Acute care physical therapist evaluation and intervention for an adult after right hemispherectomy. (Case Report).

Hemispherectomy, although radically invasive, is performed on patients whose seizures fail to respond to medical management. During a "classic" brain hemispherectomy, an entire cortical hemisphere is removed. Classic hemispherectomy initially was used in 1928 (1) as an intervention for diffuse malignant glioma, and subsequently it was used as a treatment for intractable seizures. (2-6) Reports about the late complications of the procedure, however, stimulated modification of the surgical technique. (7-9) Rasmussen (8) began performing "functional" hemispherectomies in 1974. This technique differs from the classic hemispherectomy in that it leaves one fourth to one third of the epileptogenic hemisphere (usually the frontal and occipital poles intact to provide physical support to the unaffected hemisphere. The tissue connections to the remaining brain are severed, and subcortical tissue is left in place. The hemispherectomy is functionally complete because the major pathways are removed or cut, but anatomically incomplete as portions of the supporting structures remain in place. This method has been used since its development by Rasmussen.

Hemispherectomy is performed relatively infrequently, because most seizures can be medically managed. The procedure is considered to control progressive and incapacitating seizures due to hemimegalencephaly, infantile hemiplegia seizure syndrome, infantile spasms, Lennox-Gastaut syndrome, Rasmussen encephalitis, and Sturge-Weber syndrome. (10) Hemispherectomy may be indicated when an individual has a seizure disorder uncontrolled by medical intervention, a hemiparesis contralateral to the hemisphere with seizure activity, and the absence of epileptiform discharges ipsilateral to the hemiparesis. (10) The goal of the surgery is to eliminate seizures without creating new neurological deficits. (10)

Both research and case reports indicate that hemispherectomy can be an effective intervention for seizure control. (1,11-15) About 80 % of patients have approximately 80% reduction in the frequency of their seizures after hemispherectomy. (1,11-15) Postoperative sensorimotor recovery and descriptions of subsequent functional limitations and quality of life are less well documented. In 1999, Holthausen and Strobl (16) reviewed studies from the 1950s and 1960s and described the functional outcomes of adults and children who had hemispherectomies for gliomas and infantile hemiplegia seizure syndrome, respectively. Most adults walked independently after the surgery, but were only able to flex and extend their hips and were unable to move their knees and ankles. Upper-extremity movement was essentially absent, and all sensory functions were impaired in the arm and leg. Results of the pediatric studies were mixed, but children generally had better recovery than adults. Children had decreased ankle strength (muscle strength is defined in the Guide to Physical Therapist Practice as "the force exerted by a muscle or group of muscles to overcome resistance ..." (17)(p688)) and lack of individual finger movements, but milder deficits in the more proximal parts of their limbs. The authors also compared children who had acquired lesions (eg, Rasmussen encephalitis, cerebrovascular accident) with those who had developmental disorders (eg, hemimegalencephaly, polymicrogyria). Based on their finding that individuals with acquired lesions had poorer outcomes, the authors suggested postoperative function could be predicted by the etiology of the seizure disorder and the associated potential for neural reorganization.

Vining et al (18) created a "burden of illness" scale to better describe postoperative outcomes and used it with 58 children who had hemispherectomies. Each received a score of 0 ("no burden") to 3 ("severe burden") on 3 scales: motor function, intellect, and seizures. All children had similar residual motor impairments after hemispherectomy, regardless of the underlying etiology for their seizures, and the level of function did not represent a significant change from preoperative status. Most children walked independently (many with an ankle-foot orthosis [AFO]) and used their affected upper extremity as a "helper hand," (18)(p164) but they could not perform individual finger movements. The burden of illness score, however, was often decreased postoperatively due to improvement in seizure control and reduced need for anticonvulsants. Although the authors stated that the scale needed refinement and validation, they noted that it represented an improvement in quantifying outcome because it measured more than seizure frequency.

In 1991, Muller et al (19) published the only case report to date describing the impairments of an adult undergoing hemispherectomy. They described a 46-year-old woman who had a right hemispherectomy at the age of 18 years for seizure control. The patient had "marked spasticity" (19)(p126) on the left side for 1 week after the operation that subsequently subsided. The patient regained her preoperative sensorimotor function within 2 weeks, and her only other postoperative deficit was a homonomous hemianopsia. A follow-up examination conducted 28 years after the surgery revealed impaired strength, sensation, and motor control on the left side, all of which had been present before the surgery. The patient's deficits were more pronounced in the distal parts of her extremities than in the proximal parts. The patient showed decreased ankle mobility when walking and a barely perceptible limp, but no abnormal movement of the hip and knee joints.

Muller and colleagues' case report (19) focused on the status of the patient many years after her surgery. We found no literature describing and measuring the existence or progression of impairments, functional limitations, and disabilities of adults during the acute recovery phase after hemispherectomy. In addition, none of the literature described perioperative physical therapist management for this patient population. The purpose of this case report is to describe and discuss preoperative and postoperative physical therapist examination, evaluation, and intervention during the hospitalization of an adult who underwent a right functional hemispherectomy.

Case Description

Patient

The patient ("CB") was a 27-year-old right-handed man with uncontrollable seizures. CB had a traumatic brain injury at age 5 years. Staring and drooling episodes started 2 weeks after the injury. Within 1 year, left-sided focal motor seizures began, and CB started taking anticonvulsant medications, with little improvement in seizure control. Subsequently, he developed progressive epilepsia partialis continua, a syndrome of continuous focal jerking of a body part that can last from days to years. (20) CB also had generalized seizures, during which he would lose consciousness, fall, and often strike his head.

At age 14 years, CB had a right temporal lobectomy, which failed to reduce the frequency of his seizures. He therefore underwent a partial corpus callosotomy at age 18 years. The procedure was aborted, however, prior to completion because of a hemorrhage along the sagittal sinus during the surgery. This complication resulted in a left hemiparesis, and the surgery failed to produce the desired reduction in CB's seizures.

Throughout the next 9 years, physicians prescribed numerous combinations of anticonvulsant medications to control CB's seizures, without success. CB's multiple daily seizures, left hemiparesis, and impaired cognitive and behavioral status interfered with his ability to be independent with functional activities. CB lived in a group home where he received assistance for activities of daily living (ADL). Because he was impulsive and had poor judgment, CB needed someone with him while he walked to cue him to avoid unsafe situations. He was often agitated and violent, and could not keep a job, although he earned his general equivalency diploma. CB had supportive parents, who were involved in all aspects of his care.

At age 27 years, CB began having more frequent complex partial seizures in addition to the epilepsia partialis continua, despite his drug regimen of Tegretol* (2,600 mg/d), Neurontin ([dagger]) (3,600 mg/d), and phenobarbital (270 mg/d). Partial seizures, also called "focal seizures," can cause motor, somatosensory, autonomic, or psychic symptoms, depending on the location of the epileptic focus. (21) During a complex partial seizure, consciousness also is impaired. CB was admitted to Beth Israel Deaconess Medical Center (BIDMC) for a diagnostic evaluation and medication adjustment. He was given 1-mg doses of Ativan ([double dagger]) and 8- to 10-cc doses of paraldehyde to help control his increased frequency of seizures.

During the admission, CB continued to take 270 mg of phenobarbital per day, and his Tegretol dose was reduced to 2,200 mg/d because his Tegretol level was found to be toxic. During his hospital stay, CB was diagnosed with probable Rasmussen encephalitis, a chronic inflammatory encephalitis. Rasmussen encephalitis is characterized by increasing hemiparesis, mental retardation, hemianopsia, and seizure frequency. (10) Since the disease uniformly progresses to involve an entire cortical hemisphere and does not spread to the unaffected hemisphere, hemispherectomy is often an effective intervention. BIDMC's neurosurgical team thought that the procedure could be a successful option for CB. Because he was having fewer seizures in the first few days after admission, they planned to complete CB's presurgical testing as an outpatient and have him return at a later date for an elective functional hemispherectomy.

Before the planned discharge, physical therapist examination and evaluation (details provided below and in the Table) indicated that CB was not capable of living safely in his group home. During this time, he began to have an increasing number of seizures that required increased phenobarbital (330 mg/d) and Tegretol (2,600 mg/d) doses with daily, additional doses of Ativan and paraldehyde. These drugs failed to control the seizures, leading to a decision to perform the hemispherectomy during this hospital admission.

Preoperative testing included an electro-encephalogram, which confirmed epilepsia partialis continua. The recurrent spike and slow wave complexes characteristic of seizure activity were present continuously throughout the 7-minute recording and involved broad regions of the right hemisphere. Magnetic resonance imaging (MRI) demonstrated increased white matter signal in the right hemisphere, with extensive encephalopathic and prior operative changes. Sodium amytal was injected into the right internal carotid artery to temporarily suppress function of the right hemisphere. This test revealed left hemispheric dominance for language, indicating a good prognosis for postoperative speech capabilities.

CB underwent a right functional hemispherectomy approximately 1 month after being admitted to BIDMC. During the procedure, the surgeon removed all remaining white matter in the patient's right hemisphere except for small frontal and occipital poles. The entire corpus callosum, including the splenium and rostrum, was cut to eliminate the chance of epileptic discharges spreading from the subfrontal or deep areas to the opposite hemisphere. The basal ganglia remained intact, but all connections from the cortex to subcortical tissue were severed. The pathology report of right hemisphere brain tissue revealed tissue changes consistent with the diagnosis of Rasmussen encephalitis. The patient was observed overnight in the postanesthesia care unit and then was transferred to the acute neurosurgical unit.

Hospital Course

Preoperative Status

A physical therapist interviewed CB's parents to determine his baseline level of function and the level of independence necessary for his group home. They reported that prior to admission their son had been able to walk without physical assistance despite his hemiplegia and that this level of mobility was required for his group home. CB's balance and his level of attention were identified as the areas that had worsened most in recent weeks.

During the physical therapist's preoperative bedside examination (Table), it was noted that CB was inattentive to testing and demonstrated short-term memory deficits. He had decreased safety awareness, indicated by repeated attempts to perform activities beyond his ability without assistance (eg, walking alone, which he was unable to do since he was admitted to the hospital). CB was able to transfer independently from a supine position to a sitting position and from a sitting position to a standing position. He was able to walk 15.2 m (50 ft) without an assistive device, but with occasional assistance from the physical therapist to prevent a fall. When observing his gait, the therapist noted it was characterized by circumduction of his left lower extremity (LLE) during swing phase, with placement across the midline, and excessive displacement of his trunk to the left during stance. He frequently lost his balance to the right while walking. The onset of his righting reactions appeared to be delayed, and the amplitude of the response often appeared to be ineffective for correcting loss of balance.

CB's right upper and lower extremities (RUE and RLE) had intact sensory integrity and intact motor control and muscle performance (Table). His left upper extremity (LUE) twitched constantly due to seizures and rested with his forearm pronated, elbow extended, and shoulder medially (internally) rotated and adducted. With attempts to contract his LLE extensors individually, CB could generate a synergistic movement that included the combination of ankle plantar flexion and inversion; knee extension; and hip extension, medial rotation, and adduction. During functional activities, CB could flex his hip and knee through the full range of motion (ROM) against gravity, but he could not withstand any resistance to these movements. He could generate only slight contractions in his distal LLE muscle groups. Light touch, proprioception, and sharp/dull discrimination were tested to determine sensory integrity, revealing no impairment on the right side, but frequent incorrect answers on the left side.

Postoperative Days 3 to 5

Functional status and impairments. The physical therapist re-examined the patient on postoperative day (POD) 3 (Table). CB was positioned supine in bed, was somnolent, and was difficult to arouse with stimulation. He appeared, however, to have no seizure activity. Strong verbal, tactile, and visual cues were needed to maintain his attention during the examination, and he frequently did not answer simple questions. CB did not cough spontaneously or on command. He did not respond to noxious stimuli on the left side. His left side was flaccid except for minimal biceps muscle spasticity (ie, velocity-dependent resistance to passive stretch) that was graded 1 on the modified Ashworth scale, a measure of upper-extremity spasticity. (22) During a transfer from a side-lying position to a sitting position, he required directions on how to perform the task and complete assistance to move his LLE over the side of the bed. Once seated, CB was able to reach about 12.7 cm (5 in) with his RUE in all directions and remained seated for 10 minutes with supervision.

At this point, CB was still taking phenobarbital (330 mg/d) and Tegretol (2,400 mg/d), but his physician had also prescribed Dilantin ([dagger]) (400 mg/d) and Ativan (1.5 mg/d). In addition, he was taking codeine for pain and Decadron ([section]) for inflammation.

Prognosis and plan. Because CB was an adult and had a large amount of brain tissue removed, we thought the prognosis for return to preoperative level of function was poor. We believed the potential for CB to regain functional use of his left extremities was limited and, therefore, independent ambulation was unrealistic. Although we thought he might have some minor return of his sensorimotor function, we expected he would mainly have to compensate for those impairments. For these reasons, we planned to focus on instruction in compensatory strategies, such as wheelchair mobility training. Because we were uncertain about the possibility of sensorimotor return on the left side, our plan of care also included interventions to attempt to maximize this potential. We recommended discharge to a rehabilitation hospital to continue this intervention after his acute hospitalization.

Initial goals of physical therapy focused on maintenance of full passive range of motion (PROM), protection of his flaccid left shoulder, facilitation of bed mobility and transfers, improvement in sitting balance, protection of his airways, and retention of appropriate hemodynamic responses to changes in position. Because atelectasis and aspiration are well-demonstrated risks in patients who are sedated and somnolent and have neurological impairment (particularly those with loss of gag and swallow reflexes), CB was given an incentive spirometer and instructed in its use with assistance from his nurse and parents. (23) The physical therapist also instructed the patient and his caregivers about frequent position changes, including the method and importance of transfer to a sitting position for 15- to 20-minute periods several times each day.

During physical therapy, CB also practiced anticipatory and reactive balance exercises in a sitting position to prepare him to perform seated ADL. He reached in all directions for objects with his RUE and tried to move farther out of his base of support to grasp the objects. He practiced voluntary sways and weight shifting in all directions, and the therapist nudged him lightly on the thorax in all directions without warning in an attempt to stimulate righting reactions.

The LUE and LLE joints were passively taken through the full ROM in all directions by the physical therapist while CB was in the supine or side-lying position, and CB was taught rolling and supine-to-sitting transfers. He was instructed in and practiced components of these skills individually (eg, reaching with his arm to initiate rolling or moving his lower extremities over the side of the bed when in a side-lying position to prepare to transfer to a sitting position) and then in sequence to complete the task. As he improved his ability to perform these skills, we decreased the amount of instruction and feedback that we provided.

The physical therapist and the occupational therapist issued resting splints for the patient's left ankle and left hand because we believed they were necessary to prevent contractures while he had no active movement and therefore could not move his joints through the ROM voluntarily to preserve this motion. All details of the examination, prognosis, and plan of care were discussed with CB, his parents, and the rest of the team, including the occupational therapist, the case manager, the social worker, the nurse, and the physician.

Postoperative Days 5 to 8

Functional status and impairments. Within a week of his surgery, CB was able to participate in conversation and answer questions appropriately (Table). His verbal and physical responses to questions and instructions were delayed by 2 to 3 seconds, and he continued to require repeated verbal redirection to attend to tasks that were practiced during physical therapy sessions, such as bed mobility, transfers, and balance activities. He also required re-explanation of the rationale and goals of each therapy session and firm direction to decrease his agitation. His PROM remained normal, but he still had no spontaneous movement on his left side. CB still required assistance to move the left side of his body during rolling and to transfer to a sitting position. He was able to sit for 20 minutes at the edge of the bed with supervision and was able to reach approximately 17.8 cm (7 in) out of his base of support with his RUE and withstand more forceful perturbations in all directions without losing his balance. He was able to return to midline independently when moved out of the upright position.

CB was able to transfer out of bed for the first time since surgery on POD 5. He performed a stand-pivot transfer from the bed to a chair on his right side. This required substantial assistance of the physical therapist to help him rise to a standing position and to maintain his balance as he turned. CB's LLE did not support any of his weight during the transfer. He continued to have no evidence of seizures, and the physicians reduced his Tegretol dose to 1,600 mg/d.

Prognosis and plan. Although we became more optimistic about CB's ability to regain the ability to perform mobility and ADL tasks with less physical assistance, we still believed that he did not have the potential to walk functionally and predicted that he would rely on a wheelchair for most of his locomotion and ADL. We continued to discuss the need for intensive physical therapy after discharge from the acute care hospital with the social worker and CB's insurance case manager. We recommended transfer to an acute-level rehabilitation facility after his hospitalization.

The initial goals for maintenance of full PROM and protection of the left shoulder were still appropriate. Additional goals were added after the re-examination on POD 5 to address CB's standing balance and sit-to-stand and bed-to-chair transfers. Much to our surprise, a second update of the plan of care was necessary on POD 6 when CB began to walk. He said that he wanted to go into the bathroom rather than use his bedside commode and, with significant assistance from both the physical therapist and occupational therapist, managed to walk the 3 m (10 ft) to the lavatory. With help to shift his weight onto his RLE and initiate the LLE movement, CB could then bring his left leg forward and complete the swing phase. He needed physical assistance to prevent the left knee from buckling during stance. Nonetheless, CB was able to activate his LLE muscles during ambulation, which was progress from the previous day's examination when he showed no LLE muscle activation.

In the following days, CB continued to practice mobility. He performed several transfers from supine to sitting to standing positions during each treatment session while we increased the length of time from approximately 5 to 15 minutes that CB stood and decreased the amount of assistance we provided. Gait training was continued without an assistive device because the use of a hemi-walker caused CB to walk more slowly and focus primarily on moving the device, which resulted in a less fluid and less stable gait. We discussed the possibility of obtaining an AFO with CB and his parents, who reported that he had an AFO at home due to his preoperative foot drop, but that he had always refused to wear it. We decided to defer any decision about an orthosis to CB's physical therapist in the rehabilitation hospital. Because his parents would be closer to his home at that time, they could take CB's AFO for examination and possible use. CB also would have more time to maximize his motor return before deciding on the most appropriate orthosis for long-term use.

CB continued to receive PROM for his LUE to prevent contractures. He stopped wearing the resting ankle splint because we believed that his standing activities provided sufficient stretching to prevent a decrease in the ROM of his left ankle. We recommended continued use of the resting hand splint because upper-extremity motor function had not returned. Because a Harris hemisling is able to hold the flaccid shoulder of a patient with hemiplegia in good anatomical alignment when compared with the patient's uninvolved shoulder, (24) CB also was provided with a sling of this type to wear during transfers and gait in an attempt to protect his left shoulder and decrease the risk of subluxation.

CB's treatment at this time focused on mobility training because it prepared him for return home, he was motivated to practice it, and it was the only activity that stimulated contraction of the LLE muscles. Proprioceptive neuromuscular facilitation (PNF) has been shown to improve gait speed and cadence in patients with hemiplegia of both short and long duration, (25) so we used these techniques during CB's gait training. We also incorporated PNF techniques such as manual contact, stretch, and approximation (26) during his transfer training and standing balance activities (eg, in a standing position, downward pressure was applied on the patient's pelvis while manual contact was applied to the gluteus maximus muscle in an attempt to facilitate hip extensor activation, upright posture, and improved balance). Although BC had demonstrated LLE motor recruitment only during the automatic movements of walking, we hoped that the use of these techniques would promote a return of the use of the LLE during other activities such as transfers and standing.

During this time, CB's right hand twitched at times, but he was able to taper off the Dilantin and the Ativan. He continued to take the phenobarbital and Tegretol, as he had before surgery, and he took codeine for his persistent headaches.

Postoperative Days 11 to 14

Functional status and impairments. By POD 11 (Table), CB's responses to questions and commands were only slightly delayed. His sensory integrity and spasticity were unchanged, but, for the first time, CB purposefully moved his left leg when not walking. Although CB could not perform voluntary, isolated contractions of his LLE muscles when asked, he moved his LLE over the side of the bed to sit up without assistance or cues. He was able to stand and maintain his stance for approximately 10 seconds with only the therapist's hand on the gait belt. He could not move out of his base of support or withstand any perturbations in standing, but he could ambulate 38 m (125 ft) with the assistance of 2 therapists to support his weight and compensate for his lack of balance. The left foot barely cleared the floor because of his foot drop, his left knee buckled or hyperextended during the stance phase (each deviation occurred approximately 50% of the time), and he demonstrated a Trendelenburg sign on the left because of gluteus medius muscle weakness. CB noticed his improvements, said he was pleased with his progress, and participated in physical therapy enthusiastically.

Prognosis and plan. At this stage, CB moved his LLE during transfers and gait, was making gains in mobility, and participated fully in 45- to 60-minute weekday physical therapy sessions. In light of these improvements, we thought his potential for functional ambulation was much greater than we originally anticipated. The physical therapist continued to advocate for acute-level rehabilitation. The occupational therapist agreed with the recommendation for ongoing intensive therapy as there had also been demonstrable gains in CB's ability to perform purposeful activities such as feeding and dressing. Because of CB's altered prognosis, the anticipated goals and desired outcomes had shifted. Rather than teaching multiple compensatory techniques to maintain independence in the face of severe impairment, we planned to assist CB to regain his preoperative level of function.

We updated CB's goals to include performing all transfers, standing, and walking with less support of his body weight and assistance to maintain his balance from the physical therapist. In addition, CB wanted to stand for longer periods of time and walk longer distances. We continued patient and caregiver education about correct positions and safety measures to protect CB's hemiplegic shoulder (eg, avoidance of pulling on his arm to assist him to transfer or roll), the amount and type of assistance that he needed for mobility, and the need to inspect his skin to ensure there was no breakdown since he had lost protective sensation on his left side and therefore might not feel areas of pressure. During the last several days of his acute care stay, CB made gains in motor control and mobility. He performed isolated contractions of his proximal LLE muscles on request (ie, not in the context of a functional task) for the first time on POD 12. The following day, he moved his LUE voluntarily into shoulder abduction and lateral (external) rotation, elbow flexion, and forearm supination. In addition, CB was able to ambulate longer distances during physical therapy and more fully participate in bathing and dressing activities with the nurses and during occupational therapy sessions. Because CB could voluntarily contract his muscles, he was given a strengthening program to do several times each day. It consisted of hook-lying hip extension, supine active-assisted hip abduction and adduction, sitting hip flexion, knee extension, shoulder flexion, and elbow flexion. CB completed approximately 8 to 10 repetitions of each exercise, which caused fatigue of the muscles, as indicated by decreased control of the movement on the final repetition. He also continued to perform self-ROM with his LUE according to instructions provided to him by the occupational therapist. We provided written instructions, posted reminders in his room, and asked his nurses and parents to assist him with this program 3 times each day. The patient was challenged to perform transfers and gait with less assistance and fewer verbal cues. We also began to practice functional tasks in more complex environments, such as walking in the narrower spaces of his room and with the distractions present in the hospital hallway.

Discharge Status: Postoperative Day 14

CB was alert and participating fully in therapy by POD 14 (Table). He continued to be somewhat distracted in busy environments but focused on tasks well in controlled settings. He had developed spasticity in the flexors and extensors of his LUE (modified Ashworth scale grade 2). (22) He had regained the ability to perform isolated, voluntary muscle contractions in his left hip and knee, but he could not contract his ankle or toe muscles. At discharge, CB was able to ambulate 121.9 m (400 ft) with the physical assistance of one person to help him maintain his balance. He continued to have gait deviations due to his left hemiparesis and sensory deficits. His ability to participate in self-care tasks also improved, and he completed ADL tasks such as bathing, dressing, and feeding each day during occupational therapy sessions with assistance and initial cueing to focus on the activities.

Outcomes

CB's improvement necessitated regular updates of goals during his l 1-day rehabilitation course. In this short period, he achieved almost the level of function demonstrated on admission to BIDMC. He had excellent postoperative seizure control on essentially the same drug regimen he was taking when admitted. This seizure reduction, along with his improvements in mobility and ADL, made it possible to update his long-term plan from a fully assisted living situation with questionable mobility status to full-time community ambulation and return to his group home. To facilitate progress toward these outcomes, CB was discharged to an acute rehabilitation facility on POD 14. At the rehabilitation hospital, CB continued to improve. He eventually became able to ambulate independently in the community and returned to the group home where he lived prior to the surgery. As he did before surgery, he receives continued support from his parents and assistance with high-level ADL tasks, such as money management. At a meeting 6 months after surgery, CB reported that he was working at a fast-food restaurant where his job responsibilities include bussing dishes. He also described his frequent outings to sing karaoke with his girlfriend, which was not something he had enjoyed prior to surgery. He has right-sided facial twitching several times a week, but his seizures do not interfere with his daily life, and he is able to take lower doses of most anticonvulsants than he required before the surgery (1,600 mg of Tegretol, 330 mg of phenobarbitol, and no Ativan or Neurontin).

Discussion

Our frequent examinations, evaluations, and goal updates with CB illustrate the difficulties we had establishing an accurate prognosis for this patient. Because a substantial section of CB's cortex was removed when he was an adult, we expected him to have profound and lasting impairments and functional limitations. CB demonstrated rapid recovery of lower-extremity motor control and mobility during his 14-day, postoperative acute care hospitalization. He also had a marked reduction in the frequency of his seizures.

Several factors, may have contributed to CB's recovery after his hemispherectomy: subcortical control of movement and locomotion, (27-37) intact ipsilateral cortical pathways, (38-45) the young age at which he had his initial brain injury, (46-52) and the remodeling of his nervous system after each injury in a series of previous lesions. (53-55) Appreciation of these factors would have changed our prognosis and the focus of our treatment plan from compensatory training to facilitation of the return of sensorimotor function and early transfer and gait training.

CB was able to use his LLE to walk several days before he could generate voluntary movement of the limb during a less automatic task. His ability to ambulate at such an early stage may have been due to the contributions of subcortical regions of the nervous system. Several structures in the nervous system other than the cortex are believed to help produce coordinated movement; these structures include the cerebellum (27,28) and the mesencephalic locomotor region of the brain stem. (29,30) In addition, recent research indicates that the lumbosacral spinal cord contributes to the ability to walk in animals and humans. (31-36)

This case illustrates how gait training may be useful even in the absence of voluntary, isolated muscle contractions in the lower extremities. Perhaps the physical therapist's initial passive movement of CB's LLE in standing stimulated the central pattern generators of his spinal cord to produce reflexive motions in the left leg, allowing CB to actively step in absence of voluntary LLE movements. Supraspinal input from the motor cortex, cerebellum, and basal ganglia allows adaptation of locomotion to different tasks and environmental demands. Similarly, adequate posture and balance relies on supraspinal input from the vestibular system. (37) Early in his recovery, CB was able to initiate upright activities, but he had difficulty with the modification of his gait pattern, as well as with postural control during ambulation. It is possible that CB's vestibular system provided input to his axial and limb muscles to reflexively activate his extensor muscles, contributing to his ability to stand. His difficulty with gait and postural control may have been related to increased reliance on spinal levels of locomotor control and reflexive stepping and relatively little supraspinal input during the early postoperative period.

CB also may have regained left-sided motor control through cortical input from portions of the intact left hemisphere. Two pathways can transmit cortical input to ipsilateral muscles: the corticoreticulospinal tract and the ipsilateral corticospinal tract (Figs. 1 and 2). Both of these pathways control proximal muscles to a greater extent than they control distal muscles. Thus, our observations of CB's relatively strong shoulder and hip muscles, together with his persistent foot drop and inability to perform individual finger movements on the left side, are consistent with the hypothesis that he was using these ipsilateral pathways.

[FIGURES 1-2 OMITTED]

Recent studies provide evidence that individuals rely on ipsilateral input for control of movement after cortical resection. Wieser et al (38) described a patient who had a right precentral and postcentral resection at age 6 years as a treatment for Rasmussen encephalitis. Although the patient had left hemiplegia, she could walk independently after her operation, and she had no seizures for 17 years. Due to a subsequent recurrence of seizure activity, however, the patient had a [H.sub.2][sup.15]O positron-emission tomography (PET) scan, intracarotid amytal testing, and sensory-evoked potentials to evaluate the appropriateness of a functional hemispherectomy. These tests indicated that the hemiplegic left leg was relying on ipsilateral corticospinal, corticoreticulospinal, and spinocortical pathways and that its sensorimotor function would therefore not decline with more extensive surgery. This proved to be true; a hemispherectomy was performed, and it did not affect the patient's impairments and functional limitations.

Wieser et al (38) described the transfer of left leg motor control to the left primary motor cortex in a patient with a right-sided surgical resection. This finding is inconsistent with the research of Muller et al. (39) which suggests that control more frequently shifts to cortices adjacent to the ipsilateral primary motor cortex rather than to the ipsilateral primary motor cortex. Graveline et al (40) used functional MRI to reveal transfer of upper-extremity cortical control to contralateral associative sensorimotor areas in 2 patients who had previously had hemispherectomies (at ages 9 and 17 years, with brain injuries at birth and 2 years of age, respectively). The results of other studies (41-44) support the findings that intact nonprimary motor and somatosensory areas of the cortex are responsible for residual function after hemispherectomy.

Beneke et al (45) also studied these ipsilateral cortical pathways in individuals who had hemispherectomies at a young age and in patients who had strokes as adults. The authors compared these groups, examining the effect of age at time of injury on neural reorganization and outcomes. They also compared data from the 2 groups with data from previous studies of people without neurological problems. In the studies of people without neurological problems, magnetic brain stimulation had been found to set up a fast-descending impulse in the corticospinal tract, resulting in responses only in muscles on the opposite side of the body. In contrast, all of the patients in Beneke and colleagues' study demonstrated responses in muscles on the same side as the stimulation. The group with early brain damage demonstrated responses in their proximal muscles that were equal in amplitude bilaterally, despite having had hemispherectomies. The group with late brain damage demonstrated longer latency and smaller-amplitude responses and much poorer motor outcomes than the group with early brain damage. Both groups had larger amplitude responses proximally than distally, corresponding to the decreased strength noted in distal muscles during clinical examination of these patients. The authors suggested that patients with early lesions are able to use a fast ipsilateral corticospinal or corticoreticulospinal pathway, allowing them to have extensive residual motor function. By comparison, patients with late lesions rely on slower and weaker pathways and have poorer functional outcomes.

Numerous animal studies also have provided evidence for extensive neural plasticity after hemispherectomy. (46) Rather than simply concluding that early lesions produce fewer deficits than late lesions, however, recent research on animals suggests a "critical maturational period" (46)(p632) related to the development of the nervous system of each species. Villablanca and Hovda (46) described the critical maturational period as a time frame during the development of the species when there is decreased vulnerability to brain injury. This period in development corresponds to the time when there is completion of peak nerve growth, but incomplete myelination of the neurons. This time frame also is associated with the presence of excessive synapses, which are subsequently slowly pruned away. From their work on cats with hemispherectomies, Villablanca and Hovda (46) suggested that the potential for sprouting new neurons and synapses and for preventing the normal loss of excessive neurons and synapses is maximal during the critical maturation period. Consequently, the impairments that result from a lesion during this period are far less severe than those that occur when a lesion is sustained before or after this time period. Schmanke and Villablanca (47) suggested that the critical maturational period in cats encompasses a time period from fetal day 55 to postnatal day 60.

To date, no data establish such a critical maturational period in humans. It is known, however, that the number of neurons in the human brain reaches a maximum by weeks 28 to 32 of gestation. (48) Myelination of the pyramidal tract is completed by the age of 3 years, (16) and synapse numbers peak between 2 months and 2 years of age, depending on the cortical region. (49-51) The pruning of excess neurons appears to continue into adolescence. (49,50,52) Villablanca and Hovda, (46) therefore, proposed that perhaps procedures such as hemispherectomy should be performed before the age of 3 years to maximize the ability of the brain to reorganize and recover after the surgery.

CB had his hemispherectomy at 28 years of age, far beyond the proposed critical maturational period, yet he showed excellent recovery. It is likely that his brain began its reorganization in response to the initial brain injury when he was 5 years of age. Although the myelinization of CB's pyramidal tract would have been completed by age 5 years, his nervous system would have had excessive neurons and synapses remaining at that age. Perhaps in response to the injury, CB's nervous system increased his reliance on unconventional neural pathways, such as the ipsilateral corticospinal and corticoreticulospinal tracts, reinforcing their strength and preventing the pruning of their connections.

In addition, it is likely that CB had good preservation of function because his accidental and surgical lesions occurred in stages. There is strong evidence that progressive operations result in fewer impairments than procedures that create a large lesion all at once. (53-55) Smaller, serial lesions produce less stress on the system during each episode, prevent depletion of endogenous growth factors that aid in the repair of the brain after injury, and allow for neuroplasticity and reorganization of structure and function after each injury. (53) It is possible CB's right hemisphere functions gradually transferred to his left hemisphere in response to his head injury and earlier surgeries. If that were the case, he may have simply returned to using these established pathways after his hemispherectomy.

We cannot determine how much of CB's recovery was spontaneous and how much was due to physical therapy and occupational therapy. In addition, we might argue that decreasing levels of anticonvulsants accounted for the recovery noted. CB was able to taper off Ativan and Dilantin by POD 6 and then maintain a constant medication regimen of Tegretol and phenobarbital, similar to what he was taking prior to admission. Elimination of the Ativan and Dilantin may have contributed to CB's improved function because these drugs can have negative effects on the motor system, such as weakness and ataxia. CB, however, continued to gain motor control and mobility even when his medications were not being adjusted. CB performed his first isolated, voluntary contractions of his LLE and LUE, for example, during a time when there were no modifications to his drug regimen (PODs 12 and 13). We therefore cannot attribute all of his recovery to decreasing levels of medication.

Future research may be able to determine the effectiveness of and optimal physical therapy interventions for patients who have hemispherectomies. Because the frequency of hemispherectomy is low, single-subject design studies would be appropriate. The limitations of this report are primarily related to the limited descriptive literature available regarding this procedure and its outcomes in adults to allow us to accurately plan for this patient and our subsequent reliance on post-case hypothetical analysis.

This case report presents a man with an unusual history who underwent a relatively rare procedure. The description provided highlights the need to consider neuroanatomy and plasticity of the neurological system to establish an optimal physical therapy plan of care based on an accurate prognosis.

Table.
Summary of BC's Examination Results Over the Course
of His Hospital Stay (a)

                 Preoperative                    POD 3

Cognition        [down arrow] arousal,           Very poor arousal,
                   attention, safety,              [down arrow]
                   short-term memory               comprehension

Sensation        [down arrow] light touch,       No response to noxious
                   pinprick, proprioception        stimuli on left
                   LUE and LLE

Motor function   Spasticity LUE>LLE, LUE         Flaccid LUE and LLE
                   rests in extensor
                   synergy with no
                   active movement, LLE moves
                   in extensor synergy

Balance          Mass movements and large        Sits statically and
                   sway with dynamic               reaches min distance
                   sitting activities,             out of BOS with
                   wide BOS for static             supervision,
                   stance and poor                 slouched posture
                   dynamic standing
                   balance

Transfers        Independent                     Rolls to left, side
                                                   lying to sit with
                                                   min-mod assist

Gait             15.2 m (50 ft) with min         Unable due to impaired
                   assist, LLE circumduction       cognition and severe
                   and left lateral lurch          headache

ADL              Min assist with basic ADL,      Dependent
                   mod cues to attend
                   and set up provided

                 POD 5                         POD 11

Cognition        Delay in responses,           [down arrow] delay to
                   distractible                  respond, distractible

Sensation        [down arrow] light touch,     No change
                   pinprick, proprioception
                   LUE and LLE

Motor function   No change                     LUE without movement,
                                                 LLE with proximal
                                                 muscle activation
                                                 with gait

Balance          Sits statically and           Withstands mod nudges
                   reaches mod distance out      in sitting, stands with
                   of BOS with supervision,      wide BOS and CG 10 s,
                   good posture                  unable to move out of
                                                 BOS or withstand nudge

Transfers        Independent rolling,          Supine [right arrow] sit
                   supine [right arrow] sit      with sit [right arrow]
                   and stand pivot to            stand with CG
                   right with mod assist

Gait             Unable                        3.8 m (125 ft) with
                                                 mid-mod assist of 2,
                                                 left Trendelenburg
                                                 sign, knee buckling
                                                 or hyperextension,
                                                 foot drop

ADL              Washed face with mod          Min assist for basic
                   cues for attention and        ADL with mod cues for
                   sequencing                    attention and set up
                                                 provided

                 POD 14

Cognition        Slight delay, distractible

Sensation        No change

Motor function   LUE moves against gravity,
                   flexor synergy, LLE isolated
                   movement, 2/5 hip and
                   knee, 0/5 ankle

Balance          Stands with wide BOS and CG
                   15 s, dynamic standing poor

Transfers        Independent supine [right arrow]
                   sit, sit [right arrow] stand
                   with CG

Gait             121.9 m (400 ft) with rain-mod
                   assist of one, no change in
                   gait pattern

ADL              Min assist for basic ADL with
                   initial cues to attend to task

(a) POD=postoperative day, LUE=left upper extremity,
LLE=left lower extremity, BOS=base of support,
min=minimal, mod=moderate, CG=contact guard,
ADL=activities of daily living.

* Novartis Ophtalmics Inc, 11695 Johns Creek Pkwy, Duluth, GA 30097.

([dagger]) Parke-Davis, Warner-Lambert Div of Pfizer Inc, 201 Tabor Rd, Morris Plains, NJ 07950.

([double dagger]) Baxter Healthcare Corp, Anesthesia & Critical Care, 95 Spring St, New Providence, NJ 07874.

([section]) Merck & Co Inc, West Point, PA 19486.

References

(1) Tinuper P, Andermann F, Villemure JG, et al. Functional hemispherectomy for treatment of epilepsy associated with hemiplegia: rationale, indications, results, and comparison with callosotomy. Ann Neurol. 1988;24:27-34.

(2) McKenzie KG. The present status of a patient who had the right cerebral hemisphere removed. JAMA. 1938; 111:168.

(3) Krynauw RA. Infantile hemiplegia treated by removing one cerebral hemisphere. J Neurol Neurosurg Psychiatry. 1950; 13:243-267.

(4) McKissock W. Infantile hemiplegia. Proc R Soc Med. 1955;46:431-434.

(5) Ransohoff J. Hemispherectomy in the treatment of convulsive seizures associated with infantile hemiplegia. Res Publ Assoc Nerv Ment Dis. 1954;34:176-195.

(6) White H. Cerebral hemispherectomy in the treatment of infantile hemiplegia: review of the literature and report of 2 cases. Confinia Neurol. 1961;21:1-50.

(7) Oppenheimer DR, Griffith HB. Persistent intracranial bleeding as a complication of hemispherectomy. J Neurol Neurosurg Psychiatry. 1966; 29:229-240.

(8) Rasmussen T. Hemispherectomy for seizures revisited. Can J Neurol Sci. 1983;10:71-78.

(9) Rasmussen T. Postoperative superficial hemosiderosis of the brain, its diagnosis, treatment and prevention. Trans Am Neurol Assoc. 1973; 98:133-137.

(10) Peacock WJ. Hemispherectomy for the treatment of intractable seizures in childhood. Neurosurg Clin N Am. 1995;6:549-563.

(11) Davies KG, Maxwell RE, French LA. Hemispherectomy for intractable seizures: long-term results in 17 patients followed for up to 38 years. J Neurosurg. 1993;78:733-740.

(12) Lindsay J, Ounsted C, Richards P. Hemispherectomy for childhood epilepsy: a 36-year study. Dev Med Child Neurol. 1987;29:592-600.

(13) Vera CL. Hemispherectomy, callosotomy, and frontal lobectomy for seizures. J S C Med Assoc. 1992;88:245-250.

(14) Villemure JG, Rasmussen T. Functional hemispherectomy in children. Neuropediatrics. 1993;24:53-55.

(15) Peacock WJ, Webby-Grant MC, Shields WD, et al. Hemispherectomy for intractable seizures in children: a report of 58 cases. Childs Nerv Syst. 1996; 12:376-384.

(16) Holthausen H, Strobl K. Modes of reorganization of the sensorimotor system in children with infantile hemiplegia and after hemispherectomy. Adv Neurol. 1999;81:201-220.

(17) Guide to Physical Therapist Practice. Phys Ther. 2001;81:9-746.

(18) Vining EPG, Freeman JM, Pillas DJ, et al. Why would you remove half a brain? The outcome of 58 children after hemispherectomy--the Johns Hopkins experience: 1968-1996. Pediatrics. 1997;100:163-171.

(19) Muller F, Kunesch E, Binkofski F, Freund HJ. Residual sensorimotor functions in a patient after right-sided hemispherectomy. Neuropsychologia. 1991;29:125-145.

(20) Cockerell OC, Rothwell J, Thompson PD, et al. Clinical and physiological features of epilepsia partialis continua: cases ascertained in the UK. Brain. 1996;119(pt 2):393-407.

(21) Berkow R, ed. The Merck Manual of Diagnosis and Therapy. 16th ed. Rahway, NJ: Merck & Co Inc; 1992.

(22) Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther. 1987;67:206-207.

(23) Slack S, Shucart,W. Respiratory dysfunction associated with traumatic injury to the central nervous system. Clin Chest Med. 1994;15: 739-749.

(24) Brooke MM, de Lateur BJ, Diana-Rigby GC, Questad KA. Shoulder subluxation in hemiplegia: effects of three different supports. Arch Phys Med Rehabil. 1991;72:582-586.

(25) Wang R-Y. Effect of proprioceptive neuromuscular facilitation on the gait of patients with hemiplegia of short and long duration. Phys Ther. 1994;74:1108-1115.

(26) Voss DE, Ionta MK, Myers BJ. Proprioceptive Neuromuscular Facilitation. 3rd ed. Philadelphia, Pa: Harper & Row Publishers Inc; 1985.

(27) Schultz W, Montgomery EB Jr, Marini R. Proximal limb movements in response to microstimulation of primate dentate and interpositus nuclei mediated by brain-stem structures. Brain. 1979;102:127-146.

(28) Mori S, Matsui T, Kuze B, et al. Cerebellar-induced locomotion: reticulospinal control of spinal rhythm generating mechanism in cats. Ann NY Acad Sci. 1998;860:94-105.

(29) Grillner S, Shik M. On the descending control of the lumbosacral spinal cord from the mesencephalic locomotor region. Acta Physiol Scand. 1973;87:320-333.

(30) Jordan LM. Initiation of locomotion in mammals. Ann NY Acad Sci. 1998;860:83-93.

(31) Dobkin BH, Harkema SJ, Requejo PS, Edgerton VR. Modulation of locomotor-like EMG activity in subjects with complete and incomplete spinal cord injury. J Neuro Rehab. 1995;9:183-190.

(32) Harkema SJ, Hurley SL, Patel UK, et al. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol. 1997;77: 797-811.

(33) Wernig A, Muller S, Nanassy A, Cagol E. Laufband therapy based on "rules of spinal locomotion" is effective in spinal cord injured persons. Eur J Neurosci. 1995;7:823-829.

(34) Dietz V, Colombo G, Jensen L. Locomotor activity in spinal man. Lancet. 1994;344:1260-1263.

(35) Nicol DJ, Granat MH, Baxendale RH, Tuson SJM. Evidence for a human spinal stepping generator. Brain Res. 1995;684:230-232.

(36) Barbeau H, McCrea DA, O'Donovan MJ, et al. Tapping into spinal circuits to restore motor function. Brain Res Rev. 1999;30:27-51.

(37) Kandel ER, Schwartz JH, Jessell TM. Principles of Neural Science. 3rd ed. East Norwalk, Conn: Appleton & Lange; 1991.

(38) Wieser HG, Henke K, Zumsteg D, et al. Activation of the left motor cortex during left leg movements after right central resection. J Neurol Neurosurg Psychiatry. 1999;67:487-491.

(39) Muller RA, Chugani HT, Muzik O, Mangner TJ. Brain organization of motor and language functions following hemispherectomy: a [(15)O]-water positron emission tomography study. J Child Neurol. 1998;13:16-22.

(40) Graveline CJ, Mikulis DJ, Crawley AP, Hwang PA. Regionalized sensorimotor plasticity after hemispherectomy fMRI evaluation. Pediatr Neurol. 1998;19:337-342.

(41) Bittar RG, Ptito A, Reutens DC. Somatosensory representation in patients who have undergone hemispherectomy: a functional magnetic resonance imaging study. J Neurosurg. 2000;92:45-51.

(42) Bernasconi A, Bernasconi N, Lassonde M, et al. Sensorimotor organization in patients who have undergone hemispherectomy: a study with [sup.15]O-water PET and somatosensory evoked potentials. NeuroReport. 2000;11:3085-3090.

(43) Holloway V, Gadian DG, Vargha-Khadem F, et al. The reorganization of sensorimotor function in children after hemispherectomy. Brain. 2000;123:2432-2444.

(44) Leonhardt G, Bingel U, Spiekermann G, et al. Cortical activation in patients with hemispherectomy. J Neurol. 2001;248:881-888.

(45) Beneke R, Meyer B, Freund H. Reorganisation of descending motor pathways in patients after hemispherectomy and severe hemispheric lesions demonstrated by magnetic brain stimulation. Exp Brain Res. 1991;83:419-426.

(46) Villablanca JR, Hovda DA. Developmental neuroplasticity in a model of cerebral hemispherectomy and stroke. Neuroscience. 2000;95: 625-637.

(47) Schmanke TD, Villablanca JR. A critical maturational period of reduced brain vulnerability to injury: a study of cerebral glucose metabolism in cats. Dev Brain Res. 2001;131:127-141.

(48) Rabinowicz T, de Courten-Myers GM, McDonald J, et al. Human cortex development: estimate of neuronal numbers indicate major loss during late gestation. J Neuropath Exp Neurol. 1966;55:230-238.

(49) Huttenlocher PR. Synaptic density in human frontal cortex: developmental changes and effects of aging. Brain Res. 1979;163:195-205.

(50) Huttenlocher PR, de Courten C. The development of synapses in the striate cortex of man. Hum Neurobiol. 1987;6:1-9.

(51) Rakic P. Specification of cerebral cortex areas. Science. 1988;241: 176-179.

(52) Rabinowics T. The differential maturation of the human cerebral cortex. In: Falkner F, Tanner JM, eds. Human Growth. Vol 3: Neurobiology and Nutrition. New York, NY: Grune & Stratton; 1988: 1235-1241.

53 Stein DG. Brain Repair. New York, NY: Oxford University Press; 1995:82-86.

(54) Robinson GA, Goldberger ME. The development and recovery of motor function in spinal cats, I: the infant lesion effect. Exp Brain Res. 1986;62:373-386.

(55) Robinson GA, Goldberger ME. The development and recovery of motor function in spinal cats, II: pharmacological enhancement of recovery. Exp Brain Res. 1986;62:387-400.

AL Bates, PT, MS, GCS, is Inpatient Clinical Specialist, Massachusetts General Hospital, Boston, Mass. She was Physical Therapist I, Beth Israel Deaconess Medical Center, Boston, Mass, when this patient was treated. Address all correspondence to Ms Bates at Massachusetts General Hospital, Physical Therapy Services, WACC 128, 15 Parkman St, Boston, MA 02114 (USA) ([e-mail]).

CC Zadai, PT, DPT, MS, CCS, FAPTA, is Assistant Professor, MGH Institute of Health Professions, Boston, Mass. She was Chief of the Rehabilitation Services, Beth Israel Deaconess Medical Center, when this patient was treated.

The authors acknowledge Kirsten Colton, OTR/L, who also cared for the patient in this case report, and Lily Dayan-Cimadoro, PT, DPT, MS, who provided clinical guidance during the patient's stay at Beth Israel Deaconess Medical Center. They also thank Alvaro Pascual-Leone, MD, PhD, for thoughtful discussions at the start of this project and Nancy Dookie for her assistance with manuscript preparation.

This article was submitted November 4, 2002, and was accepted January 11, 2003.