The Challenges Faced by Today’s Clinical Researchers Measuring Effects of Treatment on Motor Function in Children with Cerebral Palsy
A Review of ongoing efforts supported by the NIH
The NIH Taskforce on Childhood Motor Disorders Met March 4-5, 2006 Terry Sanger, MD, PhD, Daofen Chen, PhD, Mauricio R. Delgado, MD, FRCPC, Deborah Gaebler-Spira, MD, Mark Hallett, MD, Amy Bastian, PhD PT, Max Wiznitzer, MD, Hilla Ben-Pazi, MD, Kristie Bjornson, MS, PT, PCS, Susan Brown, PhD, Nancy Byl, PhD, PT, FAPTA, Noemi Cantin, Dr. Jorge Carranza, Robert Chen, MB, FRCPC, Edward Dabrowski, MD, Diane Damiano, PhD, PT, Scott Delp, PhD, Ruthmary Deuel, MD, Darcy L. Fehlings, MD, Eileen Fowler, PhD, PT, Marjorie A. Garvey, MD, Sharon I. Gorman, PT, MS, Mark Gormley, MD, Edward Hurvitz, MD, Mary Jenkins, MD, PT, JoAnn Kluzik, PhD, PT, Sahana Kukke, MS, Maria Lebiedowska, PhD, Mindy Levin, PhD, Colleen Lewis, Dennis Matthews, MD, Margaret Barry Michaels, PhD, PT, PCS, Cheryl Missiuna, PhD, OTR/L, Helene Polatajko, PhD, OT Reg. (Ont.), OT(C), FCAOT, Karl Rathjen, MD, Jessica Rose Agramonte, PhD, Paul Steinbok, MD, Dagmar Sternad, PhD, Ann Tilton, MD, Johan van Doornik, PhD, Jason Wingert
Note from Mindy Aisen MD, CEO United Cerebral Palsy Research and Educational Foundation. I would like to express my profound thanks to Dr. Sanger and colleagues who are serving on the NIH Taskforce on Childhood Motor Disorders. I provide a summary of the discussions which occurred at the Group’s most recent meeting. I hope this illustrates the challenges that face clinical researchers trying to understand the neurobiology of Cerebral Palsy, and trying to design clinical trials that will provide rigorous evidence to improve care and neurological functions in children and adults with motor disorders. This group is inclusive of all relevant disciplines and represents a bench to bedside philosophy. Their dedication is to be admired, and the work has only just begun.
Abnormalities of motor function in childhood are classified by clinicians as “positive” and “negative” signs.
The positive signs are characterized by excessive involuntary muscle activity (muscle tightness: increased tone, spasms, dystonic muscle contractions, spasticity), while negative signs are characterized by an inability to generate desired muscle activity (weakness, paralysis, incoordination (clumsy or inaccurate movements), tremor, inability to perform complex tasks). Medical therapy has often focused upon positive (hypertonicity) signs due to the availability of treatments that reduce muscle strength or activation. However, for many children negative signs (weakness) can be an even greater contributor to disability and decreased participation at school, home, or in social activities.
The first step in finding solutions to medical problems is having a consensus about terminology and interpretation of signs on physical examination. This allows clinician researchers to collaborate on research hypotheses and to develop collaborative clinical research programs with certainty that their colleagues are recording the same clinical signs and responses to treatments.
At a meeting in March 2005 on the campus of the National Institutes of Health, a taskforce was created to develop consensus definitions of four negative signs in childhood: weakness, reduced selective motor control, ataxia, and deficits of praxis. The latter was divided into apraxia and developmental dyspraxia based upon whether or not skills had previously been acquired. The definitions are restated here:
Weakness is defined as the inability to generate normal voluntary force in a muscle or normal voluntary torque about a joint.
Reduced selective motor control is defined as the impaired ability to isolate the activation of muscles in a selected pattern in response to demands of a voluntary posture or movement. (for example moving one finger at a time)
Ataxia is defined as an inability to generate a normal or expected voluntary movement trajectory that cannot be attributed to weakness or involuntary muscle activity about the affected joints. (being able to accurately touch a target without extra movements or tremor)
Apraxia is defined as impairment in the ability to accomplish previously-learned and performed complex motor actions, although each subcomponent of the action is possible for the child to perform.
Developmental dyspraxia is defined as a failure to have ever acquired the ability to perform age-appropriate complex motor actions that is not explained by the presence of inadequate demonstration or practice, ataxia, reduced selective motor control, weakness, or involuntary motor activity (and as in the case of apraxia, each performance of each subcomponent of the task is possible)
Following the meeting in 2005, the taskforce recognized the need to develop assessment tools that can be used to categorize and quantify negative symptoms. These clinical experts recognized that the very process of coming together to exchange ideas and develop agreement about terminology had great potential value for improving research and care.
The goal was to provide a set of measures that can be used clinically to diagnose disorders, quantify impairments, and assess treatment effects of interventions.
By providing measurement tools, clinicians and researchers can use these tools to establish inclusion in clinical research trials and measure outcome.
The dialog and interactions involved in the validation of definitions and identification of potential areas for modification of definitions has great value. For example using widely accepted and validated measures and definitions will lead e better specificity of diagnosis and closer relation to the underlying pathophysiology.
Development of appropriate assessment tools is a prerequisite to the performance of clinical trials to evaluate potential treatments; the taskforce concluded that a high priority should be placed upon the rapid development of valid, reliable, and sensitive tools.
Therefore, the NIH funded a second meeting in Stanford, California in March 2006 to have different members of the task force review the current state of the art in clinical or research assessment of “negative signs” in childhood.
Four working groups, one for each of the negative signs (with apraxia and dyspraxia being considered by a single group) reported progress they had made toward assessing the sensitivity, specificity and relavency of currently available clinical measures. The working group leaders are:
Weakness: M Delgado
Reduced Selective Motor Control: D Gaebler-Spira
Ataxia: A Bastian
Apraxia/Dyspraxia: M Hallett and M Wiznitzer
Each working group met at least once by telephone or in person prior to the March 2006 meeting in order to review current technologies. The working group conclusions were presented at the meeting. Discussion within and between groups at the meeting led to specific recommendations for further research and evaluation of particular tools.
Groups were instructed to evaluate each rating scale or instrumented measurement technique according to the following criteria (Modified from Herndon, 1996):
1. It should measure what it purports to measure (Validity).
2. It should successfully identify those who do not have the impairment (Specificity).
3. It should successfully identify those who do have the impairment (Sensitivity).
4. It should be responsive to change of the impairment yet relatively insensitive to day-to-day symptom fluctuation.
5. It should have low test-retest and inter-observer variability (Reliability).
6. It should be rapid and easy to use with little special training (Efficiency).
Groups were further instructed to separate what should be measured from conclusions about how it should be measured. In particular, it was noted that how to measure may depend upon whether the measurement needs to be performed in the clinic (where speed and portability are essential) or in a laboratory setting (where accuracy and reliability are essential).
The weakness working group evaluated Manual Muscle Testing (MMT), Hand-held torque measurement devices (HHD), Isokinetic torque measurement devices, and surface electromyography (sEMG). They also evaluated newer methods including the Myotonometer and muscle imaging using either ultrasound or MRI. The group concluded that strength should be measured as torque during isometric maximal voluntary contraction. For this purpose, manual muscle testing is insensitive and unreliable, but due to its simplicity and common use it may have value as a clinical screening tool that could be used to refer patients for more in-depth measurement. Hand-held torque measurement devices such as strain gauges or pressure sensors show promise, but careful attention will need to be paid to standardization of use, the posture in which strength is tested, ensuring maximal effort, and stabilization of proximal joints. Isokinetic measurements using a seated system such as the Biodex may be helpful, but such devices are often not suited to children, and they would be expected to be useful only in a laboratory setting. Measurement of surface EMG, muscle hardness (myotonometer), and muscle imaging studies may provide useful related information but these do not appear to provide a direct measure of strength. In particular, the group noted that the force output of a muscle may not correlate well with surface EMG between different subjects, and therefore while surface EMG may help to differentiate between different causes of weakness, it does not provide a good measure of weakness.
The use of multi-axis force/torque sensors was identified as an important need in order to determine both the strength in “off-axis” directions as well as to determine the contribution of off-axis components of force to decreased effective torque relative to a particular task. Norms for endurance and the effects of fatigue need to be established for children of different ages. It may be helpful to establish the relationship between strength and perception of effort. Methods for motivating younger children to produce adequate effort need to be developed and standardized. The relation between weakness and limitations in function needs to be investigated in order to determine goals for intervention. This is particularly important in the upper limb, where considerable weakness may be tolerable without adverse effects on performance of common skills. Other important issues include the ability to measure strength during movement, and the ability to measure trunk and neck strength. A helpful tool may be the development of muscle models that are specific to children with motor deficits such as Cerebral Palsy, in order to be able to predict the contribution of peripheral (muscle) vs. central (neurological) factors to measured weakness.
The recommendation at this time is that weakness should be quantified by the use of torque measurements about a joint during maximum voluntary isometric contraction. Further research needs to be done to standardize methods for torque measurement, establish standards for obtaining maximum voluntary contraction, and extend measurement methods to include trunk, neck, and possibly other muscle groups. The most promising current technology for both clinical and research investigation is the use of hand-held force measurement devices, and investigation of these devices should be given high priority. New technology to permit isokinetic measurement in children should be developed. Manual muscle testing will continue to have a role as a clinical screening tool but should not be used as a quantitative assessment. It is essential to exclude possible confounding factors such as practice effects, spasticity, reduced selective motor control, contractures, task-dependent weakness, bradykinesia, and other motor or cognitive signs that could lead to inaccurate measurement of strength.
The conclusions for assessment of weakness can be summarized as follows:
Pre-requisite for assessment: ability to follow instructions, and at least some active range of movement
Identification: reduced voluntary force (in response to command)
Quantification: inability to move against resistance with proximal joints stabilized
Distinction from other signs: should be assessed in and out of synergy patterns, with balance and stability concerns eliminated, and in a single joint at a time.
Reduced Selective Motor Control
Reduced selective motor control is related to the concept of obligate synergistic activation of muscles. Such synergies can occur in the arm, fingers, leg, and possibly trunk and neck, although these latter are not quantified by any current scales. It was noted that particular synergies may respond to practice, so that with training changes could be seen. An expected consequence of reduced selective motor control would be inability to activate specific muscles in combination with other muscles, and therefore the possibility of “task-dependent weakness”. Therefore selective motor control is quantified by a limitation in the maximal torque or range of motion in a joint where the limitation depends upon the posture or torque at other joints. It may also be quantified by a limitation in maximal voluntary EMG in a muscle where the limitation depends upon EMG in other muscles. There may be specific patterns of synergistic activation that are similar across children. For example, extension or flexion synergies of the leg, or a combination of adduction at the shoulder with extension at the elbow.
Instrumented measurements that show promise include the use of gait analysis with particular emphasis on the measurement of correlations between multiple joints, referred to as “angle-angle coupling”. Surface Electromyography is often reported by gait analysis laboratories and this can be particularly useful to evaluate the presence of extensor synergies in the leg. For example, involuntary activation of the gastrocnemius muscle during attempted knee extension with simultaneous ankle dorsiflexion suggests an extensor synergy coupling quadriceps activity with gastrocnemius activity. Other extensor or flexion synergies can be tested in this way. Portable or handheld surface EMG devices may allow non-quantitative screening for such synergies the clinic. The SMC is a promising clinical rating scale that directly tests the ability to isolate individual joints.
Laboratory measures could include multiple degree of freedom force and torque measurements in order to demonstrate obligate coupling between multiple joints, and changes in strength or active range of motion as a function of posture or the angle of proximal joints. Appropriate standardization of techniques, including stabilization of some proximal joints is essential. It may also be important to measure selective motor control during motion. For example, Jules DeWald presented data showing that reduced elbow extension during shoulder abduction may be well demonstrated during movement. The extent of the reduction may depend upon particular constraints on the movement, so constraining motion to particular planes, joints, or ranges of torque may be necessary for assessment. Therefore robotic systems may be needed in order to test limitations in posture or range of motion that occur during voluntary movement.
The recommendation at this time is that further study is needed for all measures. Promising areas include the use of elements of the QUEST or Fugl-Meyer as tests of the ability to achieve specific arm postures, and elements of the SMC scale as a test of lower extremity isolated joint movements. Promising laboratory measures include the use of kinematic analysis to measure angle-angle coupling, the use of multi-axis force transducers to measure changes in strength as a function of posture or torque at other joints, the use of robot manipulators to examine the effects of proximal joint posture on distal joint active range of motion during movement, and EMG evidence of obligate coupling between muscle activation. Testing the use of portable surface EMG devices in clinic will be facilitated once this method has been validated in a laboratory setting.
The conclusions for assessment of reduced selective motor control can be summarized as follows:
Pre-requisite for assessment: ability to follow instructions, and sufficient strength to move the tested limb when it is supported against gravity
Identification: inability to isolate a single muscle or joint during specific tasks
Quantification: clinical measures of the ability to achieve multi-joint postures, multi-axis force transducer measurement of change in force with changing posture of other joints, range of motion at distal joints with proximal joints constrained, and obligate coupling between multi-muscle surface EMG
Distinction from other signs: adequate strength when supported against gravity, worsening when a proximal joint is stabilized out of the synergy pattern
Several validated clinical rating scales are available for quantification of ataxia and ataxic disorders, including the International Cooperative Ataxia Rating Scale (ICARS), the SARA, and the Friedreich’s Ataxia Rating Scale (FARA). Although these rating scales have been developed either for adult disorders or for specific childhood syndromes, they contain common elements that could form the basis for an ataxia rating scale that is applicable to childhood disorders. In particular, several scales include observer assessments of accuracy and intention tremor during finger-to-nose, finger-following, and rapid alternating movements.
A small number of laboratories have investigated the kinematics of arm movements in subjects with ataxia. There is additional data from gait analysis, standing balance, and posturography that could also form the basis for quantitative measures. At this time, such measures have not yet been validated on large numbers of children.
The recommendation at this time is that a pediatric clinical rating scale based upon elements of existing ataxia scales should be developed. Promising techniques for laboratory quantification of ataxia include measurement of the effect of speed on endpoint accuracy and comparison between accuracy in single-joint and multi-joint movements. For assessment of gait ataxia, gait analysis including base of support, joint decomposition, obstacle avoidance, and deviation from a straight line should be evaluated. The evaluation of standing balance and toe-to-target tasks may differentiate between subgroups of patients with gait ataxia due to proprioceptive deficits, balance disorders, or dysmetria.
The conclusions for assessment of ataxia can be summarized as follows:
Pre-requisite for assessment: ability to follow instructions, and ability to reach or kick against gravity, or walk unsupported
Identification: hypermetria that is worse with fast movement or multi-joint movement, and/or irregularity of rate, rhythm, or force on repetitive movement, and/or wide based, stiff, and variable gait
Quantification: magnitude of dysmetria, accuracy as a function of speed and the number of unstabilized joints
Distinction from other signs: hypermetria, irregularity, greater impairment of multi-joint movements both within and outsisde of synergies, lack of crouch gait.
Apraxia and developmental dyspraxia
The distinction between apraxia and developmental dyspraxia is based upon the history of whether or not a particular complex motor action had previously been performed by the child. Therefore assessment tools for both disorders will be similar. Recent efforts to classify adult apraxia have led to at least 5 subcategories: (1) limb kinetic, (2) ideomotor, (3) ideational, (4) conceptual, and (5) dissociation. Adult clinical rating scales are under development, and the working group suggested that pediatric rating scales should remain consistent with the adult scales to the extent possible.
Elements of the Movement Disorder Society Apraxia Scale (MDSAS) adult test may be helpful. However, the MDSAS is still only in draft form and has not been validated. This test tentatively includes the following tasks: coin rotation, toothbrush pantomime, identification and use of a real toothbrush and toothpaste, identification and use of a key with the eyes closed, and imitation of the use of a key. The intent of these tasks is to separate the different adult forms of apraxia. Modification of the particular tasks may allow a similar test to be used in children, although it would remain difficult to apply to very young children.
The recommendation at this time is to develop two clinical rating scales: a short scale and a long scale. The short scale is intended as a screening test that can be used in the clinic, and would be required to have high sensitivity but relatively low specificity. The long scale is intended as a research tool and would be required to have both high sensitivity and high specificity, as well as a graded response to change in ability. Scale elements should include screening questions for current level of skills, as well as tasks that are both transitive (using an object) and intransitive (gestures), imitation of postures and tasks, and action sequences. Scoring should attempt to determine the type of error within each subtask in order to determine the subtype of dyspraxia. Possible child-appropriate transitive tasks include dressing, writing, drawing, gripping an object, opening a door, climbing on the examining table, paper folding, use of a hammer, pencil, key, or spoon. Possible intransitive tasks include wave, blow kiss, ok sign, and silence sign. It will be important to test both novel and previously-learned tasks, and to compare performance to age norms. The order of evaluation of types of dyspraxia may be important, since children with limb-kinetic apraxia may be unable to complete any other tasks. It will also be important to test different modalities of command, including verbal command and visual demonstration of tasks.
We recognize that oromotor and gait apraxia may be significant contributors to disability, but at this time we have not addressed measurement tools for these signs.
The conclusions for assessment of disorders of praxis can be summarized as follows:
Pre-requisite for assessment: ability to follow instructions, adequate strength to resist gravity, lack of obligate synergies that interfere with task to be tested, adequate coordination to perform the task.
dentification: spatial, timing, sequencing, or conceptual errors
Quantification: clinical rating scale and skill tests
Distinction from other signs: other signs must be excluded, and performance does not improve with slow speed of movement or stabilization of proximal joints.
The taskforce recognizes a significant need for further research to determine the most appropriate assessment tools and assess reliability, sensitivity, specificity, and validity of those tools. We recommend that separate tools be developed for use in the clinic as a rapid screening tool and for use in the laboratory as a quantitative diagnostic tool and outcome measure. We further recommend that efforts be directed toward particular methodologies, as summarized in the following table:
Manual muscle testing
|Dynamometer, isometric and
isokinetic single-joint torque
|Selective Motor Control
||clinical rating items
||clinical rating items
||kinematics as a function of
speed and number of free
||clinical rating items – short
|clinical rating items – long
Assessment of a child with negative symptoms should proceed in a stepwise fashion in order to eliminate confounders:
2. Ability for control of independent joint motion
3. Coordination including dysmetria, dysrhythmia, dyssynergia
4. Limb-kinetic apraxia
5. Other deficits of praxis
The goal of the working group over the upcoming year is to develop candidate rating scales and assessment tools, and to begin to obtain reliability, validity, sensitivity, and specificity data. It is expected that this will require testing each rating instrument on children with all four types of motor signs, and possibly on children with other motor signs including hypertonic or hyperkinetic signs. It is expected that a set of clinical measures will result that can be used to guide clinical treatment and can provide meaningful outcome measures for clinical trials to develop and test new therapies for children with motor disorders.
Commentary: The process of having this group interact and focus on appropriate measurement instruments has already led to important exchanges of ideas about best treatments (and the appalling lack of rigorous data) for best practices in CP. On a VERY positive note, however, the best and the brightest are now engaged in the research questions that could revolutionize the care of developmental disorders, particularly strategies for improving strength, selective motor activation, combination therapies to improved motor and functional performance (eg focal spasticity treatment coupled with CIT, functional neuromuscular stimulation, robotics, virtual reality, the list goes on..)