mean age, 57.8 years [range, 36 to 71 years]) suffering from their first stroke (cerebral infarction) in the dominant

Transcription

1 1789 Upper Limb Somatosensory Evoked Potentials as a Predictor of Rehabilitation Progress in Dominant Hemisphere Stroke Patients 0. Keren, MD; H. Ring, MD; P. Solzi, MD; H. Pratt, PhD; Z. Groswasser, MD Background and Purpose: The aim of this study was to determine the predictive yield of upper limb short latency somatosensory evoked potential (USEP) in patients with first stroke in the dominant hemisphere. Methods: Nineteen patients (average age, 58 years) were evaluated twice: on arrival at the rehabilitation center, approximately 3 weeks after the stroke, and again approximately 10 weeks later. The clinical assessment included a quantitative evaluation of motor ability, independence in activities of daily living, and communication ability. USEP was recorded during the week of the initial clinical evaluation. Special attention was paid to the relations between USEP parameters and the dynamics of the clinical condition. Results: The seven patients in whom no cortical potential could be detected showed the worst outcomes; however, the existence of cortical potentials in the remaining 12 patients did not provide a precise prediction of their "rehabilitative capacity" (ie, the extent of their progress). A correlation was established between the amplitude of the potentials recorded over both hemispheres and changes in communication ability. Additional findings included an association between shortened central conduction time over the damaged hemisphere during the first month after stroke and improvement in motor ability. Conclusions: USEP can serve as an adjuvant tool for predicting the recovery progress of stroke patients. (Stroke. 1993;24: ) KEY WoRDs * evoked potentials, somatosensory * prognosis * rehabilitation Stroke is a leading cause of disability among all central nervous system disorders.' It is critical to triage stroke patients for rehabilitation services to make best use of limited resources. Many investigators have sought to define the variables that best predict outcome after stroke.2 Given that most of the functional changes occur during the first 6 months after stroke,3 evaluation should take place early within this period. Evoked potentials (EPs) can help in monitoring the functional integrity of the central nervous system.4 They may be used in comatose patients or in patients who have communication deficits. Some authors have tried to ascertain whether upper limb short latency somatosensory evoked potential (USEP) can be used as a predictive tool, but no definitive conclusions have been drawn, probably because of the diversity of methods and populations used.2 In the present study we attempted to determine the predictive capacity of USEP in a specific population of patients (first stroke in the dominant hemisphere for handedness). The clinical evaluation included quantification of motor ability, independence in activities of daily living (ADL), and communication ability. Received April 20, 1993; final revision received August 23, 1993; accepted August 23, From the Evoked Potentials Laboratory, Loewenstein Rehabilitation Hospital, Ra'anana, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv (O.K., H.R., P.S., Z.G.), and Evoked Potentials Laboratory, Technion, Haifa (H.P.), Israel. Correspondence to H. Ring, MD, Loewenstein Rehabilitation Hospital, PO Box 3, Ra'anana 43100, Israel. Subjects and Methods Patients The study included 19 patients (11 men, 8 women; mean age, 57.8 years [range, 36 to 71 years]) suffering from their first stroke (cerebral infarction) in the dominant hemisphere, hospitalized at Loewenstein Rehabilitation Hospital. Dominance was determined by the writing hand, as reported in the anamnestic information provided by the patient and/or the family: 16 had right-hand dominance and 3 left-hand dominance. Most had a communication disability. Cerebral infarction was diagnosed by medical history and computed tomography scan of the head. Clinical Evaluation The patients had two clinical evaluations: the first on admission (approximately 3 weeks after the acute stroke; mean, days) and the second approximately 10 weeks later (mean, 71 ± 14.6 days). Each type of evaluation was carried out by the same specialist on both occasions, as described below. (1) was evaluated using two established quantitative methods. First, the motoricity index (MOT), performed by a medical doctor, scores muscle power on a three-point scale. It uses weighted scores, so that a score out of 100 can be calculated to represent the power at a joint, on a limb, or in a whole side. Second, the Rancho Los Amigos assessment (RAN),6 performed by an occupational therapist, was originally developed for the evaluation of the integrated function of the total upper extremity of the hemiparetic patient. The test consists of 17 graded activities, arranged in seven levels by degree of difficulty.

2 1790 Stroke Vol 24, No 12 December 1993 Amplitude 2 uv/div N9 N13 \, D NI P15\ - -!0 '. D -l---r---1 T I---1 Latency 5 Millisec/Div k- Recording of upper limb somatosensory evoked potentials of a 56-year-old man 3 weeks after a first stroke to the Erb/Fz dominant hemisphere. Each montage reflects the re- C7/Fz sponses recorded after stimulation of the dominant hand (D) and of the nondominant hand (). Cortical potentials C3'IFz (D) (NI and PI) were detected only C4V/Fz () nondominant hemisphere after stimulation of the nonparetic hand. Similar potentials recorded over Erb's point and the cervical regions were detected after stimulation of both hands. over the (2) Independence in ADL was assessed by two established tests. First, the Sister Kenny test (ADLS),7 performed by a nurse, has been used routinely for ADL evaluation in our hospital for more than 15 years. Second, the Barthel index (ADLB),8 performed by an occupational therapist, is the most widely used method of ADL evaluation. It measures 10 different activities at 5-point increments, from 0 to 100. A cut point of 60 represents the threshold between independence and various levels of partial dependence, and a score of 20 or below reflects total dependence. (3) Communication ability was evaluated by the Israel Loewenstein Aphasia Test (ILAT), performed by a professional communication therapist. The ILAT is actually a battery of tests, in Hebrew, and has been used in our hospital for more than 15 years. Results are ranked on a scale of 7 (1, highest level; 7, lowest level) and reflect speech, comprehension, reading, and writing abilities. Upper Limb Somatosensory Evoked Potential Studies USEP studies were carried out during the same week as the first clinical examination. Scores were recorded with an EP system (Biologic Brain Mapping III apparatus). Electrical stimuli were delivered by surface electrodes to the median nerve at the wrist at 4.1 Hz. Stimulus duration was 100 microseconds, and intensity was adjusted to just elicit a visible twitch of the thumb muscles. The patients lay in a darkened, air-conditioned room; no sedative drugs were administered. Silver-silver chloride electrodes were used for recording; impedance was kept below 3000 Q. The electrodes were placed over Erb's point, the seventh cervical vertebra, and contralateral somatosensory areas (2 cm behind C3 and C4). An electrode at F, was used as a reference. Two or more separate runs were performed for every test, and 512 nonrejected stimuli were averaged. Responses were filtered through a band pass of 1 to 3000 Hz. USEP measures included amplitudes of P15/NI and NI/PI, latencies of N9, N13, P15, NI and PI (cortical potentials), and central conduction time (CCT) (NI- N13).9-11 In each patient USEP amplitude was measured over the hemisphere contralateral to the stimulated median nerve. The difference between the two hemispheres was obtained by subtracting the amplitude over the affected hemisphere from the amplitude over the nonaffected hemisphere. Statistical Methods The spssx statistical system was used for data analysis. USEP results for the three left-dominant patients were reversed, so that in all cases the results of the dominant hemisphere could be grouped together. Student's t test was used to evaluate changes in each clinical parameter between the initial and the follow-up evaluations. Spear- TABLE 1. Differences in Performance Between Patients With Detectable Cortical Potentials and Those With No Potential Over the Affected Hemisphere Detectable Nondetectable Potential Potential (n=12) (n=7) Clinical Parameter Mean SD Mean SD P ADLB1* <.001 ADLB <.001 D-ADLB NS RAN <.001 RAN <.001 D-RAN <.04 MOT <.001 MOT <.003 D-MOT NS Comprehensionl <.001 Comprehension <.001 D-Comprehension NS ADLB indicates activities of daily living as assessed by the Barthel index; RAN, Rancho Los Amigos assessment of motor ability; and MOT, motoricity index. Numbers after tests indicate the following: 1, first test; 2, second test. D before test name indicates (score for second test) - (score for first test). *Similar findings were found for activities of daily living as assessed by the Sister Kenny test.

3 Keren et al USEP as a Predictor of Rehabilitation Progress 1791 TABLE 2. Correlations of Upper Limb Somatosensory Evoked Potential Parameters Observed Initially With Clinical Ability 2 Months Later, by Spearman's Analysis Statistical Correlations* USEP Side of Stimulation Clinical Parameters P r Conduction Latency NI D MOT Latency NI D ADLS Communication Latency PI D vs Speech Latency PI D Speech Latency NI D Speech Amplitudes P15/NI D vs MOT P1 5/NI D MOT NI/PI D RAN NI/PI D vs MOT Communication NI/PI D vs Comprehension NI/PI D Comprehension NI/PI D vs Speech P15/NI D vs Speech P1 5/NI D Speech P1 5/NI D Comprehension P15/NI D vs Comprehension USEP indicates upper limb somatosensory evoked potential; NI and PI, cortical potentials; D, dominant;, nondominant; MOT, motoricity index; ADLS, activities of daily living as assessed by the Sister Kenny test; RAN, Rancho Los Amigos assessment of motor ability. *Only significant correlations are included. man's analysis (nonparametric) was used, since it makes no assumptions regarding population distribution. For regression studies, four variables were taken as representative of the USEP and verified as independent parameters. Four dependent parameters were chosen to represent the clinical evaluation. Correlations were sought between the USEP parameters and those of the second clinical evaluation. The changes in the clinical findings from the first to the second evaluation were also assessed. The independent USEP parameters used were (1) CCT after stimulation of the dominant hand; (2) CCT after stimulation of the nondominant hand; (3) amplitudes (NI/PI) after stimulation of the dominant hand; and (4) amplitudes (NI/PI) after stimulation of the nondominant hand. The dependent clinical parameters were (1) motor ability, as measured by MOT and RAN; (2) ADL ability, as measured by ADLB; and (3) communication (speech and comprehension) ability, as measured by the ILAT. Results No NI was recorded over the affected hemisphere after stimulation of the paretic (dominant) hand in seven of the 19 patients (Figure). These patients had the worst outcomes: three remained totally dependent and the rest became partially dependent (Table 1). Significant functional improvement between the first and the second examinations was observed in all patients (Table 1). Patients without NI at the first examination started at a lower functional level than those with a detectable NI, but both groups had a statistically significant functional change. The amount of change in the clinical performance during this period did not differ significantly between patients with and those without NI, except for a borderline difference in motor ability assessed by RAN (P=.044) (Table 1); this comparison was performed only for patients with a detectable NI in the damaged hemisphere. The data regarding patient potentials detected over the affected hemisphere pertain to only 12 of the patients; data regarding the nonaffected hemisphere pertain to all 19 patients. Correlations between the USEP parameters and functional performance after 2 months of rehabilitative treatment were performed and are presented in Table 2. A significant correlation between NI latency over the dominant hemisphere and motor ability was found. Latency of PI was found to correlate with communication skills, and this was true for both hemispheres (Table 2). Other relations were found between amplitude and clinical performance. The two elements of amplitude (P15/NI

4 1792 Stroke Vol 24, No 12 December 1993 TABLE 3. Multiple Regression Analysis (Stepwise) for Upper Limb Somatosensory Evoked Potential Parameters Observed Initially With Changes in Clinical Performance Between First and Second Tests* Variables Entered D Clinical Parameters First R2* Total R2* D-MOT 1 NI/PI CCT CCT NI/PI D-RAN 1 NI/PI.34 2 NI/PI 3 CCT CCT ADL ability D-ADLB 1 CCT NI/PI CCT NI/PI Communication ability D-Comprehension 1 NI/PI CCT... 3 CCT... 4 NI/PI D indicates dominant;, nondominant; NI and PI, cortical potentials; CCT, central conduction time; D before test name, (score for second test) -(score for first test); MOT, motoricity index; RAN, Rancho Los Amigos assessment of motor ability; and ADLB, activities of daily living as assessed by the Barthel index. *Only significant correlations are included. and NI/PI) correlated with motor and communication abilities, respectively. The relation between amplitude and communication ability was found to be stronger than between amplitude and motor ability. The relation between amplitude and communication ability was even more pronounced when the difference in amplitude between hemispheres was compared (Table 2). We also sought possible relations between USEP parameters and changes in clinical performance during the rehabilitation period. Positive relations were found between CCT over the dominant hemisphere and changes in motor function alone, as evaluated by MOT (P>.046; r=.50). Direct relations were found between amplitude of NI/PI and changes in motor performance (for RAN, P>.007; r=.67); this was true also for communication ability and for amplitude of P15/NI. These relations were evident over both hemispheres and were stronger for interhemispheric latency difference (P>.002; r=.80). Regression analysis performed for USEP parameters and the changes in clinical functional ability during this period of rehabilitation yielded the following findings (Table 3). (1) In regard to motor ability, only amplitude in the damaged hemisphere was found to have a significant correlation with change in motor ability as measured by RAN. This predicted level measured by R2 was.34 and increased after the addition of the other parameters to.6. (2) In regard to ADL ability, no significant correlation was found for any one of these parameters. However, a statistically significant R2 (.4) was found for the combination of all of them. (3) In regard to communication ability, a statistically significant correlation was found only between the amplitude in the unaffected hemisphere (after stimulation of the nondominant hand) and speech ability (R2=.52). After combining the other USEP parameters, R2 measured.65. A similar and even stronger correlation was found between the USEP parameters and change in comprehension ability (between both tests). The combined R2 was.85, and for the relation between comprehension and amplitude alone, R2 was.73. Discussion Stroke-related damage to the central nervous system can be evaluated in several ways: clinical examination, imaging techniques (computed tomography, magnetic resonance imaging, single-photon emission computed tomography, and positron emission tomography)12'13 and physiological tests (EPs).14,15 Their use leads to a better

5 Keren et al USEP as a Predictor of Rehabilitation Progress 1793 understanding of central nervous system pathology, and this in turn may help in choosing the appropriate treatment plan for the short and long term. USEP is a neurophysiological tool that provides objective information about the activity of the somatosensory pathway. It was used in this study as an adjunctive mode of sensory evaluation of patients with severe language dysfunction. The purpose of the study was to assess the value of USEP as a predictor of functional ability and its improvement in stroke patients. USEP studies were performed about 3 weeks after stroke. Clinical evaluation of motor, ADL, and communication ability was performed at 3 and 10 weeks after stroke because most of the clinical and electrophysiological changes occur during this period.3 Results indicated that the absence of cortical potentials is a strong indicator of a poor functional performance. Similar findings were presented by Ring and Finnegan.16 Postrehabilitation functional performance was found to be poor in patients without cortical potentials in the first month after stroke. The presence of NIs is not predictive of the extent of change that will take place in functional ability during the first 3 months of poststroke rehabilitation. However, prolonged latencies and reduced amplitudes are associated with poorer performance levels. The significant relations observed between USEP response in the first month after stroke and clinical performance 2 months later may facilitate the use of this tool as a predictor of stroke outcome. These include the following. (1) Different components of the USEP response (ie, CCT and amplitude of NI) could be correlated to different aspects of function (motor, ADL, and communication ability). These components may represent different neurological abnormalities. (2) USEP correlates well with the severity of the neurological deficits, but it does not predict the degree of improvement over time. This may mean that USEP changes are related to neural damage but not to potential neural changes that could be activated during recovery. (3) A correlation between USEP and clinical performance was found over both hemispheres. This finding may indicate that changes in neural activity took place during the first month after stroke, and these affected the rehabilitation potential. The "end product" of EP studies is a curve. There are many ways to analyze USEP records.10 Different components of the USEP may refer to different clinical activities. There are few reports on the predictive capacity of USEP,17-23 and in all, only "gross" clinical criteria and general electrophysiological components were used. The study of Pavot et a120 included the largest group of patients (n=130), but their findings were limited because clinical assessment covered only gait and hand function. It is difficult to predict outcome within the middle band of stroke patients and even more difficult to estimate their potential for functional change. USEP has only limited prognostic value, and bedside examination is still an important tool for patient evaluation. However, together with clinical evaluation of motor and ADL ability, which has some predictive power, USEP may be a useful complementary tool, especially in patients with severe language dysfunction. This study also indicates that a combination of some components of the USEP response may have a stronger predictive value for recovery than the more commonly used single USEP component. A recommendation for future studies is to define a USEP score that may be calculated on the basis of a combination of components. It would be even more fruitful to combine such scoring with data from other evaluation techniques. References 1. Bach-y-Rita P. Brain plasticity as a basis of the development of rehabilitation procedures for hemiplegia. Scand J Rehabil Med. 1981;8: Davidoff G, Keren 0, Ring H, Solzi P, Werner RA. Assessing candidates of inpatient stroke rehabilitation: predictors of outcome. Phys Med Rehabil Clin North Am. 1991;2: De Weerd AW, Veldhuizen RJ, Veering MM. Recovery from cerebral ischaemia: EEG, cerebral blood flow and clinical symptomatology in the first three years after a stroke. Electroencephalogr Clin NeurophysioL 1988;70: Newlon PG, Greenberg RP, Hyatt MS. The dynamic of neural dysfunction and recovery following severe head injury assessed with serial multimodality evoked potentials. J Neurosurg. 1982;57: Demeurisse G, Demol 0, Robaye E. Motor evaluation in vascular hemiplegia. Eur Neurol. 1980;19: Wilson DJ, Baker LL, Craddock JA. Functional test for the hemiparetic upper extremity. Am J Occup Ther. 1984;38: Schoening HA, Anderegg L, Dergstrom D, Fonda M, Steinke N, Ulrich P. Numerical scoring of self-care status of patients. Arch Phys Med Rehabil. 1965;46: Mahoney FI, Barthel DW. Functional evaluation: the Barthel index. Md Med J. 1965;14: Keren 0, Groswasser Z, Sazbon L, Ring H. Somatosensory evoked potentials in prolonged postcomatose unawareness state following traumatic brain injury. Brain Inj. 1991;5: Chiappa KH. Short latency somatosensory evoked potentials. In: Chiappa KH, Yiannikas C, eds. Evoked Potentials in Clinical Medicine. New York, NY: Raven Press, Publishers; 1983: Walser H, Mattle H, Keller HM, Janzer R. Early cortical median nerve somatosensory evoked potentials: prognostic value in anoxic coma. Arch Neurol. 1985;42: Alberts MJ, Faulstich ME, Gray L. Stroke with negative brain magnetic resonance imaging. Stroke. 1992;23: Gilles D, Jean-Lun M, Pierre C, Aline S, Claude R, Jean-Degos D. Early and delayed SPECT using N-isopropyl p-iodoamphetamine iodine 123 in cerebral ischemia. Arch NeuroL 1987;44: De Weerd AW, Looijenga A, Veldhuizen RJ, Van Huffelen AC. Somatosensory evoked potentials in minor cerebral ischaemia: diagnostic significance and changes in serial records. Electroencephalogr Clin Neurophysiol 1985;62: Macdonell RAL, Donnan GA, Bladin PF. Serial changes in somatosensory evoked potentials following cerebral infarction. Electroencephalogr Clin Neurophysiol. 1991;80: Ring H, Finnegan JA. Somatosensory evoked potentials and rehabilitation outcome in stroke patients with hemi-inattention syndrome. Clin Rehabil. 1989;3: Liberson WT. Study of evoked potentials in aphasics. Am J Phys Med Rehabil. 1966;45: La Joie WJ, Nanjappareddy MR, Melvin JL. Somatosensory evoked potentials: the predictive value in right hemiplegia. Arch Phys Rehabil. 1982;63: Kusoffsky A, Wadell I, Nilsson BY. The relationship between sensory impairment and motor recovery in patients with hemiplegia. Scand J Rehabil Med. 1982;14: Pavot AP, Ignacio DR, Kuntavanish A, Lightroote WE. The prognostic value of somatosensory evoked potentials in cerebrovascular accidents. Electromyogr Clin Neurophysiol. 1986;26: Jacobs H, Vanderstraeten G, Van Laere M. SEPs and central somatosensory conduction time in hemiplegics. Electromyogr Clin Neurophysiol 1988;28: MacDonell RAL, Donnan GA, Bladin PF. A comparison of somatosensory evoked and motor potentials in stroke. Ann Neurol. 1989;25: Zeman BD, Yiannikas C. Functional prognosis in stroke: use of somatosensory evoked potentials. J Neurol Neurosurg Psychiatry. 1989;52: