MANAGEMENT OF UPPER EXTREMITY DYSFUNCTION FOLLOWING STROKE OR BRAIN INJURY

Cerebral function depends on a continuous supply of
oxygen. Any significant interruption of oxygenation by thrombosis,
emboli, or hemorrhage results in neuron death and subsequent deficits
in cognitive, sensory, and motor functions. Thrombosis is the most
common cause of infarction and accounts for nearly three fourths of all
CVAs (79,80). Arteriosclerosis is the most significant predisposing factor.

Cranial hemorrhage accounts for approximately one sixth
of all CVAs, and includes spontaneous intracerebral hemorrhage and
subarachnoid hemorrhage. Hypertension is commonly present in these
patients. Isolated cerebral emboli account for less than 10% of CVAs.

Predisposing factors to CVAs include atherosclerosis,
increasing age, genetic predisposition, hypertension, hyperlipidemia,
hypercholesterolemia, obesity, cardiac anomalies (arrhythmias,
myocardial infarction, hypotension, mural thrombosis), diabetes
mellitus, collagen vascular disease (vasculitis, polyarteritis),
hyperviscosity states (polycythemia, sickle cell anemia), oral
contraceptive use, tobacco smoking, severe cerebrovascular spasm
secondary to migraine headaches, and septic vasculitis (tuberculosis,
syphilis, and mucormycosis).

Distinct clinical syndromes arise from insults to
specific areas of the cerebral cortex. CVAs involving the middle
cerebral artery are the most common, and produce the typical hemiplegic
picture of greater impairment in the upper extremity, face, and speech
compared with lower extremity involvement. The middle cerebral artery
supplies the largest area of cerebral cortex. This area controls
sensory and motor function of the trunk, upper extremity, and face, as
well as the functions of speech.

The anterior cerebral artery supplies the mid-cortex in
the sagittal plane. This area of cerebral cortex controls sensory and
motor function predominantly in the lower extremity. CVAs involving the
anterior cerebral artery result in a hemiplegic picture of sensory and
motor deficits chiefly involving the lower extremity.

The posterior cerebral artery supplies the visual cortex
in the occipital region. Involvement of this artery typically results
in visual impairment. Bilateral cortical involvement may lead to severe
mental impairment, frontal release signs, loss of short-term memory,
and inability to learn.

CVAs in the vertebral basilar system are rare. Deficits
in balance and coordination arise from interruption of afferent and
efferent pathways between the brain and spinal cord. Balance reactions
are also dependent on limb control and proprioception (95,119,141,201).

Prognosis following TBI has traditionally been predicted relative to the Glasgow Coma Scale (GCS) (Table 67.2) (99,100,225). The GCS evaluates a patient’s responses to eye opening, motor responses, and verbal responses. The

Glasgow Outcome Scale is frequently used to determine outcome following brain injury (Table 67.3).
Using the GCS score obtained within 24 hours of the patient’s admission
to the hospital, a coma score of 11 or greater is associated with an
82% probability of moderate or good neurologic recovery. Lower scores
have a significantly higher incidence of severe sequelae.

Age is an important factor determining neurologic
outcome following brain injury regardless of the severity of injury.
Patients younger than the age of 20 years at the time of brain injury
experience as a group a 62% moderate or good neurologic recovery.
Patients between the ages of 20 and 30 can expect a 46% chance of
moderate or good neurologic recovery. In a series of pediatric patients
with brain injury, overall 90% achieved a moderate or good neurologic
recovery and only 8% expired or remained in a persistent vegetative
state (94,95). Young
children with a GCS score of 5 or better have a good prognosis for
recovery. In addition to having a poorer prognosis for recovery, the
cost and time required for rehabilitation of older patients is higher
than that for younger patients (39).

Although it has been demonstrated to predict mortality
accurately, recent studies have suggested the GCS as a single variable
may have limited value as a predictor of functional outcome and many
trauma centers are now using the Revised Trauma Score (RTS) to assist
with triage of multitrauma patients. The RTS combines the GCS as well
as the systolic blood pressure and respiratory rate, and is used to
predict both mortality and disability (246,247).

The duration of coma is another prognostic indicator. If
emersion from coma occurs within the first 2 weeks of brain injury, 70%
of patients can be expected to achieve a good recovery. If the coma
persists beyond 4 weeks, the chance of good recovery is much
diminished. Brain stem involvement, as indicated by the presence of
decerebrate or decorticate posturing, has a poor prognosis for outcome.
If decerebrate posturing occurs and resolves within the first week
after injury, 40% of patients will have a good neurologic recovery. If
decerebrate posturing persist beyond the first week, only 9% of
patients will achieve a good neurologic recovery. In a similar manner,
the duration of posttraumatic confusion can also be an indicator of
prognosis. If the period of posttraumatic confusion persists for more
than 4 weeks, then one third of these patients will have a poor
neurologic outcome. It should be remembered, however, that prognosis is
a probability statement, and although various factors can be used as
guidelines, none is an absolute indicator in the individual patient.

Phenol, a derivative of benzene, denatures the protein
membrane of peripheral nerves in aqueous concentrations of 5% or more.
When phenol is injected in or near a nerve bundle, phenol’s neurolytic
action on the myelin sheath or the cell membrane of axons with which it
makes contact serves to reduce neural traffic along the nerve. The
onset of the destructive process with higher concentrations of phenol
may begin to show effects several days after injection. The denaturing
process induced by phenol extends biologically on the order of weeks,
but eventually regeneration occurs. A phenol block is used as a
temporizing measure rather than a permanent intervention. In our
clinical experience and the experience of others, the effect of a
phenol block typically lasts 3 to 5 months.

Histologically, it has been shown that phenol destroys
axons of all sizes in a patchy distribution but more so on the outer
aspect of the nerve bundle, onto which phenol is dripped. When phenol
is percutaneously injected, it is likely that the nerve block will be
incomplete. This is especially useful in situations in which a spastic
muscle also has volitional capacity because under these circumstances,
it is desirable to reduce spasticity while still preserving volitional
capacity of a given muscle or muscle group.

The technique of phenol injection is based on electrical
stimulation. Motor branches are injected close to the offending muscle
or muscle group. These are referred to as motor points. A surface
stimulator is briefly used to approximate the percutaneous stimulation
site in advance. A 25 gauge Teflon-coated hypodermic needle is advanced
toward the motor nerve. Electrical stimulation is adjusted by noting
whether muscle contraction of the index muscle takes place. As the
electrode gets closer to the motor nerve, less current intensity is
required to produce a contractile response. The motor nerve is injected
when minimal current produces a visible or palpable contraction of the
muscle. Generally, 4 to 7 ml of 5% to 7% aqueous phenol is injected at
each site. As with any injection, care needs to be taken not to inject
into a blood vessel and this is done by aspirating before the injection
(13,16,28,44,70,71,72,101,102,104,107,122,123,129,130,131,145,146,156,159,229).

Botulinum toxin is a newer agent used in the localized
treatment of spasticity. Ordinarily, an action potential propagating
along a motor nerve to the neuromuscular junction triggers the release
of acetylcholine (ACh) into

the
synaptic space. The released ACh causes depolarization of the muscle
membrane, activating a biochemical sequence that leads to muscle
contraction. Botulinum toxin type A is a protein produced by Clostridium botulinum
that inhibits this calcium-mediated release of ACh at the neuromuscular
junction. Botulinum toxin A attaches to the presynaptic nerve terminal
and divides into a light and a heavy component. The light component
invades the nerve cell and interferes with fusion proteins affiliated
with vesicles of ACh, thereby preventing the release of ACh from their
storage vesicles.

Botulinum toxin injection has been used to treat a
variety of dystonias and is currently approved by the Food and Drug
Administration (FDA) for the treatment of blepharospasm, facial spasm,
and strabismus. A number of studies have reported its use in treating
spasticity in individuals with cerebral palsy, stroke, head trauma, and
multiple sclerosis (37,50,58,145,146,200,205,240).
Clinical benefit lasts 3 to 5 months but may be more variable.
Botulinum toxin is injected directly into an offending muscle, and
depending on the size of the muscle being injected, dosing has ranged
between 10 and 200 units (U). Current practice is to wait at least 12
weeks before re-injection and not to administer a total of more than
400 U in a single treatment session. Because this upper limit of 400 U
may be reached rather quickly, a different strategy is needed for the
limb requiring many proximal and distal injections. Botulinum toxin A
and phenol may be combined, the former being injected into smaller
distal muscles and the latter aimed at larger proximal ones. A 3- to
7-day delay between injection of botulinum toxin A and the onset of
clinical effect is typical. Effects will not be seen by the patient
immediately and usually a follow-up visit is arranged to check the
result. The amount of toxin given for a particular muscle is variable.

The technique of botulinum toxin injection varies. Some
physicians inject through a syringe attached to a hypodermic needle
that doubles as a monopolar EMG recording electrode. Patients may be
asked to make an effort to contract the targeted muscle or the muscle
may be contracting involuntarily as in dystonia. After inserting the
needle electrode, injection is made when EMG activity is recorded. For
deep or small spastic muscles (e.g., finger flexors), electrical
stimulation is preferred as a means of localizing the muscle before
injection.

Botulinum toxin injections have gained much popularity
in the past several years. The advantages of botulinum toxin are (a)
ease of injection and (b) the lack of residual scaring after injection.
The disadvantages of botulinum toxin are (a) high cost and (b)
stimulation of antibody formation that requires higher doses for
repeated injections. Phenol, in contrast, requires more technical
expertise to localize the nerve or motor points for injection. Phenol
is caustic and causes localized scarring of the nerve and muscle.
Phenol, however, is inexpensive and readily available.

Evaluation of spasticity focuses on identification of
three factors: (a) the clinical pattern of motor dysfunction, (b) the
patient’s ability to control muscles involved in the clinical pattern,
and (c) the role of muscle stiffness and contracture in relation to the
functional problem. For purposes of convenience, we identify six
clinical patterns of upper extremity motor dysfunction that are most
commonly seen (Table 67.7). Other variations of
motor dysfunction occur but are less common. Various muscles may
contribute to motor dysfunction across joints and limb segments in
these clinical patterns. During evaluation, focus on the following
characteristics of the involved part: (a) voluntary or selective
control, (b) spastic reactivity, (c) rheologic stiffness, and (d)
contracture. Ask five specific questions:

When many muscles cross a joint, the characteristics of
each muscle may vary. Because each muscle may contribute to motion and
movement of the joint, information about each muscle’s contribution is
useful to the assessment as a whole. Treatment depends on such
information (58,145,146).

Spasticity often masks underlying motor control. In the
upper extremity, the most common pattern of spasticity is one of
flexion. Passive range of motion of each joint should be established
first. This is tested by slow extension of the joint to avoid the
velocity sensitive response of the muscle spindle. When spasticity is
significant and passive joint motion is incomplete, perform an
anesthetic nerve block to assess whether a myostatic contracture is
present (104,105,117,118,126).
In order to evaluate passive joint motion in the entire upper
extremity, perform a brachial plexus block using a local anesthetic.

Unmasking of primitive patterning reflexes further
contributes to the motor impairment. Spasticity (hyperactive response
to quick stretch), rigidity (resistance to slow movement), or movement
dystonias may be present. The degree of spasticity within selected
muscles can be graded clinically in response to a quick stretch as
mild, moderate, or severe. There is surprising consistency between
observers using this simple grading system. Another method of
quantifying muscle tone, which is readily accessible and easily
performed at the bedside, is to measure the amount of intramuscular
pressure generated by a passive quick stretch or during functional use
of the limb. Intramuscular pressure can be measured using a wick or
slit catheter technique. The pressure generated within the muscle is
proportional to the force of contraction (5,182).

Motor control can be graded in the extremity using a clinical scale (Table 67.1) (112). The patient should be observed clinically in a variety of functional tasks.

Differentiating between the relative contributions of
pain, increased muscle tone, and contracture to a limb deformity can be
difficult. Anesthetic nerve blocks are extremely useful in assessing
joint range of motion. The

blocks
can be easily performed without the use of special devices. By
temporarily eliminating pain and muscle tone, patient cooperation is
gained and the amount of myostatic contracture can be determined. By
using local anesthetic blocks, the strength and motor control of the
antagonistic muscle group can also be evaluated (Fig. 67.9).

Clinical examination supported by laboratory studies is
the mainstay of evaluation. The clinical questions of interest
regarding a given problem include the following:

Given the degree of clinical effort, patient morbidity
and procedural costs involved in treating complicated movement
dysfunction in patients with CVA and TBI, clinical examination alone
may not be sufficient to answer these questions with a high degree of
confidence. Technology-driven laboratory assessments that may include
formal gait and motion analysis, dynamic EMG studies, and nerve blocks
may be helpful (105,112,117,118,120).
Dynamic multichannel EMG is acquired with simultaneous measurements of
joint motion (kinematics) in the upper extremity. Kinetic, kinematic,
and dynamic EMG data assist the clinician in interpreting whether
voluntary function (effort-related initiation, modulation, and
termination of activity) is present in a given muscle and whether that
muscle’s behavior is also dyssynergic (sometimes referred to as “out of
phase” behavior). In addition, responses to different rates of passive
stretch of muscle before and after local anesthetic nerve block can
help the clinician distinguish between the dynamic, velocity-sensitive
reflex resistance of spasticity versus passive muscle tissue stiffness
and contracture. Somatosensory evoked potentials (SEPs) and motor
evoked potentials (MEPs) provide information on the integrity of the
sensory and motor pathways and may be helpful in predicting recovery of
motor function after stroke (93). Combined with
clinical information, laboratory measurements of muscle function often
provide the degree of detail and confidence necessary for making
conservative and surgical treatment decisions.

Combining the findings in several previous studies of
spastic patients from our institution the following classification of
EMG activity was devised to standardize terminology and may be used for
either the upper or lower extremity (Table 67.8) (105,112,117,118,120).

Fracture malunions are common in the TBI patient. This
is due to a combination of factors. Trauma concurrent with the brain
injury may cause complex fractures, which are predisposed to malunion.
Injuries may be missed in the initial resuscitation. Fracture healing
may be acelerated by the same mechanisms that cause HO. The prognosis
of the brain injury may be sufficiently poor that optimal internal
fixation of fractures is not sought or obtained. Hemodynamic or
pulmonary instability may cause optimal fracture fixation to be
delayed. Poor patient compliance, agitation, and spasticity may alter
the initial reduction. If this is not anticipated malunion may result.

Results of HO resection are better with higher cognitive
function and when volitional movement of the involved joint is present (69).
In addition, surgical excision of functionally significant
periarticular HO often results in significant improvements in range of
motion, independence, and quality of life (62,63,69,132). Radiographically immature lesions also have a higher rate of recurrence (18,47,60,62,63,69,111,127,132,147,157,195).
It may not be feasible or desirable to allow the HO to reach maturity
before resection, because although the risk of recurrence is higher,
ankylosis of the joint and subsequent contracture worsen the over all
functional result.

Adhesive capsulitis is commonly seen in patients
following stroke and in those with RSD. They have a characteristically
painful shoulder with limited glenohumeral motion. Three clinical and
four arthroscopic stages have been identified (90,161,162,203) (Table 67.9).
The treatment in this group of patients is similar to that for the
general population. Nonsteroidal anti-inflammatory drugs, physical
therapy, and intra-articular injections are all useful. Selected cases
may benefit from manipulation under anesthesia.

Inferior subluxation of the shoulder is a common occurrence in patients with flaccid paralysis of the shoulder girdle (Fig. 67.10).
This is most commonly seen in stroke patients or in brain-injured
patients with concomitant brachial plexus trauma. The subluxation is
usually self-limiting, but occasionally, the shoulder will be
chronically subluxated, causing pain. The patients typically have no
functional use of the extremity. Patients complain of increased pain
when upright. The pain may be due to chronic stretch on the shoulder
capsule or from traction on the brachial plexus. Physical examination
shows a positive sulcus sign, with little to no active motion of the
involved shoulder. There is a prominence of the acromion and atrophy of
the deltoid. There may be contracture of the shoulder in adduction and
internal rotation. Subacromial or intra-articular injections of local
anesthetics do not relieve the symptoms. Radiographs show inferior
subluxation of the humerus on the glenoid (Fig. 67.10). Brachial

plexopathy must be ruled out by using diagnostic EMG.

Conservative treatment may include electrical
stimulation to the deltoid and supraspinatus muscles use of a sling.
This relieves the symptoms by elevating the humeral head in the
glenoid. Although this technique is usually successful in the short
run, this is frequently unacceptable to the patient as a permanent
solution. A surgical solution to this problem of excessive laxity is
the biceps suspension procedure (Fig. 67.11).

Several procedures have been described to treat
paralytic inferior subluxation of the shoulder, but none has gained
widespread acceptance. Braun (31) has advocated
using the coracoacromial ligament to suspend the humeral head. The
lateral portion of the coracoid process together with the
coracoacromial ligament is detached and transferred to the humeral
head, where it is fixed using a cancellous lag screw. Garland has
advocated detaching the proximal end of the long head biceps tendon,
looping the tendon over the clavicle and securing it back on itself.
With time, the paretic biceps muscle tends to stretch and the humerus
once again subluxates inferiorly. Shoulder arthrodesis has also been
performed but is not well accepted by the patient because it produces a
rigid joint, which interferes with passive positioning, hygiene, and
nursing care.

Overactivity of the supraspinatus muscle can cause
spastic abduction posturing. The deformity is usually dynamic, becoming
more prominent with ambulation, transfers, or other attempted
activities (Fig. 67.12). The affected arm is
held in an abducted posture, making balance while ambulating difficult.
Patients complain that their balance is thrown off because of bumping
into furniture, doorways, and people in crowds. Diagnosis requires
examination of the patient at rest and during a variety of activities.
It is also helpful to elicit from caretakers or family members any
history of activities that trigger this posture. Use dynamic EMG to
confirm that spasticity of the supraspinatus muscle is causing the
deformity.

It is possible to lengthen the supraspinatus effectively
by means of a slide procedure. This procedure has been used
successfully to correct spastic abduction deformity.

The arm is adducted tightly against the lateral chest
wall, and shoulder internal rotation causes the forearm to lie against
the middle of the chest. The tendon of pectoralis major is often
prominent when the examiner attempts to abduct and externally rotate
the shoulder but other muscles contribute to the deformity. The
glenohumeral joint normally functions as a universal joint, enabling
the hand to reach an almost spherical volume of locations in
three-dimensional space. When patients attempt to reach forward,
spastic adductors and internal rotators can severely restrict
acquisition of targets in the environment and on the body. The
patient’s ability to stabilize, push, or apply force to an object is
also compromised. From the perspective of passive function goals such
as skin care and axillary hygiene, spastic adductors and internal
rotators hinder efforts of caregivers to gain access to the axilla to
provide needed care. Restricted motion may impair dressing, washing,
and bathing, and promote skin irritation and maceration. Passive
manipulation of the shoulder during personal care may cause pain and
trigger spastic resistance in reactive muscles.

Muscles that contribute to spastic adduction and
internal rotation dysfunction of the shoulder include latissimus dorsi,
teres major, the clavicular and sternal heads of pectoralis major, and
subscapularis. Involvement of latissimus dorsi and teres major should
be considered when hyperextension posturing of the shoulder is
observed. Antagonistic activity in these muscles may be masking a
patient’s potential for active flexion. Diagnostic lidocaine block to
the thoracodorsal nerve or the lower subscapular nerve may unmask that
voluntary potential. When the pectoralis major is chronically spastic,
the musculotendinous insertion of pectoralis major is prominent and
tight. However, the two heads of this muscle may be differentially
spastic, and EMG recordings from each or diagnostic

lidocaine
blocks to medial and lateral pectoral nerves may help to distinguish
whether one or both heads are pathophysiologically active. Release of
all four muscles may be required to relieve the deformity in a
nonfunctional extremity (31,104,126).
In patients who have evidence of underlying control of muscle function
despite the presence of dyssynergy, the pectoralis major, latissimus
dorsi, and teres major muscles can be fractionally lengthened at their
muscle tendon junctions. Alternately, the teres major muscle can be
partially released from its origin on the scapula and allowed to slide
distally in the same manor as the supraspinatus slide.

Upright posture favors hypertonia in the “antigravity”
elbow flexors of the upper limb. In the patient without motor control,
severe flexion posturing can lead to skin maceration in the antecubital
fossa, malodor, and skin breakdown. In reality, a continuum of
volitional control is seen. Many patients complain that their elbows
persistently “ride up” when they stand up and walk. They also complain
that their flexed elbow hooks door frames and other people, and that
putting on a shirt or jacket is a struggle. Kinesiologically, the elbow
lengthens and shortens the upper extremity. Consequently, active
dysfunction is characterized by impaired reaching for objects in the
environment, placing them elsewhere or bringing them to the body.

Spastic extension of the elbow is much less common than
spastic flexion. These patients have frequently had a brain stem
infarct or injury. They complain of difficulty reaching their face for
activities of daily living.

Ulnar neuropathy occurs in stroke and brain-injured
patients for a number of reasons. Prolonged elbow flexion with traction
on the nerve can lead to decreased volume of the cubital tunnel,
resulting in nerve compression. Support of the torso by leaning on a
chronically flexed elbow may result in direct compression of the nerve.
HO, particularly when it is in a posterior location, may involve the
ulnar nerve causing a neuropathy. The patients are often limited in
their ability to complain about ulnar nerve symptoms because of limited
cognitive and communicative abilities. The diagnosis is usually
suspected because of intrinsic atrophy, and it is confirmed using nerve
conduction studies. A 2.5% incidence has been shown in patients with
TBI at a large brain injury referral center (113,214).

Treatment is ulnar nerve transposition, often at the
same time as elbow flexor lengthenings, flexor releases, or resection
of HO. Subcutaneous transposition is preferred to avoid stimulation of
HO formation.

Pronation deformity of the forearm in an UMN lesion is more common than supination deformity. Pronation bias

makes it difficult for a person to reach for a target underhand,
whereas supination deformity impairs reaching for targets that require
overhand reach. Many activities of daily living depend on active
supination. The use of feeding and grooming utensils and clothes
fasteners becomes problematic when spastic or contracted pronators
restrict supination. Physical examination reveals a fully pronated
resting position of the forearm. When passive supination range of
motion exceeds active supination range, the possibility of pronator
muscle dyssynergy during active supination should be suspected. Muscles
that potentially contribute include pronator teres and pronator
quadratus. Dynamic EMG studies of pronator teres, pronator quadratus,
and biceps greatly augment clinical examination. Clinical examination
does not easily predict which of the pronators might be retaining
volitional capacity and which might be spastic. Both pronator muscles
may show varying degrees of volition and spasticity. Interestingly,
flexor spasticity of the powerful biceps often coexists with a pronated
forearm deformity.

During the period of functional recovery, phenol or
botulinum toxin may be injected into either or both pronators,
depending on clinical and laboratory analyses. In the period of
residual deficits, surgical lengthening of pronator teres and pronator
quadratus may be performed depending on their individual voluntary
capacities and the clinical goal is to improve active supination
function by reducing pronator dyssynergy. The possibility of
lengthening the pronator teres has long been recognized, but fractional
lengthening of the pronator quadratus is of recent vintage. Release of
the flexor-pronator origin was previously a common operation. We do not
advocate this procedure because it does not target the individual
muscles responsible for deformities. When an excessive amount of the
pronator teres origin was released or when there was no function in the
pronator quadratus muscle, an iatrogenic supination deformity occurred.
The supination deformity was much less functional than a pronated
forearm. When the pronators are contracted and not active volitionally,
muscle releases may be considered, although passive functional
advantages, other than cosmesis, may be difficult to come by.

Spastic supination is a far less common deformity but is
also associated with elbow flexion deformities. Most often, this
deformity is seen as a complication of the surgical flexor-pronator
origin release. The biceps, supinator, or both may cause supination
deformity. Physical examination supplemented by dynamic EMG may be used
to determine the relative contribution of each.

Perform correction of the supination deformity in
conjunction with correction of the elbow flexion deformity. As
described previously, in the functional extremity, perform a biceps
Z-lengthening. In a nonfunctional extremity, perform a distal biceps
release. Often, at the conclusion of this procedure, the arm is able to
achieve a functional range of pronation. If not, attention must be
turned to the supinator.

Muscles that potentially contribute to wrist flexion
include the flexor carpi radialis (FCR), flexor carpi ulnaris (FCU),
palmaris longus (PL), flexor digitorum sublimis (FDS), and flexor
digitorum profundus (FDP). Singly or in combination, these muscles may
have variable features of spasticity, contracture, and voluntary
control. Because they have a larger cross-sectional area, wrist flexor
muscles are generally stronger than their extensor counterparts.
Despite a net balance of forces favoring flexion, the extent to which a
patient may have voluntary control over wrist extensors should be
investigated. Dynamic EMG studies and temporary diagnostic motor point
blocks are helpful. When wrist hyperextension deformity is present,
wrist extensors are typically volitional and spastic, wrist flexors are
often poorly volitional and mildly spastic, and the fingers are tightly
clenched into the palm.

Begin clinical examination by observing “resting”
posture of the wrist. FCR, FCU, or both may bowstring across the wrist,
and radial or ulnar deviation suggests their respective involvement. A
clenched fist points to extrinsic finger flexors as having a role. If
fingernails dig into the palm, FDP is likely to be involved. If the
proximal interphalangeal (PIP) joint is markedly flexed but the distal
interphalangeal (DIP) joint is not, involvement of FDS is likely.

Laboratory examination of the flexed wrist deformity
includes recordings from FCR, FCU, ECR, ECU, FDS, and FDP. Kinesiologic
study includes “isolated” muscle group testing along with whole limb
movements such as reaching, grasping, and releasing objects. In
addition, passive stretch of the wrist flexors and finger flexors
provides an indication of spastic reactivity as seen
electromyographically. Dynamic EMG findings are not easily predictable
from clinical examination. Some patients show extensive activation of
wrist extensors during extension and reaching efforts. Nevertheless,
the wrist remains flexed because tension in the wrist flexors and
finger flexors overbalances extensor forces. EMG activity in FCR is
often present during attempts at wrist extension, and activity in FDS
or FDP may be present. In patients with ulnar deviation, EMG activity
is typically seen in FCU on reaching effort, but “isolated” testing
also suggests that patients can often activate FCU voluntarily. Because
EMG activity is not correlated with force production, diagnostic nerve
blocks are often helpful in unmasking movement. Temporary chemical
“weakening” of a dyssynergic wrist flexor may unmask strength in the
wrist extensors sufficient to improve active wrist motion. A similar
hypothesis can be used for the extrinsic finger flexors after dynamic
EMG reveals whether the FDS or FDP is generating antagonistic activity
acting to restrain wrist extension. Motor point block of the target
muscle group or median or ulnar nerve blocks at the elbow may be
performed to examine for active wrist extension during reach. Combined
median and ulnar nerve blocks at or above the elbow also reveal the
presence of muscle contracture.

When a patient has underlying voluntary control,
surgical treatment for flexed and hyperextended wrists are myotendinous
lengthenings. Selective muscle releases, wrist fusion, and proximal row
carpectomy can be considered in the presence of severe deformities.
Subtotal carpectomy

combined
with radio-carpal or radio-metacarpal fusion has been suggested for the
treatment of severe flexion contracture either not amenable to, or
refractory to, soft-tissue releases alone (185,189).

In patients with a severely contracted hand, a
single-stage procedure consisting of superficialis-to-profundus
transfer, wrist flexor release, FPL lengthening, wrist arthrodesis,
carpal tunnel release, and ulnar motor branch neurectomy or intrinsic
release can provide comprehensive correction. Such a definitive
approach to severe deformities is associated with acceptable morbidity
and eliminates the possibility of recurrence or undercorrection from
untreated intrinsic pathology (189).

Spastic forearm flexor muscles causing wrist and finger
flexion deformities are treated during the physiologic recovery phase
by phenol motor point blocks (70,71 and 72).
Because of the large sensory components of both the median ulnar
nerves, perform motor point injections of the wrist and finger flexors
using surface stimulation to localize the motor points. The blocks can
be supplemented with functional electrical stimulation of the wrist and
finger extensor muscles, and by casting or splinting techniques.
Perform gentle passive range of motion of the wrist and fingers. When
motor control is present, a program of active exercise and functional
training is also employed. Electrical orthoses, which rely on
functional electrical stimulation (FES), are available and have been
shown to provide improvement in hand function in people with spinal
cord injury or stroke. These devices are still in their early stages of
development; however, technical advances in the field of FES provide
hope for more widespread applicability in the future.

Extension deformity of the wrist causes hygiene problems
and may prevent release in patients with poor digital extension. Median
nerve compression may also be caused by prolonged extension. If median
nerve compression is diagnosed, perform carpal tunnel release.

Preoperative evaluation is performed to determine which
extremities have sufficient volitional control of the muscles to allow
surgical procedures aimed at restoring function to the hand. Often,
severe deformities are present, but there is insufficient or no
volitional activity in the muscles. In these cases, perform contracture
releases to decrease pain, improve position and cosmesis of the hand,
and to ease basic skin care and hygiene. The criterion for determining
which procedures are most appropriate are summarized in Table 67.10.

The spastic clenched fist deformity is common in brain
injury or stroke involving the upper extremity. This pattern results
from unmasking of the primitive grasp reflex. The fingers are typically
clasped into the palm. Fingernails may dig into palmar skin, and access
to the palm for washing may be compromised. When access is chronically
restricted, skin maceration, breakdown, and malodor occurs. Patients
may complain of pain when they or their caregivers attempt to pry the
fingers open in order to gain palmar access. Some relaxation of finger
tightness may occur if the wrist is positioned in extreme flexion. The
deformity, however, is often accompanied by wrist flexion as well.

The degree of motor control may be masked by the severe
amount of tone present in the finger flexors. First, establish passive
range of motion. Following this, ask the patient to open and close the
fingers and to flex and extend the wrist. If no active wrist or finger
extension is seen, it is still important to assess whether there
appears to be active control of finger flexion. In the continuum of
neurologic impairment and recovery, control of wrist and finger flexion
is seen before active control of extension. Place a finger in the
patient’s palm and ask the patient to grasp. Often, an increase in the
pressure of grasp can be felt indicating underlying muscle control.

Next, perform an anesthetic block of the median nerve in
the antecubital space to eliminate flexor tone temporarily. A block of
the ulnar nerve in the cubital canal can supplement relaxation. With
the flexor muscles relaxed, the activity of the extensor muscles can be
more accurately evaluated. When extensor control returns, it is
generally seen in the extensor indicis proprius muscle first (105,117,118,120).

Muscles that contribute to the clenched fist deformity
include the FDS and FDP. If the PIP joints flex while the DIP joints
remain extended, spasticity of FDS rather than FDP may be suspected.
Dynamic EMG studies have shown that the flexor digitorum superficialis
muscles exhibit a marked degree of spasticity, whereas the flexor
digitorum profundus muscles are often normal or minimally spastic (105,117,118,120).
Despite the marked increase in tone, it often has some underlying
volitional control. The flexor profundus has less spasticity and better
volitional control. Volitional control of the finger extensors is
present in 50% of patients with spastic flexion deformities.

The intrinsics may be also be spastic along with the
extrinsics, but an intrinsic plus posture (i.e. combined MCP flexion
and PIP extension) is not seen because spastic extrinsic flexors
dominate by flexing the PIP joints. Some degree of contracture of the
extrinsics is typical of the chronically clenched fist. From the
perspective of active functional potential, some degree of volitional
control may also be present in either or both sets of extrinsic finger
flexors. Spastic finger flexors may override and mask the patient’s
potential to extend the fingers. Sometimes, a patient presents with
spasticity in just one or two muscle slips of either FDP or FDS. For
example, we have seen a number of cases of index finger flexion traced
to a spastic FDP muscle slip for that finger alone.

During the period of residual deficits and remediable
function, a variety of orthopaedic options are available. When
volitional control is demonstrated in the extrinsic flexor muscles by
dynamic EMG, the fractional lengthening is indicated. In a hand with
skin maceration and malodor from a clenched fist deformity in which no
volitional movement is detected, more significant lengthening of the
flexor tendons is required. In this situation, perform a
superficialis-to-profundus (STP) tendon transfer.

The thumb-in-palm deformity is heterogeneous in appearance and may be secondary to spasticity of multiple muscles

including the flexor pollicis longus muscle and the median and ulnar innervated thenar muscles (19,48,96,104,143,144,189,193,199,204,220).
The thumb is held within the palm, the DIP joint of the thumb is
commonly flexed, and the thumb is unable to function during key grasp
or in three-jaw chuck grasp (i.e., in opposition to the pads of the
index and third fingers). In addition, skin maceration and breakdown
can occur if proper hygiene is prevented.

Clinically, spasticity of the flexor pollicis longus is
indicated by flexion of the interphalangeal joint. Some patients may be
able to extend the thumb if the wrist is flexed, suggesting that a
spastic flexor pollicis longus (FPL) may be impeding active thumb
extension when the wrist is more extended and FPL is tighter. The
thumb-in-palm deformity may result from spastic activity in FPL,
adductor pollicis (AP), or the thenar muscles, particularly flexor
pollicis brevis. Adduction of the thumb metacarpal indicates spasticity
of the AP muscle and possibly the first dorsal interosseous muscle. A
quick stretch of the thumb into abduction often elicits a clonic
response. An anesthetic block of the ulnar nerve in Guyon’s canal at
the wrist temporarily eliminates intrinsic tone. This will demonstrate
the presence of any myostatic contractures and will also confirm that
the AP was an offending muscle in the deformity. Contracture of the
skin of the web space and interphalangeal joint contracture of the
thumb may also develop over time. If some volitional potential in thumb
extensors or thumb abductors is present, lengthening of the spastic FPL
and AP will facilitate key grasp. Dynamic EMG and lidocaine blocks are
helpful to elucidate the specifics of motor control.

In the period of residual deficits and remediable
function, orthopaedic treatment consists of fractional lengthening of
the FPL at the myotendinous junction combined with a thenar muscle
slide, in which the origins of the thenar muscles are detached from the
transverse palmar ligament while preserving the neurovascular pedicle.
Fractional lengthening of the FPL at the myotendinous junction will
improve thumb extension. This is generally performed in conjunction
with wrist or digital flexor lengthening. In order to provide a
functional lateral pinch, it is desirable to stabilize the
interphalangeal joint of the thumb. This may be done using a Moberg
screw or a Herbert screw (Fig. 67.22).

In those cases with a fixed adduction contracture, perform surgical lengthening of the thenar muscles (19,49,96,104,143,144,189,193,199,204,220).
Generally, all of the thenar muscles are spastic or contracted, and a
proximal myotomy is required to reposition the thumb and decrease the
underlying tone in order to improve pinch function. Avoid distal
releases because these often result in a hyperextension deformity of
the metacarpophalangeal joint of the thumb (19).

A modified extensor pollicis longus (EPL) tendon
rerouting procedure in combination with intrinsic muscle releases has
been suggested for the treatment of severe deformities. In this method,
the EPL tendon is transected proximal to Lister’s tubercle. The distal
stump is passed through the first compartment tunnel and is sutured to
the proximal EPL stump. For mild deformities the EPL, rerouting may be
performed alone (193).

When spasticity of the extrinsic flexors is present, intrinsic spasticity should be expected (15,21,25,28,29,103,104,105,106,107,108,114,115,117,121,122,125,126,133,136,145,146,189)
However, intrinsic spasticity and contracture are frequently masked by
the presence of extrinsic flexor spasticity or contracture. Extension
of the fingers at the metacarpophalangeal joints may be blocked by
spasticity of the interossei and lumbrical muscles of the hand. Another
manifestation of intrinsic spasticity is the tendency to swan-neck or
boutonnière positioning of the fingers. When a release or tendon
lengthening of the spastic extrinsic flexor muscles has already been
done, an intrinsic positive deformity of the hand is unmasked. These
hand deformities can be painful and disfiguring. Such contractures
often lead to maceration of the palmar skin and recurrent nail bed
infections from poor hygiene.

The degree of tension caused by the intrinsic muscles
can be demonstrated by comparing the amount of proximal interphalangeal
joint flexion obtained with the metacarpophalangeal joints, both flexed
and extended. If there is less proximal interphalangeal joint flexion
with metacarpophalangeal joint extension, then the intrinsic tendons
are tight. Perform this test both before and after a lidocaine block of
the ulnar nerve at the wrist in order to distinguish between intrinsic
tone and contracture.

Boutonnière deformities are commonly associated with
intrinsic spasticity. They result from a combination of intrinsic
spasticity combined with FDS tone. Swan-neck deformities may also
result from increased intrinsic tone (Fig. 67.24).
The central extensor band is relatively shortened relative to the
lateral bands because of tension exerted by the intrinsics and long
extensor. In both of these cases, take care to distinguish between
deformities caused by the intrinsic spasticity that should improve with
treatment and deformities resulting from the more usual mechanisms such
as traumatic central slip injury with lateral band subluxation or
traumatic mallet finger, because if not caused by spasticity the
deformity will not improve with treatment for spasticity.

Because it is impossible to delineate fully the relative contributions and balance of spasticity and contracture of

the intrinsic and extrinsic muscles by clinical assessment alone, we
routinely obtain dynamic EMG studies of the intrinsic muscles before
embarking on treatment of hand deformities. This is especially
important before considering any surgical intervention.

In the period of residual deficits and remediable
function, three treatment options are available. The procedure chosen
is based on considerations of contracture and the presence or absence
of volitional activity in the intrinsic muscles. When no significant
intrinsic contracture is present and dynamic EMG indicates that there
is no volitional control in the intrinsic muscles, perform a neurectomy
of the motor branches of the ulnar nerve in the palm (Fig. 67.7).
Leave the sensory branches intact to preserve protective sensation in
the hand. When a contracture of the intrinsic muscles is present and
the dynamic EMG study shows no volitional activity, perform a release
of the lateral bands of the extensor hood mechanism at the level of the
proximal phalanx (Fig. 67.25). In these cases,
neurectomy of the motor branches of the ulnar nerve is performed
simultaneously to prevent recurrence of the intrinsic plus deformity
from spasticity of the interosseous muscles. When there is either a
dynamic or static intrinsic plus deformity but the EMG demonstrates
volitional control, release the interossei from their proximal origins
on the metacarpals and allow to slide distally (207,208).
A static deformity is one in which a myostatic contracture is present.
A dynamic deformity is one in which the deformity results mostly from
increased tone with little or no fixed contracture.

A less common deformity pattern is the intrinsic minus
hand. In these patients, the intrinsic muscles have normal or weakened
tone, but there is spasticity of the extrinsic finger flexors. There
may be increased tone in the extrinsic extensors as well. This pattern
results in a claw hand posture,

with
hyperextension of the metacarpophalangeal joints and flexion of the
proximal and distal interphalangeal joints. Ulnar neuropathy must be
considered as a possible diagnosis. Hyperextension contracture of the
metacarpophalangeal joint capsule is common. When it is present, the
contractures require surgical release. Treatment of this deformity may
also require lengthening of the extrinsic digital flexors or STP
transfers, as described earlier.

Release of the dorsal capsule of the metacarpophalangeal
joints from the metacarpal is often needed. Zancoli capsulodesis may
also be required to restore metacarpophalangeal flexion and place the
hand in a more functional and cosmetic position (220,248). Postoperatively, splint the hand with the metacarpalphalgeal joints flexed 70° for 3 weeks.