Orthopaedic clinical decision-making to optimize the walking ability of
children with cerebral palsy should be based on five sources of information
(Fig. 1).
Clinical History
The clinical history is obtained by reviewing the patient's medical record
and interviewing the child and his or her care providers. The medical history
identifies associated medical problems and all previous treatments. The social
history assesses the family dynamics and support resources required to
undertake the arduous rehabilitation that is required after musculoskeletal
surgery. A functional assessment of both the cognitive and the motor
neuro-development is performed. Finally, the child's and care providers'
perceived functional problems and goals for interventions to improve gait
should be identified.
Physical Examination
The physical examination begins with observing the patient's gait. This
should be done in a systematic fashion in two planes. The ranges of motion of
all major joints from the pelvis to the foot are measured. The accuracy and
repeatability of these measurements are improved by using standardized
techniques of limb-segment manipulation and using a goniometer and an angle
guide29-33.
Skeletal alignment in the transverse, coronal, and sagittal planes is assessed
in a similar systematic fashion. A neurological examination is performed to
identify any deficits of selective control of muscle activation, muscle
strength, and balance.
Diagnostic Imaging
Diagnostic imaging is used to better assess skeletal alignment. It is most
valuable for evaluation of the foot and ankle and for assessment of the
alignment of the lower extremity in the coronal and sagittal planes. It is
best to make these radiographs while the patient is standing. Common patterns
of malalignment of the foot and ankle include ankle valgus, pes equinus, pes
planovalgus, and pes
equinovarus1,7.
Rotational malalignments, which are common in the femur and tibia, cannot be
evaluated well on plain radiographs. The value of computerized tomography in
the assessment of rotational malalignment of the femur and tibia is more
controversial34.
Concerns about the accuracy of this measurement technique are related to the
difficulty in defining and determining the axis of the proximal part of the
femur, the femoral neck, and the femoral head.
Quantitative Gait Analysis
Quantitative gait analysis utilizes high-speed motion-picture cameras,
retroreflective markers on the surface of the skin aligned with palpable
skeletal landmarks, and force platforms to measure various aspects of
gait.
Kinematic data, presented as a waveform, describe the motion occurring
simultaneously in three planes. Kinetic data (moments and powers), also
presented as a waveform, define the motion occurring about the joints during
the gait cycle. Analysis of kinetic data provides a biomechanical basis for
the understanding of pathologic gait and helps to provide a rationale for the
selection of various orthopaedic, neurosurgical, and orthotic
interventions.
Dynamic electromyography documents the timing of individual muscle
activation during the gait cycle. Other modalities that may be used include
pedobarography (dynamic assessment of foot pressure distribution and
progression) and energetics (dynamic assessment of oxygen consumption,
reflecting the physiologic costs associated with walking).
Interpretation of these data is improved by understanding the technologies
and techniques used to measure movement. Kinematic calculations assume a
constant relationship between the skin surface marker and the underlying
skeletal
landmark2,7,35.
However, this relationship varies depending on the patient's soft-tissue and
bone anatomy. Kinetic data require determination of the relationship between
the ground reaction force vector and the joint center. The location of the
joint center is estimated from the skin surface markers. The model that is
most widely used to make these estimates for the pediatric population is based
on the measurement of twenty-five hip and pelvic radiographs of children of
unstated age with presumably normal hip
joints35. The
effects of hip joint pathoanatomy and skin-surface-marker motion artifact on
kinetic calculations have not been systematically
established36.
Dynamic electromyography measures the electrical potential generated by a
muscle when it is activated. The magnitude of the electrical signal is related
to the force being generated by the muscle. This relationship can be
determined when the magnitude of that potential can be normalized to that
associated with a maximum manual muscle contraction. However, children with
cerebral palsy may have impaired selective control of muscle activation,
compromising the ability to assess a maximum manual muscle contraction. In
such a situation, the dynamic electromyography data can be normalized to the
maximum value occurring during the gait cycle, or more commonly presented as a
raw
signal2,37.
In this circumstance, the dynamic electromyography data can provide valuable
information concerning the timing of muscle activation but cannot indicate
relative muscle strength or weakness. An appreciation of these limitations
greatly enhances the critical and correct application of the quantitative data
to the clinical decision-making process.
Quantitative analysis of normal gait has identified four important
components or prerequisites: (1) stability of the limb in stance phase, (2)
clearance of the limb in swing phase, (3) effective shifts of the limb from
stance to swing and from swing to stance phases, and (4) occurrence of these
components in a fashion that promotes maximum efficiency of energy
expenditure.
In pathologic gait, these four components are disrupted to varying degrees
and in varying combinations. Gait deficits or deviations, as identified by
quantitative gait analysis, associated with the pathologic gait of children
with cerebral palsy may be classified as primary, secondary, or
tertiary2,4,7,38.
Primary deficits are directly related to the underlying disorder of the
central nervous system and include spasticity, impaired balance, and impaired
motor control. In general, orthopaedic interventions do not directly address
these primary deficits. Secondary deficits or deviations occur as a
consequence of growth and development of the musculoskeletal system and are
generally sequential and progressive over time. Such deficits result in
greater disability as children grow into young adults. Examples include the
progression of deformity of muscle-tendon units from completely dynamic
deformity (overactivity with no fixed shortening) in early childhood to
myostatic deformity (fixed or structural shortening or contracture) seen in
later childhood and preadolescence. The same sequential progression is true
for skeletal deformities, in which flexible and passively correctable
segmental malalignment at the foot and ankle is followed in time by rigid
skeletal deformity. Finally, the interpretation of quantitative gait data is
facilitated by the recognition of common patterns of pathologic gait of
children with cerebral palsy. The most common patterns— jump gait,
crouch gait, stiff gait, recurvatum gait, and intoeing or outtoeing
gait—have distinct kinematic and kinetic
profiles2,5,7,9,10,39,40-44.
Examination Under Anesthesia
The clinical assessment of whether a muscle-tendon unit deformity is
dynamic or myostatic is best performed when the child is examined under
general anesthesia. When the child is awake, the presence of spasticity can
limit the distinction between dynamic overactivity and fixed shortening of the
muscle-tendon unit. Spasticity is not present under general anesthesia, so
that any limitation of range of motion is due to a myostatic deformity
(contracture) of the muscle-tendon unit or intra-articular pathology such as
adhesions or bone deformity.
The Diagnostic Matrix
Clinical decision-making with use of the diagnostic matrix is best carried
out at an interpretation session attended by clinical and technical members of
the motion analysis laboratory team. The certainty of selecting the best
intervention is proportional to the consistency of the data within the
diagnostic matrix. The selection and results of a specific treatment are
optimized when the problems being considered are common and the data from all
sources are consistent. It is not possible to be as confident about the
selection of intervention or its results when the problems are unusual and the
data within the matrix are not consistent. This inconsistency may be a
consequence of attempts to classify the continuous spectrum of a clinical
disease process into a categorical framework and/or deficiencies in the
technology and techniques of quantitative gait analysis. Sorting out these
inconsistencies frequently involves the input of members of both the clinical
and the technical team. For example, discrepancies concerning motion about the
hip and pelvis between observational gait analysis and quantitative gait
analysis are generally resolved by giving greater weight to the quantitative
data. This is done because visual assessment of the movement about the hip and
pelvis during the gait cycle is difficult and frequently misleading, whereas
the accuracy of the kinematic assessment is much greater in all three
planes24,35,42,45,46.
On the other hand, discrepancies among the clinical examination, diagnostic
imaging, and quantitative gait analysis data about the foot and ankle should
be resolved in favor of the data derived from the clinical examination and
diagnostic imaging. The reason for this is that the foot and ankle model used
in quantitative gait analysis is
simplistic35,47-49.
For example, ankle dorsiflexion is defined by the relative alignment of the
shank and foot axes in the sagittal plane, and the foot axis is defined by a
single marker over the forefoot that combines the movements occurring through
multiple foot segments into a single measure. With a major skeletal segmental
malalignment, such as pes planovalgus, the model overestimates the amount of
ankle dorsiflexion occurring during midstance, giving inaccurate kinematic
data.
With use of the diagnostic matrix, it is possible to identify the
indications for the selection of specific musculoskeletal surgical
interventions to optimize the gait of children with cerebral palsy.
Iliopsoas Recession
Clinical history: The most common symptom indicating the need for
this procedure is the inability to stand straight and walk with the upper body
erect.
Physical examination: A hip flexion contracture (>30°) is
an indication for iliopsoas resection. The patient should be examined in the
supine position (Thomas test) and the prone position (Staheli
test)29. The
magnitude of a hip flexion contracture is frequently overestimated by these
techniques because of the difficulty in stabilizing the pelvis, and the
measurements are inconsistent because of variability in pelvic positioning.
The current recommendation is to position the pelvis so that the line
connecting the anterior superior and posterior superior iliac spines is
vertical50. This
technique facilitates more consistent positioning of the pelvis and ensures
that the angle measured during the clinical examination is the same as that
used in the kinematic model in gait analysis for the calculation of hip motion
in the sagittal plane. Observing the child while he or she walks in a
"high kneel" pattern (i.e., walks on the knees) removes the
influence of tight hamstrings on pelvic alignment
(Figs. 2-A and 2-B). This will
unmask any tightness of the hip flexors.
Diagnostic imaging: No imaging studies are required.
Quantitative gait analysis: With this analysis, which identifies
kinematic and kinetic indicators (Figs.
3-A, 3-B, and
3-C), pelvic motion in the
sagittal plane shows an anterior tilt with a "double bump"
waveform pattern during stance phase. Hip motion in the sagittal plane shows
diminished extension in terminal stance, with a decreased dynamic range
throughout the gait cycle. Kinetic analysis of the hip moment in the sagittal
plane shows an increased internal extension moment in midstance, with delayed
crossover to an internal flexion moment in midstance or terminal stance.
Examination with the patient under anesthesia: This is frequently
required to resolve the roles of dynamic and myostatic deformities of the hip
flexor muscles.
Management: There is considerable controversy concerning the
appropriate indications for iliopsoas recession to improve gait in children
with cerebral
palsy40,50-56.
A better understanding of the biomechanics of hip, pelvic, and trunk motion is
needed to clarify the indications for this procedure in these children. When
hip flexion contractures are present, they are seldom an important problem for
a child with cerebral palsy who is able to walk. In addition, these
contractures are frequently overestimated on clinical examination. The
kinematic and kinetic variables identified as indications for this procedure
are frequently the consequence of other primary gait deficits, such as
weakness of the abdominal and hip extensor muscles and impaired balance and
position senses. Similar quantitative profiles may be seen as tertiary or
compensatory deviations, such as in a severe crouch gait due to insufficiency
of the ankle plantar flexors and overactivity of the knee flexors. These
disorders increase hip flexion and possibly anterior pelvic tilt, shifting the
body's center of gravity forward relative to the hip center, increasing the
external hip flexor moment and delaying the flexor-to-extensor moment
crossover in stance phase. Such compensatory gait patterns resolve
spontaneously following correction of the underlying primary or secondary
deficits.
Femoral Rotation Osteotomy
Clinical history: Candidates for this procedure, or their parents,
report intoeing, knee knocking, and tripping.
Physical examination: The patient has increased femoral
anteversion, which is best appreciated when he or she is examined in the prone
position with the hip in full extension. Hip rotation is abnormal, with
increased internal rotation and limited external rotation, indicating
increased femoral
anteversion24,29.
The value of another physical examination maneuver, the trochanteric
prominence angle test, has recently been shown to be limited by obesity,
scarring associated with previous surgery, and variable alignment of the
greater trochanter with respect to the axis of the femoral
neck57,58.
Observational gait analysis may reveal internal rotation of the patella
relative to the line of progression when the child walks. This visual
assessment is often confounded by increased, asymmetric pelvic rotation during
the gait
cycle24,42.
Diagnostic imaging: Computerized tomography scans may be utilized
to assess femoral anteversion when the findings of the clinical examination
are compromised or confusing. The most widely accepted techniques are
performed in either two-dimensional or three-dimensional (volumetric
reconstruction)
formats34.
Quantitative gait analysis: Kinematic data reveal increased
internal hip rotation throughout the gait cycle
(Fig. 4).
Examination with the patient under anesthesia: Anesthesia may be
necessary to evaluate a patient with suspected femoral anteversion. Relative
internal-external rotation of the hip is tested as described above, and when
necessary the examination can be augmented by the use of fluoroscopy to
visualize the alignment of the femoral head and
neck59.
Management: There is a relationship between increased femoral
anteversion (a static, structural skeletal deformity) and increased internal
hip rotation (a dynamic gait deviation). The latter occurs as a compensation
to restore the lever arm available for the hip abductor muscles, which is
diminished by the former. When both are present, surgical correction of the
increased femoral anteversion will result in resolution of the gait deviation
and normalization of hip rotation during
walking24. Not all
children with cerebral palsy and increased femoral anteversion exhibit
increased internal hip rotation when walking, and not all children with
cerebral palsy and increased internal hip rotation during the gait cycle have
increased femoral anteversion.
Medial Hamstring Lengthening
Clinical history: The patient or parents may complain that the
child cannot stand up straight, walks with the knees bent, and has anterior
knee pain with fatigue when walking a distance.
Physical examination: Straight-leg raising is limited
(<60°), the popliteal angle is diminished (<130° when measured
within or behind the popliteal space), and there is a spastic response to a
fast stretch of the
hamstrings29.
Diagnostic imaging: Patella alta with elongation and fragmentation
of the inferior pole of the patella is frequently seen on lateral radiographs
of the knee. These changes are presumably a consequence of chronic overload of
the extensor mechanism due to an increased and prolonged external knee flexion
moment occurring throughout stance
phase60.
Quantitative gait analysis: Kinematic, kinetic, and dynamic
electromyography data are useful (Figs.
5-A, 5-B, and
5-C). Knee motion in the
sagittal plane shows increased knee flexion during the loading response,
variable knee alignment in midstance and terminal stance, and decreased knee
extension in terminal swing. The knee moment in the sagittal plane shows an
increased internal extension moment throughout stance phase. Dynamic
electromyography shows prolonged activity of the medial hamstrings into
midstance.
Examination with the patient under anesthesia: The relative
contribution of dynamic and myostatic components of hamstring deformity is
determined with this examination. It is best done by measuring of the
popliteal angle, with direct palpation of both the medial and the lateral
hamstrings.
Management: The hamstring muscles cross both the hip and the knee
joint, and appropriate medial hamstring lengthening will improve knee
extension at initial contact and in terminal
swing6,40,52,54,56,60-63.
Increased anterior pelvic tilt is not seen following unilateral or bilateral
medial hamstring
lengthening50,64.
Lateral hamstring lengthening is necessary only in teenagers with a severe
crouch gait (>40° of knee flexion in stance) and should otherwise be
avoided to minimize the risk of excessive weakness following surgery. Severe
hamstring tightness may cause posterior pelvic
tilt50. In this
situation, the increased anterior pelvic tilt that may occur following
bilateral medial and lateral hamstring lengthening is not harmful. Knee
flexion contractures that exist following hamstring lengthening are best
corrected by the use of serial casts in the postoperative period.
Rectus Femoris Transfer
Clinical history: The patient or parents complain that the child
has stiff knees, toe dragging, and tripping.
Physical examination: The patient has a positive prone rectus test
(Duncan-Ely
test)1,7,29-31.
The child is placed in the prone position with the hips and knees fully
extended. The test is considered positive when the pelvis elevates (as a
result of hip flexion) as the knee is slowly passively flexed and/or there is
a "catch" on rapid passive knee flexion.
Diagnostic imaging: No imaging studies are required.
Quantitative gait analysis: Kinematic and dynamic
electromyographic data are useful (Figs.
6-A and
6-B). Knee motion in the
sagittal plane shows a diminished dynamic range of motion (<80% of normal),
delayed and diminished peak flexion in swing phase, and a blunting of the
flexion wave in swing phase. Dynamic electromyography shows an inappropriate
midswing phase of the rectus
femoris65.
Coactivation of the vastus lateralis in midswing may or may not be present and
does not seem to affect the outcome of a rectus femoris
transfer66.
Examination with the patient under anesthesia: No additional
information is gained from an examination under anesthesia.
Management: Diminished and delayed peak knee flexion in swing
phase may be the consequence of decreased velocity and diminished stride
length. In addition, a combination of factors occurring during stance phase,
at the stance-to-swing interval, as well as abnormal activity of the rectus
femoris in midswing can all contribute to the disruption of knee flexion in
swing phase. Currently, the rectus femoris transfer is indicated to maintain
or improve the magnitude and timing of peak knee flexion in swing phase when
hamstring lengthening is being
performed6,62,63,66-74.
The rectus femoris transfer is best performed at the time of hamstring
lengthening, but it may also be done following inappropriate isolated
hamstring lengthening that has resulted in a stiff knee gait pattern. Although
it is considered a component of the quadriceps femoris muscle group, the
rectus femoris is actually a two-joint muscle (the other three components are
single-joint muscles) whose activation timing and functional importance during
the gait cycle are distinct from those of the remainder of the quadriceps. For
this reason, transfer of the rectus femoris does not compromise knee extensor
function in stance
phase71,72.
Gastrocnemius Lengthening
Clinical history: Patients and their families report toe walking,
toe dragging, tripping, and intoeing.
Physical examination: The patient has a diminished passive range
of ankle dorsiflexion, a sustained spastic response to an applied fast stretch
of the gastrocnemius muscle (clonus), and an increased deep tendon reflex at
the ankle29. It is
helpful to consider the foot and ankle as consisting of three segments
(hindfoot, midfoot, and forefoot) and two columns (medial and lateral). The
alignment of each segment is described relative to the adjacent proximal
segment. The three most common patterns of segmental malalignment of the foot
and ankle in children with cerebral palsy are equinus (seen most commonly in
younger children), planovalgus (seen most commonly in children with spastic
diplegia), and equinovarus (seen most commonly in children with spastic
hemiplegia)1,4,7,8,49,75,76.
Equinus malalignment consists of plantar flexion deformity of the hindfoot,
with the other segments in normal alignment and the columns having an
appropriate length. Planovalgus malalignment consists of hindfoot plantar
flexion and eversion, midfoot pronation, and forefoot supination. The lateral
column is functionally shorter than the medial column. Equinovarus
malalignment consists of hindfoot equinus and inversion, midfoot supination,
and forefoot pronation. The lateral column is functionally longer than the
medial column.
Diagnostic imaging: Weight-bearing plain radiographs of the foot
and ankle are routinely utilized. When malalignment is found on physical
examination, it is evaluated further on the radiographs, with the investigator
looking for structural abnormalities.
Quantitative gait analysis: Kinematic, kinetic, and dynamic
electromyographic data are routinely obtained (Figs.
7-A,
7-B,7-C,7-D).
Ankle motion in the sagittal plane shows excessive plantar flexion in stance
and swing phases. All three ankle rockers in stance phase are disrupted, with
an absence of the first (or heel) rocker, flattening or inversion of the
second (or ankle) rocker, and a premature, diminished third (or forefoot)
rocker. The ankle moment in the sagittal plane shows an absence of the
internal dorsiflexion moment in the loading response, an increased internal
plantar flexion moment in midstance (the double bump pattern), and a decreased
internal plantar flexion moment in terminal stance. The ankle power in the
sagittal plane shows premature power generation in midstance and diminished
power generation in terminal stance. Dynamic electromyography of the
gastrocnemius is notable for premature activation of the stance-phase burst,
beginning at initial contact or in terminal swing.
Examination with the patient under anesthesia: The relative
contributions of the gastrocnemius and soleus muscles to tightness of the
ankle plantar flexors are best determined with an examination under
anesthesia. These contributions are assessed by measuring ankle dorsiflexion
with the knee flexed (which relaxes the gastrocnemius and allows evaluation of
the soleus) and extended (greater reduction in ankle dorsiflexion motion with
the knee extended, compared with the motion when the knee is flexed, is
attributed to myostatic deformity of the
gastrocnemius)29.
Management: Discrepancies between the passive range of motion of
the ankle noted on physical examination and the dynamic ankle motion measured
during gait may be the consequence of (1) an increased spastic response
following the disrupted first rocker, which further limits the dynamic range
of motion at the ankle in the second rocker; (2) the fact that the external
forces applied about the ankle during gait (particularly in adolescence and
the teenage years, primarily as a result of increased body weight) are greater
than those that can be applied by the examiner during the clinical assessment;
or (3) the fact that foot and ankle segmental alignment during the clinical
examination may not be consistent with dynamic foot and ankle segmental
alignment during gait (because of variable lever arm alignment). Discrepancies
between the findings of observational gait analysis and those of quantitative
gait analysis at the foot and ankle may be the result of the assumption that a
toe strike at initial contact, easily appreciated visually, implies increased
or excessive ankle plantar flexion. However, increased knee flexion, with
neutral ankle alignment, also results in a toe-strike pattern at initial
contact4,7,49,56,77,78.
Quantitative gait analysis (kinematic data) effectively documents the absence
or presence of excessive ankle plantar flexion in this situation.