Multidimensional assessment of motor function in a child with cerebral palsy following intrathecal administration of baclofen.

Studies of the effects of intrathecally administered baclofen in patients with spasticity (defined as presence of a velocity-sensitive increase in resistance to passive stretch and clonus[1]) have elucidated baclofen's effects on impairments such as abnormal reflex activity, muscle spasms, and bladder dysfunction, especially for patients with spinal cord injury or multiple sclerosis (see Campbell et al[2] for a review). The drug's effect on voluntary movement and daily function, however, remains less clear. Some authors[3,4] argue that decreasing spasticity will not necessarily lead to improvement in function. Spasticity may even be a useful adaptation to disability when force-control capacity is impaired. In addition, anecdotal reports of decreased strength with baclofen use suggest that the drug could even lead to deterioration of functional skills.[5,6] Corcos and colleagues,[7] however, reported that the positive signs of spasticity can interfere with movement control. In a case report on a patient with traumatic brain injury, baclofen appeared to improve patterns of muscle activation and speed of voluntary movements.[8]

A recent report[9] documented that intrathecally administered baclofen, used for periods of up to 48 months, reduces spasticity as assessed by the Ashworth Scale in children with central nervous system (CNS) dysfunction such as cerebral palsy (CP). Improved upper-extremity function and activities of daily living were found for those children who were capable of self-care.[10] Another study of children with spinal cord injury who were nonambulatory suggested that intrathecal administration of baclofen was accompanied by improvements in sleep and reduction of spasms and caregiver burden.[11] Long-term effects on other aspects of mobility, such as gross motor function and independence in transfer activities, and on disability in daily life roles have not been reported for children with CP.

This case report describes how we systematically examined the relationship between spasticity and functional improvement in a child with CP. Changes in impairment, such as range of motion (ROM) and positive signs of CNS dysfunction, were examined in relationship to changes in functional performance. Our goal was to determine whether, following intrathecally administered baclofen, the child's voluntary movement improved, with carryover into functional activities and independence in daily life. We also attempted to explore any negative effects of baclofen. Short-term effects were examined against baseline and a double-blind placebo-controlled phase; long-term outcomes were based on open observations using measures similar to those used during the short-term observational period.

Case Description

Subject

The subject, an 11-year-old boy, developed spastic diplegia subsequent to premature delivery at 26 weeks in the presence of maternal eclampsia. He was selected for this case report because he has been followed the longest of any patient with CP receiving baclofen in our center. He and his parents gave informed consent for entering the protocol described in this report. At the beginning of the period of observation, the subject was able to walk for short distances with a walker and bilateral ankle-foot orthoses (AFOs), but he primarily used a manual wheelchair for community mobility. He was not receiving regular physical therapy, and no additional therapy was provided during the period of time covered by the report.

Procedures

Table 1 shows the assessments we performed during each of the seven phases of the period of observation. First, the subject was tested to determine his baseline function and suitability for baclofen by documentation that he had increased velocity-sensitive stretch reflexes upon rapid passive extension of the lower-extremity muscles (phase 1-baseline). The modal level of spasticity in his lower-extremity muscles, as assessed with the Ashworth Scale,[12] was reduced from 4 to 2 following lumbar intrathecal bolus injection of 75 [micro] g of baclofen in saline (phase 2-drug trial), and he had no allergic reaction or other complications. On the basis of this trial dosage, the subject and his family decided to proceed with pump implantation for chronic infusion of baclofen. A permanent subcutaneous drug pump(*) was implanted in the abdominal area for infusion of baclofen into the subarachnoid space via an indwelling catheter.[6] The subject received gradual adjustment of the initial baclofen dosage until spasticity was diminished as much as possible without hindering standing stability. This dosage-adjustment period (phase 3-dosage adjustment) lasted 1.7 months, after which the subject entered a double-blind placebo-controlled clinical phase (phase 4-drug/placebo).

Table 1.
Assessments Performed During Each Phase

                               Phase(a)
                                             2-Drug    3-Dosage
                               1-Baseline    Trial     Adjustment
No. of months after pretest                            1.7
Tests of impairment
  Ashworth Scale               +             +         +
  Spasms                       +             +         +
  Babinski reflex              +             +         +
  Clonus                       +             +         +
  Single-joint movements       +
  Multijoint movements
  Strength                     +             +         +
  Range of motion              +
Tests of function
  Gross Motor Function
    Measure                    +
Interview
  Dressing                     +
  Feeding                      +
Tests of disability
  Self-rating

                                4-Double Blind      5-Open Drug
                               Drug    Placebo    12 mo    24 mo
No. of months after pretest    2.2     3.8        12       24
Tests of impairment
  Ashworth Scale               +       +          +        +
  Spasms                       +       +          +        +
  Babinski reflex              +       +          +        +
  Clonus                       +       +          +        +
  Single-joint movements       +       +          +        +
  Multijoint movements
  Strength                     +       +          +        +
  Range of motion              +       +          +        +
Tests of function
  Gross Motor Function
    Measure                    +       +          +        +
Interview
  Dressing                     +       +          +        +
  Feeding                      +       +          +        +
Tests of disability
  Self-rating                                     +

                               6-No     7-0pen
                               Drug     Drug
No. of months after pretest    44.7     46.3
Tests of impairment
  Ashworth Scale               +        +
  Spasms                       +        +
  Babinski reflex              +        +
  Clonus                       +        +
  Single-joint movements       +        +
  Multijoint movements         +        +
  Strength                     +        +
  Range of motion
Tests of function
  Gross Motor Function
    Measure
Interview
  Dressing
  Feeding
Tests of disability
  Self-rating

(a) Phase 1=baseline (no-drug pretest), phase 2=drug trial (75 ? g of baclofen in saline), phase 3=dosage-adjustment period, phase 4=double-blind placebo-controlled clinical phase, phase 5=open drug trial (140-250 ? g/d of baclofen), phase 6=no drug (broken catheter), phase 7=open drug trial (200240 ? g/d of baclofen).

During this 2-month period, the pump was first filled with baclofen (drug 1 at 143.28 [micro] g/d) for several weeks, followed by substitution of placebo (drug 2) for several more weeks, based on double-blind random assignment. Following the double-blind phase, baclofen was administered continuously for 40 months at levels ranging from 140 [micro] [micro] g/d to an optimal level of 250, [micro] g/d (phase 5-open drug). Drug administration was temporarily interrupted when the catheter connecting the pump to the subarachnoid area of the spinal cord broke (phase 6-no drug). A new catheter was implanted, and the intrathecal administration of baclofen was reestablished at 200 ? g/d (phase 7-open drug) and gradually increased to 240 ? g/d, the continuing current level.

Outcome measures are divided into those that assess impairment in organ systems; functional limitations in mobility, dressing, and eating; and disability in daily life roles at home and in school or recreational activities, as defined by the National Center for Medical Rehabilitation Research (NCMRR).[11,13] Rater reliability was assessed only for the gross motor scale used, but all assessments were carried out by personnel experienced in use of the measures, typically the same person each time and at the same time of day. Persons performing any given set of assessments were unaware of the results obtained by others, and all assessments were masked as to the substance administered during phase 4.

Clinical Assessment of Reflex Impairments

Clinical tests of spasticity and other reflexes were administered by a nurse during each phase of the observational period. They included the Ashworth Scale (1 [no resistance to movement] to 5 [completely rigid in flexion or extension]),[12] quantification of spasms (0 [no spasms] to 4 [more than 10 spontaneous spasms per hour]),[14] briskness of the stretch reflex (0 [absent] to 5 [clonus]), and elicitation of the Babinski reflex in response to a tactile stimulus on the planter surface of the foot (present or absent).[15]

Kinematic, Kinetic, and Electromyographic Assessment of Impairment in Reflexes and Selective Motor Control

Constrained movements (single joint) of the elbow and ankle and unconstrained (multijoint) isotonic movements of the knee and elbow were assessed in a motor control laboratory. For each movement, kinematics (displacement, velocity, and acceleration) and electromyographic (EMG) activity in the appropriate agonist and antagonist muscles were recorded. The complete method for these assessments is described elsewhere.[7,16-18]

For the constrained elbow and ankle movements, the subject sat in a chair and performed single-joint active elbow flexion and extension or ankle dorsiflexion and planter flexion in a low-friction manipulandum, a metal bar to which the moving limb was attached. The elbow movements were performed over a distance of 54 degrees (for both flexion and extension) in a transverse plane, with the shoulder initially abducted at 90 degrees. The initial position for the elbow joint was 70 degrees for flexion movements and 110 degrees for extension movements (full extension=0 [degrees]). The ankle movements were performed in the sagittal plane, and the hip and knee joints were initially kept close to 100 degrees. The subject was asked to move over a distance of 15 degrees for both planter flexion and dorsiflexion. The initial position for the ankle joint was 110 degrees for dorsiflexion movements and 90 degrees for plantar-flexion movements (90 [degrees]=anatomical position).

For the unconstrained knee movements, the subject sat in his wheelchair and performed knee flexion movements over one target distance in the sagittal plane. The hip and the knee were initially positioned at 100 degrees (0 [degrees]=anatomical position), and the ankle was positioned at 110 degrees. The unconstrained elbow movements were also performed in the sagittal plane. The subject was seated in a wheelchair and performed elbow flexion movements over 36 degrees. For these movements, the elbow was initially placed at 90 degrees, with the shoulder in the neutral position. The subject performed 11 trials in each direction for each voluntary isotonic movement.

Strength

Maximal voluntary contraction (isometric MVC) was assessed in both upper and lower extremities with a torque transducer attached to the manipulandum. The manipulandum was fixed at a position corresponding to 90 degrees of flexion at the elbow joint or to 110 degrees of planter flexion in the ankle joint. Three trials were used for each isometric contraction.

Range-of-Motion Impairment

Physical therapists assessed passive ROM with the subject in the supine position. Range of the following motions was assessed with a plastic universal goniometer: hip flexion and adduction, knee flexion,([dagger]) and ankle planter flexion. Hip flexion was tested with the opposite thigh flexed against the abdomen; knee flexion was assessed with the hip at 90 degrees of flexion. Hip adduction and ankle planter flexion were assessed with the hips and knees extended. In each case, 0 degrees equals the anatomical position. Elongation of the muscles was performed slowly by one therapist to avoid eliciting stretch reflexes. The muscles were held at their greatest length for 2 to 3 seconds before the measurement was made by a second therapist. Because reliability of ROM assessments in spastic muscles is problematic[19-21] and was not assessed for this case report, only changes of at least 10 degrees or more were considered clinically meaningful. The choice of 10 degrees was made based on Stuberg and colleagues' report.[20] We used the same methods for assessing the lower-extremity muscle groups, and the 95% confidence intervals for intrarater measurements ranged from [+ or -] 0.9 to 7.6 degrees.

Functional Limitations

Motor skills were assessed by a physical therapist using the Gross Motor Function Measure (GMFM). The GMFM was designed to assess gross motor function in children with CP and has been validated for sensitivity to change in children with CP who are undergoing physical therapy.[22] The test has 88 items, which are each scored on a scale of 0 (does not initiate activity) to 3 (completes activity). The items are divided into five separate scales: lying and rolling; sitting; crawling and kneeling; standing; and walking, running, and jumping activities. Item ratings are summed for each section and transformed into percentage scores based on maximum possible performance for each subscale and for overall performance. A second physical therapist videotaped the test session and also rated performance on the GMFM. Exact agreement on individual item responses between two therapists across three test sessions ranged from 75% to 87%. Performances that led to discrepant ratings between therapists were reviewed on the videotape, and differences were resolved. Clinically meaningful change was defined as follows, based on GMFM validity studies that evaluated the functional importance of various average degrees of change[22]: no change or negative change=less than 1.4% change in GMFM score, small change=1.4% to 6.9% increase, medium change=7.0% to 24.5% increase, and large change=24.6% or greater increase.

Ratings of Self-perceived Change

After 8 months of continuous baclofen therapy (phase 5), the subject and his mother rated self-perceived change on a five-level ordinal scale (worse, no change, slight improvement, moderate improvement, or marked improvement). Changes in spasticity, stiffness, and strength and in function during walking, transfers, dressing, and manipulation were rated. Specific abilities in feeding and dressing were also ascertained by interview of the subject following each GMFM assessment.

Disability in Role Functions in Daily Life

Following all data collection in phase 4 and during phase 5 (following 1 and 2 years of baclofen therapy), the subject and his mother were interviewed with open-ended questions to identify any changes of particular importance to them in daily life, both positive and negative.

Outcomes

Clinical Reflex Assessments

Figure 1 illustrates the reduction of spasticity on the Ashworth Scale after 30 and 60 minutes during the phase 2 bolus injection and its return to baseline levels after 2 hours. Clinical changes in lower-extremity spasticity and other reflexes are summarized in Table 2. Modal levels of spasticity on the Ashworth Scale in lower-extremity muscles decreased from 3 or 4 during no-drug periods to 2 or 3 during all phases when the subject was receiving baclofen. Ashworth Scale ratings for the upper extremity did not change throughout the 2-year period and were consistently 2 or below, in keeping with the diagnosis of spastic diplegia. Spasm frequency and ankle clonus decreased during baclofen administration, from 4 to 0 and from 5 to an average of 2.5, respectively. During phase 3, the Babinski reflex was suppressed. The Babinski reflex returned when the catheter failed and persisted when drug administration was first reestablished in phase 7.

[Figure 1 ILLUSTRATION OMITTED]

Table 2.
Assessment of Impairment: Clinical Reflex Measures

                       Phase(a)
                                     2-Drug    3-Dosage
Reflex Measure         1-Baseline    Trial     Adjustment
Ashworth Scale mode
  (1-5)                4             2         2
Spasms (0-4)           4             0         0
Babinski reflex(b)     +             +         -
Clonus mean (0-5)      5             2.5       2.5

                       4-Double Blind      5-Open Drug
                       Drug    Placebo    12 mo    24 mo
Reflex Measure
Ashworth Scale mode
  (1-5)                2       3          2        2
Spasms (0-4)           0       4          0        0
Babinski reflex(b)     -       -          -        -
Clonus mean (0-5)      2.5     5          2.5      2.5

                       6-No
                       Drug     7-Drug
Reflex Measure
Ashworth Scale mode
  (1-5)                4        2 (3 for hips)
Spasms (0-4)           4        0
Babinski reflex(b)     +        +
Clonus mean (0-5)      5        2.5

(a) See Tab. 1 footnote for description of phases.

(b) positive sign (+)=present, minus sign (-)=absent.

Kinematic, Kinetic, and Electromyographic Findings

Constrained voluntary movements. Figure 2 depicts voluntary movement of the elbow into flexion (Fig. 2A) and extension (Fig. 2B) with use of baclofen (solid line) and a placebo (broken line) during the double-blind period (phase 4). Baclofen appeared to have a negligible effect on the speed of the elbow flexion movement as measured by peak velocity (Fig. 2A), but the acceleration trace appeared to be smoother with baclofen than with the placebo. The agonist muscle activity was also more phasic, with an absence of clonus and less muscle activity after 700 milliseconds, during baclofen administration than during placebo administration. During elbow extension (Fig. 2B), the subject speeded up his movement. Note that although the agonist and antagonist muscles were initially coactivated, this coactivity was followed by the main antagonist burst, enabling the subject to brake the movement in a shorter period of time than he was able to do during the time when baclofen was not administered.

[Figure 2 ILLUSTRATION OMITTED]

The data in Figure 3 show the kinematic and EMG activity in the double-blind phase (phase 4) during voluntary ankle dorsiflexion. During placebo administration (broken line), there was more oscillation at the end of the dorsiflexion movement, even though movement speed was similar to that observed during baclofen administration. There was also generalized clonus during placebo administration. During baclofen administration, however, the subject achieved a similar movement speed with smoother control by using a shorter burst of activity in the agonist (tibialis anterior muscle) and less antagonist (coleus, medial and lateral gastrocnemius muscles) activity. Similar but less pronounced results were observed during voluntary ankle planter flexion.

[Figure 3 ILLUSTRATION OMITTED]

Unconstrained voluntary movements. Figure 4 depicts the changes in voluntary knee and ankle movements performed during the time of the catheter malfunction (no baclofen, phase 6, broken line) and after its change (baclofen, phase 7, solid line). The subject was asked to perform extension only at the knee joint "as fast as possible." He was unable to isolate movement to the knee, so the ankle joint also moved. Baclofen did not seem to help him to dissociate movements of the two joints. Improvement in movement speed at the knee (focal) joint was not remarkable while baclofen was being administered, but there were smaller and more phasic amounts of agonist (quadriceps femoris muscle) and antagonist (biceps femoris muscle) EMG activity less muscle activity at the ankle joint, and less clonus. In the unconstrained elbow movement task, the movement at the focal joint was 20% faster during baclofen administration than during placebo administration, with less movement at the shoulder (nonfocal) joint. Less clonus and smaller and more phasic agonist (biceps brachii muscle) and antagonist (triceps brachii muscle) EMG activity were also observed during administration of baclofen.

[Figure 4 ILLUSTRATION OMITTED]

Strength. Figure 5 depicts MVC during elbow flexion (Fig. 5A) and extension (Fig. 5B) and the agonist and antagonist muscle activity with administration of baclofen (solid line) and the placebo (broken line) during the double-blind trial (phase 4). Torque was greater during placebo administration than during baclofen administration. This greater torque was achieved by generating larger amounts of both agonist and antagonist muscle activity. During the placebo phase, there was muscle activity typical of clonus, which resulted in torque fluctuations. Large phasic responses in the biceps brachii muscle, whether acting as agonist or antagonist, were reduced during baclofen administration, and the torque traces showed less dramatic fluctuations. During the placebo phase, the subject was unable to relax his muscles in the initial position, as can be observed by comparing the EMG activity of antagonist muscles in the 100 milliseconds before the onset of agonist activity. Similar results were obtained in MVC of the ankle in both planter flexion and dorsiflexion. Decreased strength was observed in all phases when baclofen was administered to the subject.

[Figure 5 ILLUSTRATION OMITTED]

Range of Motion

Table 3 presents the ROM data recorded during the assessment of impairment in the lower extremities. Comparisons in the table are relative to the baseline, using the criterion of 10 degrees or more to constitute a valid change; footnotes provide a comparison between drug and placebo in the phase 4 trial. Following the introduction of baclofen in the clinical trial (phase 4), ROM in hip extension bilaterally and in dorsiflexion of the left ankle improved by 10 degrees or more over the baseline values. During placebo administration, measurements similar to those at baseline were again found, except that right hip abduction was decreased (5 [degrees] with the placebo versus 16 [degrees] at baseline). Comparison of the clinical trial drug phase measurements with those obtained for the placebo showed that four ROMs increased by 10 degrees or more, two ROMs decreased by 10 degrees or more, and one ROM remained the same (ie, differed by less than [+ or -] 10 [degrees]) (Tab.3).

Table 3
Assessment of Impairment: Possive-Range-of-Motion Measures
(in Degrees)

                           Phase(a)
                                         2-Drug    3-Dosage
Measure                    1-Baseline    Trial     Adjustment
Right hip abduction         16
Leh hip abduction           11
Right hip extension        -25
Left hip extension         -40
Right knee extension       -68
Right ankle dorsiflexion   -18
Left ankle dorsiflexion    -22

                             4-Double  Blind    5-Open Drug
Measure                      Drug     Placebo   12 mo    24 mo

Right hip abduction          Same(b)   --       Same
Leh hip abduction            Same(c)   Same     Same      Same
Right hip extension            +(b)    Same      --        --
Left hip extension             +(b)    Same     Same      Same
Right knee extension         Same(c)   Same     Same      Same
Right ankle dorsiflexion     Same      Same      (*)      Same
Left ankle dorsiflexion        +(b)    Same      (*)       --

Measure                        6-No   7-Open
                               Drug   Drug
Right hip abduction
Leh hip abduction
Right hip extension
Left hip extension
Right knee extension
Right ankle dorsiflexion
Left ankle dorsiflexion

(a) See Tab. 1 footnote for description of phases. Plus sign (+) =improvement of 10 degrees or more over baseline, minus sign (-) =deterioration of 10 degrees or more over baseline, same=within 10 degrees of baseline, asterisk (*)=measurement could not be made according to the protocol because hip and knee extension to an anatomical position in the supine position could not be achieved.

(b) Better than on placebo by 10 degrees or more.

(c) Worse than on placebo by 10 degrees or more.

The improvement in ROM recorded during the double-blind drug phase was not maintained 1 year later. Relative to baseline measurements (but not relative to measurements for the placebo), right hip extension had worsened. After 2 years, ROM was worse relative to baseline measurements (but not relative to measurements for the placebo) in both hip extension and abduction on the right, and ROM was worse relative to baseline and placebo phase measurements in left ankle dorsiflexion. Right ankle dorsiflexion was worse after 2 years than when the placebo was administered to the child (baseline measurements were intermediate between drug phase and placebo phase measurements). The subject had experienced a growth spurt and attained puberty during this period. Orthopedic assessment of the hip joints and muscle contractures was recommended by the therapist who assessed ROM.

Functional Changes

Gross motor function. The subject's baseline GMFM score was 140 (54.4%), with the lowest scores in the standing and walking sectors. The highest item passed was ability to take eight steps with two-hand support. Clonus of the ankle muscles was evident whenever the subject put weight on his feet, even with minimal weight bearing in the seated position. Table 4 presents the changes in the GMFM scores over the 2-year period. In the double-blind phase (phase 4), the change in GMFM scores over baseline values while baclofen was administered (as revealed when the drug code was broken) was 1.2%, which we deemed to be not clinically important. Changes included a new ability to crawl backward down four steps, but the subject was unable to take any steps with two-hand support. The tester's notes indicated that the subject's trunk swayed noticeably during independent sitting items, and some clonus was still evident.

Table 4.
Assessment of Function: Gross Motor Function Measure (GMFM)

                           Phase(a)
                                        2-Drug   3-Dosage
Measure                    1-Baseline   Trial   Adjustment

GMFM raw score             140
Sector percentage scores
   Lying/rolling            92.16
   Sitting                  95.00
   Crawling                 69.05
   Standing                 12.82
   Walking                   2.78
Average percentage scare    54.4

                             4-Double Blind       5-Open Drug
Measure                      Drug       Placebo   12 mo   24 mo

GMFM raw score               142         143      159     156
Sector percentage scores
   Lying/rolling              94.12       96.08    98.04   98.04
   Sitting                    95.00       95.00    95.00   91.67
   Crawling                   76.19       76.19    76.19   76.19
   Standing                   12.82       12.82    30 77   25.64
   Walking                     0.00        0.00    11.11   12.50
Average percentage scare      55.6        56.0     62.2    60.8

                             6-No        7-Open
Measure                      Drug        Drug

GMFM raw score
Sector percentage scores
   Lying/rolling
   Sitting
   Crawling
   Standing
   Walking
Average percentage scare

"See Tab. 1 footnote for description of phases.

During administration of the placebo, the subject's GMFM score was 56%. The difference between scores obtained while baclofen was administered and scores obtained while the placebo was administered in the clinical trial phase was 0.4% in favor of the placebo and was not considered to be clinically meaningful.

During the 1-year follow-up (phase 5), the changes we noted on the GMFM were, in our opinion, clinically meaningful, registering a medium positive change of 7.8% above baseline values or a small positive change of 6.2% over the score obtained with administration of baclofen in the clinical trial. The major change occurred in areas that were previously the most difficult (de, in the standing and walking sections) (Tab. 4). Items accomplished that had not been observed at either the baseline or placebo phase measurements were the ability to cruise sideways while standing at a table, the ability to stand briefly on one foot while holding on to a table with only one hand, and the partial ability to squat from a standing position. The subject previously had been unable to initiate (score of 0) each of these activities. His ability to walk with two-hand support improved from 8 steps at the baseline measurement and 0 steps during administration of baclofen and the placebo in phase 4 to 10 steps at the 1-year follow-up. The subject discarded his AFOs because clonus during weight bearing was reduced. At 2 years after initiation of baclofen therapy, the GMFM score was 60.8%, a decrease of 1.4 points from the 1-year score of 62.2%. He continued to be able to walk 10 steps with two-hand support, but he could not cruise sideways at a table, pivot in a sitting position, or squat from a standing position as well as he had after 1 year of baclofen therapy. The 2-year score, however, continued to represent a small positive change over baseline and placebo phase scores.

Client's and Mother's Reports of Changes

Phase 4-randomized drug/placebo.

After all other data were collected in phase 4, the subject provided anecdotal comments to an interviewer regarding his perception of the immediate effects of baclofen use, including negative effects. Although everyone involved in phase 4 was unaware of whether baclofen or the placebo was given or when it was given, if a drug has effects, those persons receiving the medication often have perceptions reflecting their experience. We asked what these perceptions were in order to better understand any changes, positive or negative, from the subject's perspective. When the subject suspected he was receiving baclofen, he perceived his body to feel lighter and his walking to be improved, although stair climbing was somewhat more difficult to do at first because he had to use his own "muscle strength and couldn't depend on tone for using my legs." He thought his endurance for walking was somewhat decreased during the short-term drug trial, but he valued the fact that he could move more freely. A new activity he reported with pleasure was the ability to cross his legs in a sitting position without using his hands to position them. The subject's mother noted that during the periods that the subject believed he was receiving baclofen, he was able to sit up straight in a chair with his buttocks well back in the seat and transfers were easier to accomplish without assistance.

Phase 5-long-term outcome. The ordinal ratings of self-perceived change completed by the subject and his mother after 1 year indicated "marked" improvement in overall spasticity, stiffness, and strength; "moderate" improvement in walking and dexterity; and "marked" improvement in transfer activities and dressing. Increased ease of transfers because of improved lower extremity mobility was noted by the family to be the most important overall functional change. The subject described his movements as "feeling like my muscles are now working rather than resisting attempts to move."

Activities of daily living. The subject was independent in eating activities at the start of the period of observation, so no change was expected, but he reported that he became capable of pouring his own soft drink from a can. His ability to dress independently also improved during baclofen administration. Initially, he needed assistance with putting on socks, a polo shirt, and his underwear. While baclofen was administered (phase 4), he was immediately able to manage all of these pieces of clothing independently. While the placebo was administered, he again needed assistance with these articles of clothing. During long-term assessment (phase 5), the subject continued to dress independently, needing assistance only with buttons and shoelaces.

Disability in Daily Life Roles: Long-term Outcomes

When questioned about the presumed effects of baclofen that mattered most, for better or worse, the subject reported special pleasure in the new freedom to transfer in and out of a car from his wheelchair without assistance, to fasten his own seat belt, and to manage toilet hygiene independently. He also believed that when playing wheelchair basketball, his ball handling improved. The subject proudly reported at the 1-year follow-up that he had won an award in school for being most improved in physical fitness. He had completed a 5.2-m (50-yd) "dash" and walked 41.8 m (400 yd) with his walker to achieve this citation.

In summary, administration of baclofen appeared to produce an immediate reduction in spasticity, clonus, and coactivation of antagonists. Lower extremity movements were smoother, and speed and control of upper-extremity movements increased despite decreased strength. During the observational period, ROM increased in the hips and one ankle, and independence in eating, dressing, and transfer activities improved. After 1 year, gross motor skills showed a moderate increase, and the subject reported further increases in independence in car transfers and toileting, receipt of a fitness award, and discontinuation of orthotic device use. Range of motion, however, decreased for the right hip after 1 year and for the hips and left ankle after 2 years.

Discussion

Reduction of Impairment

During the period of observation, intrathecally administered baclofen appeared to reduce the subject's spasticity, as measured by the Ashworth Scale and the subject's self-report. Ankle clonus, resistance to passive movement, excessive coactivation of antagonist muscles during voluntary movement, and Babinski reflexes are common signs associated with spasticity in people with CP.[23] Although these signs decreased during baclofen administration (Tab. 2 and Fig. 3), movement speed at the lower-extremity joints was not altered. The subject was observed to generate more phasic bursts of activity for all muscles controlling elbow motion and to speed up movement during baclofen administration (Fig. 2). We believe that the changes in upper-extremity control may also be related to improved lower-body stability as a result of less irradiation of activity to lower-extremity muscles during arm use, suggesting the need for more exploration of how baclofen achieves its effects on upper-extremity activity. Diffusion of baclofen to the cervical cord, albeit in lower concentrations, occurs following lumbar injection, so the effects on upperextremity control could be direct, indirect (de, improved because of reduction of lower-extremity associated reactions), or both.[24] The subject also obtained more selective control of multijoint reaching movements while receiving baclofen. These observations support the idea that decreasing abnormal postural tone and reflexes can result in improved movement control, despite the presence of decreased strength as assessed by MVC.

The Relationship of Reduced Impairment and Functional Improvement

Other researchers[2,3] have argued that decreasing the positive signs of CNS dysfunction will not necessarily lead to improvement in functional performance, but we disagree and believe that this case report supports our view. The subject's decrease in muscle strength (Fig. 5) was not enough to result in overall deterioration in the performance of voluntary movements because the GMFM score stayed essentially the same after the introduction of baclofen (increases in a few skills offset by decreased number of steps with two-hand support) and differed little from performance during placebo administration. Our data suggest that improved patterns of muscle activation and control of voluntary movements occurred along with immediate improvements in transfer, dressing, and eating skills. We therefore provide some evidence in support of previous reports of improved upper-extremity function.[10]

The reversal of movement in conjunction with stretch reflex activity, as shown in Figure 3, supports the observation of Corcos and colleagues7 that clonus can provoke a reversal of direction in fast voluntary movements as a result of rapid reflex activation in an elongating antagonist muscle. Furthermore, reduction of ankle clonus and spasticity during baclofen administration was accompanied by smoother voluntary movements. The reduction of spasticity, clonus, and coactivation of antagonist muscles appears to relate to changes in the way the movements were controlled and may have contributed to the subject's functional improvements and enhanced quality of life. During baclofen administration, the subject was able to pour a soft drink into a glass from a can and reported improved ability to intercept a ball. By reducing clonus, generating more phasic bursts of EMG activity, delaying the activation of antagonists, and decreasing coactivation of muscles, the subject may have been more able to achieve the muscle force requirements of those tasks with improved control.

Although the use of baclofen did not appear to increase movement speed or selective activation of individual joint movements in the lower extremity (Fig. 4), it appeared to contribute to faster and better-controlled movements in the upper extremity (Fig. 2) and to decrease movement fluctuations in the lower extremity (Fig. 3). The changes in movement control may have led to the functional improvements reported by the subject and his mother, as well as the perceptions of his body that he described. We believe that the changes in movement control may have contributed to the subject's increased community mobility and independence in self-care and transfers at an age when children with CP frequently experience diminished physiologic fitness, mechanical efficiency, and ability to use ambulation for community mobility.25 Whether improved function might occur in other patients treated with baclofen or other antispasticity interventions cannot be predicted from a report on a single case. Controlled studies are warranted to further explore the linkage between changes in impairments related to motor control and functional outcomes of importance to the individual.

Effects of maturation or learning from repeated testing should be considered as possible alternative explanations for the changes observed over time. We believe that these explanations are unlikely to be valid, however, because many of the changes were reversed during no-drug periods. In addition, several of the changes observed on long-term follow-up that might be attributed to maturation were observed immediately after the initiation of baclofen treatment. The only major relapse was in ROM, which could be the result of the growth spurt that occurred at puberty and the fact that the subject discarded his AFOs. In addition, he increased his walking with reduced interference from spasticity and clonus, but also with slightly reduced strength, another factor that may have contributed to decreased ROM. Based on our clinical experience, we believe it to be unlikely that new functional skills will develop spontaneously in adolescence to the degree noted here. A reduction in mechanical efficiency is more typical.[25]

More optimal drug levels were gradually obtained over the first year of therapy, which may have allowed further functional improvements to occur. The change in the subject's GMFM scores from baseline testing to 1 year after the introduction of baclofen was 7.8%. In research on validity of the GMFM, ratings changed by an average of 2.3% over a 6-month period of physical therapy for children with spastic diplegia who were over 6 years of age.[22] Therefore, we consider the change in the subject's GMFM scores to be small to moderate but clinically significant. Scores on the GMFM have also been shown to be related to the physiological fitness of children with CP, as assessed by aerobic and anaerobic power.[26] Scores on the standing, walking, running, and jumping dimensions of the GMFM (those on which our subject showed the most change after 1 year) were related to peak anaerobic power and mean anaerobic power of the legs, whereas only low correlations were found with power in the arms. Although we did not measure fitness in our subject quantitatively, changes in the subject's GMFM scores in the standing and walking dimensions and his receipt of a fitness award in school suggest that his physiological fitness improved. In the future, wellcontrolled studies should assess this possibility with quantitative measures, such as perceived effort during transfers, mobility activities, and recreation, or with measures of actual oxygen consumption under standardized conditions. New assessment tools such as the Pediatric Evaluation of Disability Inventory[27] and the Canadian Occupational Performance Measure[28] would also be useful to better quantify changes in independence and self-selected goals.

The changes in our subject's GMFM scores can be compared with an average change in GMFM scores of 10% to 12% in much younger children 1 year after selective posterior rhizotomy.[29] In contrast, however, children experience decreases in strength immediately after rhizotomy and undergo intensive physical therapy (2 hours per day for 5 days for 1 month and 3-5 hours per week over the next 5 months) to regain presurgical levels of performance and go on to make functional gains. Studies are needed to compare rhizotomy and baclofen effects.

Negative Effects

In addition to decreased strength, which did not appear to hinder function after careful dosage adjustment, ROM decreased after long-term drug use. Although growth leading to decreased flexibility is an alternative explanation and measurements of passive ROM are known to be relatively unreliable in the presence of spasticity,[21] we consider the ROM findings to be a potentially negative effect of baclofen use. The subject discarded his AFOs as unneeded when his spasticity and clonus were reduced by baclofen, but the reduction in hip abduction ROM at the end of 2 years and the failure to maintain the short-term increases noted in hip and ankle ROM are of concern. Reports of complications in selective posterior rhizotomy for spasticity reduction have noted the occurrence of hip dislocation,[30,31] and the same risk may exist with baclofen use. Consequently, we have encouraged our subject to have regular orthopedic evaluations and to consult with his local physical therapist regarding the advisability of using night splints to maintain ROM. We did not recommend discarding the AFOs, and we believe that discarding the AFOs may not have been the best decision, even though our subject considered the freedom from orthoses to be an important outcome.

Recommendations

Crenna and colleagues[32] have demonstrated that the gait of children with CP can be impaired by various combinations of spasticity, muscular coactivation, paresis, and muscle hypoextensibility, which are unique to each child. In our subject, both spasticity and coactivation were apparently reduced during baclofen use. Whether children with coactivation as the primary problem might be helped by the drug is unknown but could be studied. We also recommend investigation of the combined effects of physical therapy and intrathecally administered baclofen. Therapy might assist children in making use of improved movement control and in counteracting weakness. Finally, we believe that our work demonstrates the need to perform motor control laboratory studies of kinematic, kinetic, and EMG activity to document the rather subtle changes in neuromotor control that can result in functional improvements and the perception of decreased effort. We also subscribe to the NCMRR recommendation that research measurement protocols should relate changes in impairment to changes in functional limitations and disability in daily life.[13] Clearly, more work is needed to develop an optimal protocol for assessing the effects of intrathecally administered baclofen on impairment, functional limitations, and disability in children with CP.

Acknowledgments

We thank Roni Cromwell, Kyung Lee, Thubi Kolobe, and Rich Conti for helping with parts of the data collection. Special thanks to Janet Gianinno and Michelle York for the outstanding services provided to the patient and for the clinical tests of spasticity, and to Gerald Gottlieb for advice on motor control testing and data analysis.

(*) Medtronic Inc, 7000 Central Ave NE, Minneapolis, MN 55432-3576.

([dagger]) Only right-knee ROM is reported here because of a recording error on the left side.

References

[1] Lance JW. Symposium synopsis. In: Feldman RG, Young RR, Koella WP, eds. Spasticity: Disordered Motor Control. Miami, Fla: Symposia Specialists; 1980:485-500.

[2] Campbell SK, Almeida GL, Penn RD, Corcos DM. The effects of intrathecally administered baclofen on function in patients with spasticity. Phys 7 hen 1995;75:352-362.

[3] Landau WM. Spasticity: the fable of a neurological demon and the emperor's new therapy. Arch Neurol. 1974;31:217-219.

[4] Sahrmann SA, Norton BJ. The relationship of voluntary movement to spasticity in the upper motor neuron syndrome. Ann Neurol. 1977;2:460-465.

[5] Penn RD. Intrathecal baclofen for spasticity of spinal origin: seven years of experience. J Neurosurg. 1992;77:236-240.

[6] Penn RD, Corcos DM. Management of spasticity by central nervous system infusion techniques. In: Youmans JR, ed. Youmans Neurological Surgery, Volume 5. 4th ed. Philadelphia, Pa: WB Saunders Co; 1996:3687-3700.

[7] Corcos DM, Gottlieb GL, Penn RD, et al. Movement deficits caused by hyperexcitable stretch reflexes in spastic humans. Brain. 1986;109;1043-1058.

[8] Latash ML, Penn RD, Corcos DM, Gottlieb GL. Effects of intrathecal baclofen on voluntary motor control in spastic paresis. J Neurosurg. 1990;72:388-392.

[9] Albright AL, Ceervi A, Singletary J. Intrathecal baclofen for spasticity in cerebral palsy. JAMA. 1991;265:1418-1422.

[10] Albright AL, Barron WB, Fasick MP, et al. Continuous intrathecal baclofen infusion for spasticity of cerebral origin. JAMA. 1993;270:2475-2477.

[11] Armstrong RW, Steinbok P, Farrell K, et al. Continuous intrathecal baclofen treatment of severe spasms in two children with spinal-cord injury. Dev Med Child Neurol. 1992;34:731-738.

[12] Ashworth B. Preliminary trial of carisoprolol in multiple sclerosis. Practitioner. 1964;192:540-542.

[13] National Institutes of Health. Research Plan for the National Center for Medical Rehabilitation Research. Washington, DC: US Dept of Health and Human Services; March 1993. NIH Publication No. 93-3509.

[14] Penn RD, Savoy SM, Corcos D, et al. Intrathecal baclofen for severe spinal spasticity. N Engl J Med. 1989;320:1517-1521.

[15] Latash ML, Penn RD, Corcos DM, Gottlieb GL. Short-term effects of intrathecal baclofen in spasticity. Exp Neurol. 1989;103;165-172.

[16] Almeida GL, Hong D, Corcos DM, Gottlieb GL. Organizing principles for voluntary movement: extending single joint rules. J Neurophysiol. 1995;74:1374-1381.

[17] Gottlieb GL, Corcos DM, Agarwal GC. Organizing principles for single joint movements, I: a speed-insensitive strategy. Neurophysiol. 1989;62:342-357.

[18] Almeida GL, Corcos DM, Latash ML. Practice and transfer effects during fast single-joint elbow movements in individuals with Down syndrome. Phys Ther. 1994;74:1000-1016.

[19] Harris SR, Smith LH, Krukowski L. Goniometric reliability for a child with spastic quadriplegia. J Pediatr Orthop. 1985;5:348-351

[20] Stuberg WA, Fuchs RH, Miedaner JA. Reliability of goniometric measurements of children with cerebral palsy. Dev Med Child Neurol. 1988;30:657-666.

[21] O'Dwyer N, Neilson P, Nash J. Reduction of spasticity in cerebral palsy using feedback of the tonic stretch reflex: a controlled study. Dev Med Child Neurol. 1994;36:770-786.

[22] Russell D, Rosenbaum P, Gowland C, et al. Gross Motor Function Measure Manual. 2nd ed. Hamilton, Ontario, Canada: McMaster University; 1993.

[23] Leonard CT, Hirschfeld H. Myotatic reflex responses of nondisabled children and children with spastic cerebral palsy. Dev Med Child Neurol. 1995;37:783-799.

[24] Kroin JS, Ali A, York M, Penn RD. The distribution of medication along the spinal canal after chronic intrathecal administration. Neurosurgery. 1993;33:226-230.

[25] Lundberg A. Longitudinal study of physical working capacity of young people with spastic cerebral palsy. Dev Med Child Neurol 1984;26:328-334.

[26] Parker DF, Carriere L, Hebestreit H, et al. Muscle performance and gross motor function of children with spastic cerebral palsy. Dev Med Child Neurol. 1993;35:17-23.

[27] Haley SM, Coster WJ, Ludlow LH, et al. The Pediatric Evaluation of Disability Inventory: Development Standardization and Administration Manual. Boston, Mass: New England Medical Center Publications; 1992.

[28] Law M, Baptiste S, McColl MA, et al. The Canadian Occupational Performance Measure: an outcome measure for occupational therapy. Can J Occup Ther. 1990;57:82-87.

[29] McLaughlin JF, Bjornson KF, Astley SJ, et al. The role of selective dorsal rhizotomy in cerebral palsy: critical evaluation of a prospective clinical series. Dev Med Child Neurol. 1994;36:755-769.

[30] Abott R. Complications with selective posterior rhizotomy. Pediatr Neurosurg. 1992;18:43-47.

[31] Greene WB, Dietz FR, Goldberg MJ, et al. Rapid progression of hip subluxation in cerebral palsy after selective posterior rhizotomy. J Pediatr Orthop. 1991;11:494-497.

[32] Crenna P, Inverno M, Frigo C, et al. Pathophysiological profile of gait in children with cerebral palsy. In: Forssberg H, Hirschteld H, eds. Movement Disorders in Children. Basel, Switzerland: S Karger AG, Medical and Scientific Publishers; 1992:186-198.

GL Almeida, PhD, PT, is Associate Professor, Instituto de Reabilitacao de Campinas, Universidade Estadual de Campinas, Campinas, Sao Paulo, Brazil.

SK Campbell, PhD, PT, FAPTA, is Professor and Director of Graduate Studies, Department of Physical Therapy, College of Associated Health Professions, University of Illinois at Chicago, 1919 W Taylor St (M/C 898), Chicago, IL 60612-7251 (USA) ([e-mail]). Address all correspondence to Dr Campbell.

GL Girolami, PT, is Director, Pathways Center for Children, Glenview, Ill.

RD Penn, MD, is Professor, Neuroscience Institute, Neurosurgery Division, Rush-Presbyterian-St Luke's Medical Center, Chicago, Ill.

DM Corcos, PhD, is Associate Professor, School of Kinesiology, College of Associated Health Professions, and Department of Psychology, University of Illinois at Chicago, and Department of Neurological Sciences, Rush-Presbyterian-St Luke's Medical Center.

The protocol was approved by the human subjects research committees of Rush-Presbyterian-St Luke's Medical Center (ORA 78101221) and the University of Illinois at Chicago (IRB H89-395).

Some of the data in this report were presented at the Presidential Symposium of the 1995 meeting of the American Academy for Cerebral Palsy and Developmental Medicine in Philadelphia, Pa.

This work was supported in part by Medtronic Inc (RDP), National Institutes of Health grants R01 NS 28176 and R01 AR 33189 (GLG, DMC) and KO4-NS 01508 and R01-NS 28127 (DMC), and US Public Health Service Maternal and Child Health Bureau grant MCJ IL 179590 (SKC).