortical” if there was no visible cortical involvement and they involved the caudate nucleus, internal capsule, putamen, or globus pallidus. CT scans that showed no detectable infarcts were classified as subcortical. “Mixed” infarcts involved both cortical structures and the basal ganglia or internal capsule.
Self care and ambulation were scored using the Barthel Index and Barthel ambulation subscore.12 These assessments were made by rehabilitation staff who were unaware of the test hypothesis.
Data were analyzed using the Macintosh version of Statview512+ marketed by BrainPower Inc. Two-tailed, unpaired t tests were performed on interval scale data. The Kruskal-Wallis test was used for ordinal scale data. The χ2 statistic or Fisher’s exact test was used as appropriate to assess categorical data. All analyses were considered significant if the probability statistic was ≤.05.13
Forty-one patients met the inclusion criteria. Initial agreement was noted for 18 of 41 patients for both reviewers using the arterial templates published by Kinkel. Table 1⇓ shows the spectrum of arterial branch vessel patterns observed. Twelve patients had purely cortical strokes, 9 had purely subcortical strokes, and 20 had mixed infarcts involving both cortical and subcortical structures. Table 2⇓ displays demographic data comparing the three groups.
Table 3⇓ presents the results of the admission neurological examination. No significant differences were found in the Mini-Mental State Examination scores.14 Significant differences were found in the motor scores of the proximal upper extremity and both the proximal and distal lower extremity, suggesting that if significant strength was found in these areas, the infarct was likely to be purely cortical.
Table 4⇓ displays data on standard measures of self-care function. On admission to the rehabilitation unit, all patients were dependent on assistance for at least a portion of their self care. Total Barthel Index scores and the Barthel walking subscores were not statistically different among the three groups.
Outcome measures are shown in Table 5⇓. Virtually all patients remained at least partially dependent for function of activities of daily living. Weak trends suggested that the purely cortical patients do slightly better than the others, but these trends did not reach statistical significance in this sample.
Fig 1⇓ is a life-table analysis of the probability of recovering the ability to walk 150 ft with assistance. Only two patients eventually ambulated 150 ft unassisted, one of whom became independent with self care. Fig 2⇓ is a life-table recovery curve showing the probability of reaching a Barthel score of 60, a point at which assisted home care becomes practical.
This study is the first to examine the prognostic value of CT in a single clinical stroke syndrome using standardized outcome measures. Previous studies have used clinically heterogeneous populations, nonstandardized outcome measures, or both. Our data suggest that anatomic information obtained by CT does not add to the prognostic accuracy inherent in the HHH syndrome as defined by the extent of neurological impairment.
Our finding is consistent with previous neuroimaging-based classification systems in which clinical factors were more robust predictors of outcome than lesion location or size. Rascol et al15 studied the CT localization of pure motor strokes and incidentally commented on outcome. He reported that lateral lenticulostriate lesions were associated with “major sequelae” in three of four cases, whereas most patients with anterior choroidal or medial lenticulostriate lesions had “favorable” outcomes or “complete recovery.” Other reports16 17 studied isolated subcortical strokes but did not focus on either functional recovery or the spectrum of associated neurological impairments produced.
Clinicoanatomic correlation is probably most advanced in the field of aphasiology. Comparisons of clinical syndromes with imaged lesions show that only finely detailed lesion mapping such as that done by Naeser et al18 has yielded useful information in predicting either the rate of recovery or the final level of language disability. These studies have emphasized the effect of subcortical lesion location on language recovery. No similar attempts at finely mapping subcortical motor pathways in humans have yet been published. Little information is available on humans concerning the variability of primary motor pathways within either the centrum semiovale or the internal capsule.
There are several reasons why CT lesion localization might not improve outcome prediction. Early CT scanning may not reflect the true extent of the lesion. The limitations of CT imaging are well known and include both overestimation and underestimation of lesion size, “fogging” of lesions subacutely, and relatively poor anatomic resolution.19 Additionally, a more refined anatomic analysis in a larger group of patients might yield more information.
No studies have evaluated MRI and stroke outcome prediction, but MRI does not always clearly distinguish ischemic tissue from early infarct.19 Diffusion-weighted MRI may hold some promise for early lesion definition. Preliminary data from MR spectroscopy of stroke lesions suggest that preservation of normal N-acetyl group levels and phosphocreatine-creatinine ratios is associated with better stroke recovery.20
This study using CT technology is a first step in melding clinical and neuroradiological information into a useful outcome prediction schema. It is unusual in that we have examined functional recovery for a group of patients with a single clinical syndrome using standard outcome measures. We have demonstrated that within the HHH syndrome there is little correlation between cortical, subcortical, or mixed lesion location and either neurological impairment or extent of functional recovery.
- Received May 19, 1995.
- Revision received July 18, 1995.
- Accepted July 18, 1995.
- Copyright © 1995 by American Heart Association
Jongbloed L. Prediction of function after stroke: a critical review. Stroke. 1986;17:765-776.
Reding MJ, Potes E. Rehabilitation outcome following initial unilateral hemispheric stroke: life-table analysis approach. Stroke. 1988;19:1354-1358.
Dromerick A, Reding MJ. Medical and neurological complications during stroke rehabilitation. Stroke. 1994;25:358-361.
Reding MJ, Winter SW, Hochrein SA, Simon HB, Thompson MM. Urinary incontinence after stroke: a neurologic-epidemiologic perspective. J Neurol Rehabil. 1987;1:25-30.
Chalsen G, Fitzpatrick K, Navia R, Bean N, Reding MJ. Prevalence of the shoulder-hand syndrome in an inpatient stroke rehabilitation population. J Neurol Rehabil. 1987;2:51-57.
Bromfield E, Reding MJ. Relative risk of deep venous thrombosis or pulmonary embolism post stroke based on ambulatory status. J Neurol Rehabil. 1987;2:51-57.
Medical Research Council. London, UK: Balliere Tindall; 1985.
Kinkel W. Computerized tomography in clinical neurology. In: Baker AB, Baker LB, eds. Clinical Neurology. Philadelphia, Pa: JB Lippincott; 1988:37-58.
Mahoney F, Barthel D. Functional evaluation: the Barthel Index. Md Med J. 1965;2:61-65.
Rimm A, Hartz A, Kalbfleish J. Basic Biostatistics in Medicine and Epidemiology. Norwalk, Conn: Appleton Century Crofts; 1980.
Rascol A, Clanet M, Manelfe C, Guiraud B, Bonafe A. Pure motor hemiplegia: CT study of 30 cases. Stroke. 1982;13:11-17.
Levine RL, Lagreze HL, Dobkin JA, Turski PA. Large subcortical hemispheric infarctions: presentation and prognosis. Stroke. 1988;45:1074-1077.
Weiller C, Ringlestein B, Reiche W, Thron A, Buell U. The large striatocapsular infarct: a clinical and pathophysiological entity. Stroke. 1990;47:1085-1091.
Naeser MA, Palumbo CL, Helm-Estabrooks N, Stiassny-Eder D, Albert ML. Severe nonfluency in aphasia: role of the medial subcallosal fasciculus and other white matter pathways in recovery of spontaneous speech. Brain. 1989;112:1-38.
Bryan RN, Levy LM, Whitlow JM, Preziosi TJ, Rosario JA. Diagnosis of acute cerebral infarction: comparison of CT and MRI imaging. AJNR. 1991;12:611-620.
Corey CC, Griffey RH, Matwiyoff NA, Rosenberg GA. Multivoxel 1H-MRS of stroke. Neurology. 1992;42:1408-1412.