The Shoulder and Arm

Directed physical examination is the most reliable
method of determining the nature of the injury. Gross inspection of the
infant may demonstrate partial or complete paralysis of the affected
limb. Provocative testing by eliciting neonatal reflexes to induce
elbow, wrist, and digital extension may be used. The Moro reflex is
demonstrated by placing the infant face up on a soft, padded surface.
The head is gently lifted to remove the body weight from the pad. The
head is then released suddenly but quickly supported. The infant may
have a “startled” look as the arms fling out sideways with the palms up
and the thumbs flexed. As the reflex ends, the infant draws the arms
back to the body, elbows flexed, and then relaxes. Unilateral absence
of the Moro reflex may indicate neural or bony injury. Similarly, the
asymmetric tonic neck reflex, activated by turning the head to one
side, may be used to elicit upper extremity movement. As the head
turns, the arm and leg on the same side extends while the opposite
limbs bend.

Osseous injuries to the shoulder girdle should be sought
as the cause of the paralysis or as a coexisting injury. Soft tissue
swelling may indicate a fracture of the clavicle or the proximal
humeral physis. Fractures of the clavicle are often initially
unappreciated, being discovered as a lump over the site of injury at
about 10 days after birth. Findings in proximal humeral physeal
fractures can be subtle, such as irritability with attempted arm
movement or pseudoparalysis. Plain films of the shoulder should be
obtained with comparison views of the opposite side. Because the
proximal humeral ossification centers do not coalesce until about 5 to
7 years of age, other studies may be necessary to confirm the
diagnosis; arthrography, ultrasound, and magnetic resonance imaging
(MRI) have been used to help delineate the position of the largely
cartilaginous proximal humeral fragment.

The use of invasive and noninvasive diagnostic tools has
attempted to differentiate between root avulsions and extraforaminal
ruptures. One study examined the results of myelography, myelography
combined with computed tomography (CT), and MRI and comparing them with
operative findings in infants. CT myelography demonstrated a
truepositive rate of 94%; similar results were obtained with MRI with
the additional advantage of permitting more distal visualization of the
plexus.

Electromyography (EMG) and nerve conduction velocities
(NCV) have also been used to further define the extent of the
neurological deficit. In the early postinjury period, documentation of
nerve recovery can be performed; however, the presence of motor
activity in a given muscle unit may not accurately reflect the extent
of useful motor function. Multiple investigators in helping to predict
functional outcome have emphasized the importance of serial clinical
examinations. Gilbert and Tassin first described the importance of
monitoring the return of biceps function as an indicator of overall
neurologic recovery. They noted that if normal biceps function did not
return by 3 months of age, the outcome at 2 years of age was abnormal. Michelow et al.
found increased accuracy in predicting outcome by combining elbow
flexion with return of wrist, finger, and thumb extension and shoulder
abduction. A general consensus exists that total plexus or C5-C7
involvement, or the presence of Horner’s syndrome, predict a poorer
prognosis for spontaneous recovery.

In the early postinjury period while awaiting neurologic
recovery, nonsurgical treatment is directed at preserving joint motion.
Daily passive motion of all affected joints should be performed with
adjunctive splinting as necessary to prevent contracture. Conservative

measures
should be continued until sufficient motor recovery has occurred or the
child is able to cooperate with a postoperative regimen following a
reconstructive procedure.

Controversy remains in regard to the indications for and
timing of operative intervention for brachial plexus injuries. The
timing of plexus exploration with microsurgical nerve graft
reconstruction or nerve transfers for upper root avulsions is based on
the overall extent of injury and return of biceps function. Infants
with total plexus involvement with Horner’s syndrome and no return of
biceps function at 3 months or partial plexopathy with no return of
biceps function between 3 and 6 months may be considered for surgical
intervention.

Chronic untreated or incompletely recovered plexopathies
may develop significant contractures impairing activities of daily
living. Upper-trunk lesions result in weak external rotators against
intact internal rotators and adductors; this imbalance may lead to
posterior glenohumeral subluxation or dislocation with secondary
glenohumeral changes. In an attempt to avert this complication, a
subscapularis muscle slide has been described to increase passive
external rotation at 1 year of age for those unresponsive to physical
therapy.

In the slightly older child with an internal rotation
contracture, external rotator and abductor weakness, and posterior
glenohumeral subluxation without significant glenoid deformity, the
Sever-L’Episcopo procedure has been successfully performed to address
the anterior contracture in addition to providing active external
rotation. First, open release of the anterior capsule with division of
the subscapularis and pectoralis major increase passive external
rotation. Second, transfer of the teres major and latissimus dorsi
tendons to the rotator cuff permits active external rotation.

If osseous deformity with flattening of the humeral head
and posterior glenoid is present from chronic subluxation or
dislocation, a derotational osteotomy of the proximal humerus becomes a
reasonable option. A transverse osteotomy is performed followed by
external rotation of the distal segment to improve the utility and
function of the existing arc of motion by rotating it into a more
central position.

Multiple studies have demonstrated glenohumeral
arthrodesis improves stability and function of the upper extremity
after brachial plexus injuries with a resultant flail shoulder. Despite
the loss of glenohumeral motion, residual scapulothoracic and elbow
motion reliably permit activities at waist and midelevation levels.

Symptomatic glenohumeral arthrosis following a chronic
plexus injury may require prosthetic replacement. Patients suffering
from debilitating pain unresponsive to conservative measures including
anti-inflammatory medications and activity modification may benefit
from a joint replacing procedure. However, given that resultant motion
and glenohumeral stability is highly dependent on the integrity and
function of the surrounding rotator cuff and deltoid musculature, pain
relief is a more reliable surgical goal.

As with other joints in the body, the AC joint is
subject to inflammatory and degenerative processes. Primary
osteoarthritis of the AC joint appears to be related to the normal
aging process. The relatively small surface area of this joint is
subject to high loads and shear stresses with disc degeneration
occurring as early as the second decade and degenerative changes by the
fourth decade in the majority of specimens obtained from 151 patients.

In spite of these findings, symptomatic primary
osteoarthritis of the AC joint is relatively uncommon. In the
evaluation of pain around the shoulder, consider other possible
coexisting conditions. AC arthrosis may be only one of a number of
pathologic disorders such as rotator cuff impingement or glenohumeral
arthritis resulting in the symptom complex. Similarly in rheumatoid
arthritis, a recent prospective study of 74 patients demonstrated that
although the AC joint was affected more often than the glenohumeral
articulation, both joints were affected in almost half (42%) of the
cohort.

Acute or repetitive injury can result in posttraumatic
arthritis. Acromioclavicular sprains and separations and fractures of
the clavicle, particularly those with intra-articular extension, can
produce a painful joint. Osteolysis following repetitive microtrauma
has gained increased recognition as a potential source of AC joint
symptoms. Most commonly associated with weight-lifting activities, the
so-called “weightlifter’s clavicle” may result from repeated injury to
the subchondral bone resulting in microfractures. The resorptive
response following this trauma produces the characteristic osteolysis.

The patient with isolated AC joint symptoms typically
report discomfort and aching along the anterior and superior aspect of
the shoulder. Occasionally, the pain may radiate into the trapezius,
deltoid, and down the arm; irritation of the AC joint has been shown to
mimic symptoms similar to cervical radicular pain.

Common daily activities elicit and exacerbate the pain.
Reaching overhead to high shelves, behind the back to a back pocket or
to unhook a bra, or to the opposite shoulder or axilla for daily
hygiene increase contact between the acromial and clavicular facets.
Pain at night may also occur, disrupting sleep when the patient rolls
onto the affected shoulder.

A patient may also have a history of more physically
demanding heavy-load or repetitive overhead activities resulting in or
exacerbating the pain. These patients may include laborers, throwers,
golfers, swimmers, and racquet-sport athletes. Weight-training
individuals often report onset of symptoms after bench pressing, dips
on the parallel bars, and push-ups.

Physical examination may initially reveal joint
prominence and asymmetry secondary to osteophyte formation or
synovitis. Localized tenderness over the AC joint is present, and pain
may be reproduced with certain provocative maneuvers such as adducting
the arm across the body to the opposite shoulder (the cross-body
adduction test) or placing the shoulder into internal rotation with the
hand behind the back (Figure 12-1).

Glenohumeral and scapulothoracic motion are usually
largely unaffected. With the exception of chronic cases that may
develop some loss of terminal internal rotation and cross-body
adduction, passive motion is rarely limited other than by discomfort.
Greater restrictions in motion may indicate underlying capsular
tightness or glenohumeral arthrosis.

A complex or unclear clinical picture can be clarified
with direct local anesthetic injection into the AC joint. Elimination
of symptoms several minutes following an injection may serve as a very
useful diagnostic tool as well as a reliable indicator of probable
success or failure with a distal clavicle excision procedure.

Anteroposterior (AP) and axillary views of the AC joint
are usually sufficient for initial radiographic evaluation. The Zanca
view provides an unobstructed AP view of the AC joint by angling the
x-ray beam 10 to 15° cephalad. Stress views obtained with traction or
weights on the affected extremity are not routinely indicated in
evaluation of degenerative AC joint disorders but may be useful in
diagnosis of AC joint instability following trauma.

The radiographic findings will depend on the underlying
pathological process. Degenerative disease manifests itself similarly
to other joints with narrowing of the joint space, subchondral cysts
and sclerosis, and osteophyte formation (Figure 12-2).
In contrast, osteolysis of the distal clavicle may demonstrate relative
osteopenia, widening or tapering of the distal clavicle, and expansion
of the joint space (Figure 12-3).

Three phase technetium-99m bone scans may be useful in
cases where plain radiographic findings do not correlate with the
clinical examination. In the active patient whose history and physical
exam are inconsistent with radiographic findings,

bone scintigraphy may detect subtle osteolysis as well as infectious or neoplastic conditions.

The initial approach to a symptomatic degenerative or
osteolytic AC joint is nonoperative treatment. Activity modification,
heat, and nonsteroidal anti-inflammatory medications along with
judicious use of intra-articular steroid injections may adequately
alleviate symptoms. Physical therapy, while not directly addressing the
AC joint pathology, may be useful as an adjunct to decrease or prevent
restricted motion as well as address any associated rotator cuff
disease. Although some authors recommend 6 months prior to initiating
operative treatment, this must be adjusted to the patients’ activity
level, symptoms, and degree of disability.

If a course of nonoperative treatment has failed,
surgical intervention may be considered. Open distal clavicle excision
was first described independently by Mumford and Gurd in 1941 and remains a reliable procedure allowing direct

visualization of the AC joint ensuring adequate bone removal.

As the technology has evolved and more procedures are
performed arthroscopically, the technique of arthroscopic distal
clavicle resection has become more common. Esch et al.
described a subacromial approach to excision of the distal clavicle in
patients undergoing a subacromial decompression procedure. A superior
direct approach to the AC joint was subsequently developed with the
theoretical advantage of preserving subacromial and capsular anatomy (Figure 12-4).
The advantages of either arthroscopic approach over open resection are
reduced morbidity with smaller incisions and avoidance of detaching the
deltoid and trapezius. However, more technical skill is required in
addition to the potentially higher risk of inadequate resection (Figure 12-5).

The complications of treatment are similar for both open
and arthroscopic distal clavicle resection. Persistent pain or weakness
can occur and may be attributable to diagnostic error, inadequate bone
resection, or joint instability.

About 5 to 10% of patients with shoulder pain manifest
glenohumeral degenerative joint disease. In primary osteoarthritis, a
source cannot be identified but may result from a combination of
genetic and environmental factors resulting in articular cartilage
injury. Several secondary forms of arthritis have been described in the
shoulder including posttraumatic, rheumatoid, inflammatory,
postinstability repair, and rotator cuff arthropathy. The incidence of
shoulder osteoarthritis is higher in women and appears to increase with
age for the typical patient in the sixth or seventh decade of life.

The anatomic findings of glenohumeral osteoarthritis
were well characterized by Neer with the gross pathologic findings
similar across most of the patient population. Articular cartilage loss
with eburnation and sclerosis of the bone is most pronounced in the
area of the humeral head that contacts the glenoid between 60 and 100°
of abduction. Subarticular cysts may be present. Large peripheral
osteophytes occur most commonly at the inferior margin of the joint,
blocking rotation and enlarging the diameter of the head (Figure 12-6).

The glenoid is typically smooth but eburnated with
marginal osteophytes. These excrescences are easily palpated but
usually obscured from direct view by the overlying capsular and
ligamentous structures. Posterior glenoid wear is a common finding in
conjunction with an internal rotation contracture and posterior
subluxation of the humeral head.

Patients frequently present with an insidious onset of
pain over the course of several months to years. Pain localized to
“deep” in the joint or along the anterior and lateral deltoid is
common, although patterns vary. Symptoms may initially be merely a
nuisance but ultimately can become debilitating. Maneuvering the
affected extremity in space to perform basic activities such as daily
hygiene may become difficult, secondary to pain and loss of motion.
Interference with sleep may also occur with pain at rest and at night,
often with the patient rolling onto the affected side.

On examination, active and passive motion are often
restricted with a relative compensatory increase in scapulothoracic
motion. Contracture of the surrounding soft tissues, distortion of the
bony anatomy, osteochondral loose bodies, and pain may all contribute
to the loss of motion. Movement may elicit palpable and audible
crepitation or grating as the articular surfaces devoid of cartilage
rub against one another. Joint line tenderness with palpation is common
with the posterior margin usually more pronounced due to less overlying
soft tissue. Aside from weakness due to pain inhibition and disuse,
strength is usually well-preserved with a low incidence of tears of the
rotator cuff in association with osteoarthritis.

Patients with degenerative glenohumeral osteoarthritis
share similar roentgenographic features. Plain radiographs in the AP
and axillary planes are usually sufficient to evaluate the extent of
disease. The absence of articular cartilage manifests as a loss of the
joint space along with sclerosis of the subchondral bone and
subarticular cysts in both the humerus and glenoid. Flattening of the
humeral head and large peripheral osteophyte formation, particularly
along the anterior and inferior margins, results in an apparent
increase in the head diameter. Although rotator cuff tears are uncommon
in primary osteoarthritis, as previously mentioned, maintenance of the
subacromial space provides a simple initial indicator of rotator cuff
integrity (Figure 12-7).

Assessment of glenoid bone stock is crucial in
preoperative planning. The AP view is useful in determining the amount
of medial glenoid bone loss. Posterior glenoid bone erosion is common
and is best visualized on the axillary view. The

combination
provides crucial information in predicting the ability to resurface the
glenoid. Although not routine in the initial workup of primary
osteoarthritis, an MRI, usually obtained to determine the status of the
rotator cuff, or CT, may also yield additional information about
glenoid alignment and bone quantity (Figure 12-8).

Symptomatic osteoarthritis may adequately respond to an
initial trial of conservative treatment. Judicious use of oral
anti-inflammatory medications, taken with the caveat of possible
gastrointestinal and renal side effects, may make symptoms tolerable.
Activity modification may diminish exacerbating events but can be
difficult or impossible to institute when symptoms are elicited with
activities of daily living. Intra-articular steroid injections may
provide longer lasting relief but remain unpredictable in duration and
carry the catastrophic risk of infection. Aside from maintaining
generalized strength and conditioning of the muscles, physical therapy
has no formal role in treatment and may in fact worsen symptoms.

Surgical treatment is reasonable after failure of
nonoperative therapy. Several options have been described ranging from
arthroscopic debridement to prosthetic arthroplasty with distinct
indications for each intervention.

Arthroscopic debridement has had mixed results with the
extent of disease as a primary influence on outcome. Success with this
technique has been demonstrated in patients with evidence of early
osteoarthritis. One study found arthroscopy as a reasonable option when
the humeral head remains concentric within the glenoid and where some
joint space remains on the axillary radiograph. Another group found
that patients with Outerbridge IV changes could gain significant
improvement in pain relief and function with results deteriorating with
lesions greater than 2 cm in diameter.

Arthrodesis is seldom, if ever, an accepted treatment
for glenohumeral osteoarthritis. Appropriate situations may include
painful arthritis with permanent loss of motor function (i.e., brachial
plexus injuries or insufficiency of the deltoid and rotator cuff),
chronic infection, failed revision arthroplasty, severe refractory
instability, or bone loss following tumor resection. In the young
patient who is likely to place excessive demands on a prosthesis with
heavy manual labor, consider an arthrodesis. Although it remains a
viable salvage procedure to achieve pain relief, fusion results in
significant functional limitations such as loss of internal and
external rotation and use of the extremity above shoulder level.

Resection arthroplasty also has limited utility in
treatment of glenohumeral osteoarthritis. This treatment modality, now
essentially reserved for failed arthroplasty with extensive bone loss
and salvage after severe infection, results in limited motion and
variable pain relief.

Prosthetic arthroplasty, either replacement of only the
humeral head (hemiarthroplasty) or the head and the glenoid (total
shoulder replacement), remains the treatment of choice for most cases
of degenerative osteoarthritis of the glenohumeral joint. In the early
1950s, Neer introduced his technique and design for humeral head
replacement for complex shoulder fracture-dislocations. Current
technology includes a polyethylene glenoid component and adaptation of
the humeral prosthesis with most modern systems providing an array of
modular stem and head sizes (Figure 12-9).

Unconstrained prosthetic arthroplasty provides
predictable pain relief as well as functional improvement with an
intact functional rotator cuff and appropriate physical therapy. Both
total shoulder replacement and hemiarthroplasty can reliably provide
improvement in pain in more than 90% of patients; however, humeral head
replacement alone may result in less pain relief, relief that
deteriorates with time, and inferior functional outcomes (Figure 12-10).

Although the results of shoulder arthroplasty are good
to excellent in the vast majority of cases, potential complications are
a concern with any prosthetic reconstruction. With an overall incidence
reported to

be
approximately 14%, problems may occur, including instability, rotator
cuff tear, glenoid and humeral component loosening, intraoperative
fracture, nerve injury, and infection. Revision surgery can
successfully manage many of these causes of failed shoulder
arthroplasty, but overall results are inferior compared to primary
procedures.

Any injury or disease process that results in
interruption of the blood supply to the humeral head can lead to
osteonecrosis. Reported causes include trauma, corticosteroid use,
alcohol abuse, radiation, dysbarism, cigarette smoking, and systemic
diseases such as Gaucher’s disease, rheumatoid arthritis, systemic
lupus erythematosus, and infection with human immunodeficiency virus.
Sickle cell hemoglobinopathy is the most common cause

of osteonecrosis worldwide. Occasionally, no cause can be identified.

A patient’s symptoms and physical findings may reflect
the extent of disease. Patients will typically present with pain that
is not well localized and worse with activity. Night pain and pain at
rest may occur but not as commonly as with other disorders of the
shoulder, such as osteoarthritis or rotator cuff disease. Active motion
may be affected early in the disease process by pain inhibition, but
passive motion is often preserved unless capsular contracture and
secondary osteoarthritis are present. Strength is usually preserved
except in those cases with underlying cuff pathology or systemic
disease affecting the muscles.

Radiographic workup should initially consist of plain
radiographs in orthogonal planes. Cruess’s modification of the
Ficat-Arlet classification of osteonecrosis of the femoral head is the
most widely used system for evaluation and treatment planning. The
continuum ranges from stage I where

changes
are not visible on plain radiographs; stage II characterized by focal
sclerosis; stage III with subchondral collapse and loss of the head’s
spherical contour (the crescent sign); stage IV with an area of
collapsed articular surface; and stage V with signs of secondary
arthritis on both the humeral and glenoid surfaces (Figure 12-11).
Signal intensity on MRI depends on water and fat variations and
therefore can reveal osteonecrosis prior to radiograph changes (Figure 12-12).
Bone scanning has been used for early detection of the disease, but
recent work has called its utility into question with failure to detect
more than half of the lesions identified by MRI and histology.

Disease progression and symptom development is varied
with relationship to etiologic factors and extent of involvement. In
general, clinical progression is slow with most patients presenting
with advanced radiographic changes. Patient with sickle cell disease
tend to have the most benign course. Corticosteroid-induced
osteonecrosis appears to fare better than posttraumatic causes with Hattrup and Cofield finding almost twice as many trauma-related cases requiring arthroplasty at 3 years after diagnosis.

A course of nonoperative treatment should initially be
instituted with oral anti-inflammatory medication and physical therapy
for range of motion and strengthening exercises. These interventions
are often more successful in the shoulder than in the hip for several
reasons. The glenohumeral joint is not subjected to the same
weight-bearing forces, the articulation is less conforming than the
acetabulum and can tolerate greater deformity, and scapulothoracic
motion can partially compensate for lost glenohumeral motion.

If nonsurgical options have failed, several forms of
operative management may provide relief of symptoms. Core decompression
has been described with some good results in earlier stages of disease.
LaPorte et al. reported their experience with
63 shoulders followed for an average of 10 years. Results demonstrated
success rates as high as 94% for stage I disease and falling
precipitously to 14% for stage IV osteonecrosis.

Prosthetic replacement, either hemiarthroplasty or total
shoulder arthroplasty, may be appropriate in stage IV and V disease and
select patients with stage III involvement (Figure 12-13). Glenoid resurfacing is used when the glenoid is involved with substantial secondary arthritis. In a recent study, 88

shoulders followed for an average of 8.9 years were treated with either
hemiarthroplasty or total shoulder arthroplasty with 79.5% reporting
subjective improvement and 77.3% with no to moderate pain. Inferior
clinical results were noted in posttraumatic osteonecrosis and superior
results in corticosteroid-induced disease.

Careful evaluation should be performed prior to
initiating operative treatment and should be based on patient function,
symptomatology, and radiographic findings. Rutherford and Cofield
reviewed the data on 33 shoulders and concluded that even with
extensive radiographic changes (stage IV and V), mild symptoms will not
necessarily progress and can be treated nonoperatively.

Involvement of the shoulder may present with an
insidious onset of pain, swelling, and loss of motion. All
synovial-lined joints may be affected including the acromioclavicular,
sternoclavicular, and glenohumeral joints, although the latter can be
the only symptomatic location in up to two-thirds of patients.

Initially, pain may restrict active motion while
preserving passive motion. Development of soft tissue contracture and
bony destruction may eventually affect passive motion as well. Strength
may deteriorate with muscle atrophy and rotator cuff disease, which is
found in as high as 75% of patients with RA.

Radiographic changes are variable with early involvement
exhibiting only minimal findings with more extensive changes late in
the course of disease. Plain radiographs may reveal only osteopenia of
the humeral head and glenoid. Symmetrical

marginal
erosions along the inferior humeral head and subchondral cysts develop
and may eventually involve large portions of the head. Glenoid
destruction occurs with disease progression demonstrating central or
peripheral erosions (Figure 12-14).
Osteosclerosis and osteophytosis is uncommon in RA and typically
reflects quiescence of the inflammation and development of secondary
osteoarthritis.

Superior migration of the humeral head may result from
incompetence of the rotator cuff. Continued contact between the head
and the undersurface of the acromion can produce erosion and thinning
of the acromion process and possible fracture (Figure 12-15).

MRI is useful in delineating the soft tissue component
of the disease. The extent of synovial proliferation and joint effusion
can be visualized. More importantly, the integrity of the rotator cuff
can be determined with pathology ranging from mild inflammation to
full-thickness tears.

Bone destruction and alteration in the normal osseous
architecture is best evaluated with CT scan. In the preoperative
planning process, quantification of bone loss, particularly when
considering implantation of a glenoid component, may be necessary; CT
scan has been shown to provide superior characterization of bony
defects compared to conventional radiography.

Initial nonsurgical management of rheumatoid arthritis
affecting the shoulder is appropriate in the early phases of disease
with minimal bony destruction. Physical therapy may be helpful in
maintaining active and passive motion during acute exacerbations and
gradually adding resistive exercises to preserve strength.

Optimization of medical therapy is the primary means of
controlling the disease process prior to development of extensive
radiographic changes

and
severe symptoms unresponsive to conservative measures. Limited trials
of corticosteroid injections may be useful in reducing the symptoms of
acute inflammation unresponsive to oral medications aimed at the
underlying disease process. Communication with the patient’s
rheumatologist is essential prior to embarking on surgical treatment.
When the rheumatologist’s medical armamentarium of nonsteroidal
anti-inflammatory medications, antimetabolic drugs, steroids, and
disease-modifying drugs has been exhausted, surgical intervention
should be considered. However, once joint destruction is demonstrated,
early replacement before loss of the rotator cuff and glenoid vault
preclude glenoid resurfacing should be considered.

In patients with joint swelling and pain but relative
preservation of the glenohumeral articular surfaces, bursectomy and
synovectomy may be useful. Synovectomy may also help slow disease
progression in aggressive cases of synovitis.

Hemiarthroplasty or total shoulder arthroplasty should
be considered for patients with severe pain and functional limitation
unresponsive to more conservative modalities. The indications for
humeral head replacement are an irreparable rotator cuff tear or
inadequate glenoid bone stock precluding adequate fixation of the
polyethylene component (Figure 12-16). Most
patients who undergo prosthetic arthroplasty for RA can expect some
pain relief and improvement in shoulder function, but results are less
reliable than in the treatment of osteoarthritis.

The presenting symptoms of posttraumatic arthritis are
similar to other forms of degenerative arthritis of the shoulder. Pain
level is variable but is usually the primary complaint. Motion may be
restricted by pain, bone deformity, and contractures of the surrounding
capsule and rotator cuff. Strength may be compromised as well from
injury to the rotator cuff and deltoid or from alterations in rotator
cuff attachments seen in proximal humerus malunions and nonunions.
Neurologic injury should always be considered and appropriate
evaluation with an EMG should be performed if there is clinical
suspicion.

The radiographic findings are varied and depend on the underlying inciting injury. Fractures of the

proximal humerus may demonstrate malunion or nonunion of the
tuberosities, head segment, or surgical neck or articular surface
collapse from posttraumatic osteonecrosis. Arthritis resulting from
anterior instability may show a Hill-Sachs lesion on the posterolateral
humeral head, a so-called Reverse Hill-Sachs lesion on the anteromedial
head in posterior dislocations, persistent anterior or posterior
subluxation with corresponding glenoid bone loss, or a chronic
dislocation with bony injury to the humerus and glenoid. The presence
of metallic implants used for fracture fixation or instability repair
have the potential to encroach on the glenohumeral joint and may be
evident on radiographic examination.

Further testing with CT or MRI may be useful in
preoperative planning in more complex cases where the normal anatomical
relationships have been distorted by the injury.

A course of nonoperative therapy should be attempted as
an initial form of treatment. Activity modification, gentle exercises,
and anti-inflammatory medications may provide enough symptomatic relief
to avoid surgical intervention.

Operative intervention is warranted if nonoperative
therapy has failed. When arthritis has developed secondary to fracture
nonunion or malunion, osteotomies or soft tissue procedures alone will
not address the articular incongruity. Humeral head and glenoid
articular destruction would likely most benefit from prosthetic
replacement. In cases of arthritis following instability repair, mild
arthritic change and significant loss of external rotation due to
tightening of the anterior soft tissue structures may benefit from
anterior releases. Studies examining arthroplasty for arthritis
following instability repairs have shown reduction in pain and improved
function but high rates of revision surgery secondary to instability,
component failure, and pain due to an unresurfaced glenoid at long-term
follow-up. The difficulties following previous injury or surgery
including extensive scarring and distortion of the normal bony and soft
tissue anatomical relationships make these cases challenging for even
the most experienced of surgeons.

The findings of Neer et al.
suggested that the precipitating event in RCTA was, as its name
suggests, the massive, untreated rotator cuff tear. They noted,
however, that only 4% of shoulders with full-thickness cuff tears
progressed to RCTA. Thus, aside from pure mechanical dysfunction, other
possible contributing factors must be involved to account for the
pathologic findings (Figure 12-17).

Normally, the rotator cuff stabilizes the head against
the glenoid and provides the mechanism for rotation of the glenohumeral
joint. Full-thickness tears of the cuff may disrupt these functions as
well

as provide a conduit for escape of synovial fluid essential for cartilage nutrition.

Studies in the rheumatology literature have emphasized
the role of crystal deposition and the subsequent inflammatory cascade
and injury in the pathogenesis of a clinically similar disease. McCarty et al.
hypothesized that diseased synovium and cartilage release basic calcium
phosphate crystals in shoulders clinically similar to RCTA but what
they termed Milwaukee shoulder syndrome. The ensuing inflammatory
response results in injury to the rotator cuff and articular cartilage.

Although the precise cause still remains unclear, these differing theories may in fact represent a single disease entity.

Patients with cuff-tear arthropathy will often present
with predictable symptoms related to arthritis and cuff deficiency. A
long history of progressively worsening pain and stiffness is typical.
Both active and passive motion may be affected secondary to joint
incongruity, pain, and lack of a competent rotator cuff. Dislocation or
rupture of the tendon of the long head of the biceps occurs in most
patients and may be detected on exam. Many patients also exhibit
significant shoulder swelling resulting from synovial fluid
communication between the glenohumeral joint and subacromial bursa (fluid sign).
Aspiration of this fluid may reveal a bloody or blood-streaked
effusion. Large ecchymoses down the arm are not unusual, and often spur
fruitless workups for clotting abnormalities. Shoulder girdle muscular
atrophy, weakness, and loss of active and passive motion will be
evident resulting from disuse, pain, joint incongruity, and the large
rotator cuff tear.

Destructive, arthritic changes of the osseous structures
of the shoulder are characteristic of RCTA. Early in the disease
process, radiographic changes may only reflect a large rotator cuff
tear with proximal humeral migration and abutment against the
undersurface of the acromion. Progression will involve extensive
erosion of the proximal humerus and narrowing of the glenohumeral joint
(Figure 12-18). A standard axillary radiograph may reveal subluxation or dislocation of the proximal humerus.

Additional imaging will most often not be necessary. MRI
is typically not indicated in the evaluation of RCTA, given the drastic
changes seen on plain radiography suggestive of a massive, chronic
rotator cuff tear. Similarly, CT scan does not have specific utility in
preoperative planning aside from assessment of bone loss.

As in most conditions affecting the shoulder, a trial of
nonoperative management is reasonable. Oral nonsteroidal
anti-inflammatories, limited trials of intra-articular cortisone,
activity modification, and gentle exercises may provide some
symptomatic relief.

Several surgical options have been described for treatment of the cuff-deficient arthritic shoulder.

Arthrodesis, now considered a salvage procedure by most surgeons, may
rarely be indicated in the setting of the arthritic shoulder with an
irreparable cuff tear and an incompetent deltoid. Constrained shoulder
arthroplasty, involving a coupled humeral and glenoid articulation,
theoretically provided a stable fulcrum around which the remaining
deltoid could rotate the humerus. However, the tremendous stresses
generated from this construct have been shown to lead to rapid
component loosening, catastrophic implant failure, and fracture.

Nonconstrained prosthetic shoulder arthroplasty is
currently the surgical treatment of choice for this patient population
with pain relief as the primary goal. Despite providing a painless
articulation with prosthetic replacement, the remaining cuff defect
will result in a persistently weak shoulder. Multiple studies examining
the results of total shoulder arthroplasty have emphasized the
importance of the rotator cuff in centering the humeral head on the
glenoid during active motion. Loss of this stabilization effect from an
unrepaired or irreparable cuff tear has been shown to result in
eccentric loading and early failure of the glenoid component. Although
every attempt should be made at reconstructing the rotator cuff, it is
now recommended that large defects with proximal humeral migration
should be treated with hemiarthroplasty alone with preservation of the
coracoacromial arch to prevent anterosuperior migration of the
component (Figure 12-19).

In the early 1990s, Grammont introduced a reverse
shoulder prosthesis with a concave humeral articulation and a spherical
glenoid. Early results in Europe have been promising with respect to
stability and function, however, glenoid survival remains a concern.
Long-term experience with this prosthesis is lacking, and it has only
recently been approved for use in the United States.

The glenohumeral joint is a diarthrodial
ball-and-socket-type articulation with tremendous capacity for motion
in multiple planes. This motion is

obtained
through the sacrifice of joint stability. Often likened to a “golf ball
on a tee,” the humeral head articulates with the small, shallow glenoid
surface. The labral, capsuloligamentous, and musculotendinous
structures surrounding the joint provide compensation for the lack of
inherent bony restraint.

The rotator cuff is composed of four musculotendinous
units that provide multiple functions including motion about the axis
of the joint. The subscapularis, which originates on the anterior
surface of the scapula and inserts onto the lesser tuberosity of the
humerus, provides internal rotation, particularly at the end range of
motion. The supraspinatus originates from the supraspinatus fossa of
the scapula, and inserts on the greater tuberosity of the humerus and
contributes to shoulder abduction. External rotation power is produced
solely by the infraspinatus and teres minor, both of which originate
from the scapula and insert onto the greater tuberosity.

Despite its ascribed name, current understanding of
rotator cuff function has emphasized its role as a prime stabilizer of
the humeral head rather than its motor function. The cuff compresses
the joint and provides resistance to sliding and translation during
dynamic activity. In midrange positions of the shoulder when passive
soft tissue restraints are lax, almost all stability is imparted by
cuff function (Figure 12-20).

Extrinsic mechanical and intrinsic factors have both
been proposed as the origin of rotator cuff disease. No one unifying
theory exists addressing their relative contributions to these
conditions.

The mechanism of external impingement is based on
compression of the rotator cuff between the unyielding surfaces of the
humerus and coracoacromial arch, which is composed of the coracoid,
coracoacromial ligament, and the anterior acromion (Figure 12-21). Inherent morphologic variations of the acromion (flat, curved, or hooked) (Figure 12-22)
or age-related acromial spur formation may contribute to the role of
mechanicalbased pathology of the rotator cuff. Biomechanical

analysis
has demonstrated a correlation between a hooked acromion and increased
contact with the underlying cuff. Furthermore, contact of the acromial
undersurface has also been shown to occur predominantly with the
supraspinatus, consistent with the majority of pathology originating in
this location. More recent studies have described the entity of
internal impingement where the articularsided surface of the cuff
repetitively contacts the glenoid rim. Tensile overload and subtle
glenohumeral instability can result in repeated microtrauma in the
overhead throwing athlete in the absence of external, or subacromial,
impingement.

Intrinsic properties of the rotator cuff have been the
focus of other studies examining the causes of rotator cuff disease.
Higher incidence of articularsided partial-thickness rotator cuff tears
may be partially explained by histologic studies demonstrating
variations in collagen with thinner, less uniform bundles near the
articular surface. Thicker, longitudinally oriented fibers found near
the bursal surface were found to have an ultimate failure stress twice
as high as those found near the articular side. Tenuous vascularity
along the articular surface of the cuff shown in injection studies has
also been suggested as a factor in the pathogenesis of undersurface
partial tears.

The symptom characteristics of rotator cuff disease may
be somewhat variable depending on the duration and extent of tendon
involvement. The patient will typically have a history of slowly
escalating pain or pain arising acutely from a single traumatic event.
A dull ache in the anterior and lateral deltoid is typical, often with
radiation down the arm to the elbow. Occasionally, the pain may radiate
into the trapezius and parascapular musculature, but these findings
should alert the examiner to possible cervical spine pathology; an
examination of the neck is essential to rule out other causative or
potentiating sources of shoulder complaints. Symptoms may initially
occur only with vigorous activities such as lifting or sports requiring
motion above shoulder level. A painful arc of motion between 60 and
120° of elevation is present in most patients. Early in the course,
rest and occasional use of anti-inflammatory medication may provide
significant relief. Secondary stiffness can develop from inflammation
and immobility of the shoulder due to pain inhibition, although
stiffness is surprisingly uncommon with full-thickness tears.
Eventually, pain may progress to symptoms at rest or with use of the
extremity for daily activities such as brushing the teeth, combing the
hair, reaching the perineum, unhooking a bra, or even donning a coat.
Pain preventing or interrupting sleep is common, particularly when
rolling onto the affected side.

Physical exam findings are essential to accurately
assess the status of the rotator cuff. The presence and severity of
certain provocative maneuvers and signs will often aid in determining
the degree of rotator cuff injury.

Abnormalities in the normal contour of the shoulder
girdle may be found on visual inspection. Atrophy in the supraspinatus
or infraspinatus fossae may indicate chronic massive tearing of the
associated tendons or compression of the suprascapular nerve. Loss of
the rotator cuff’s compressive function may result in anterosuperior
prominence of the humeral head. Associated proximal rupture of the long
head of the biceps is identified as a painless soft tissue bulge along
the upper arm (Popeye muscle). A discrete fluid-filled mass or swelling
around the shoulder may also become evident with full-thickness tears
as synovial fluid escapes into the subacromial space (fluid sign).

The diagnostic challenge is often in distinguishing pain
from mechanical impingement from other sources of shoulder pain. The
Neer impingement sign (Figure 12-23) (pain with forced passive forward elevation) and Hawkins sign (Figure 12-24)
(pain with internal rotation and 90 degrees of forward elevation) may
reproduce symptoms. Diminution of pain on repeat testing after
injection of 10 mL of 1% lidocaine into the subacromial space (the
impingement test) is useful not only in verifying the diagnosis as well
as discerning it from other causes of shoulder pain but also in
obtaining a more accurate measure of strength inhibited by pain.

Strength is highly dependent on the specific tendons involved and their degree of integrity (Figure 12-25).
Strength is generally preserved with partial tears but can be falsely
diminished by pain. External rotation lag signs may help differentiate
partial from full-thickness tears; the inability to hold the affected
extremity in external rotation in abduction or at the side is highly
suggestive of a complete tear (Figure 12-26).
The hornblower’s sign, an inability to externally rotate the elevated
arm, is another indication of significant pathology of the posterior
cuff (Figure 12-27). Patients with large cuff tears may also shrug the affected shoulder in an attempt to elevate the arm (Figure 12-28). Tears of the subscapularis are often missed but can be detected by noting a loss of terminal internal

rotation strength, increased passive external rotation as compared to a
normal contralateral side, and inability to perform the belly-press and
Gerber lift-off tests (Figure 12-29).

Plain radiographic studies should include AP views in
the plane of the scapula and glenohumeral joint as well as an axillary
lateral. Subtle findings suggestive of rotator cuff pathology such as
an acromial or AC joint spur, cystic changes in the greater tuberosity,
or a sourcil (or eyebrow) sign with sclerosis on the undersurface of
the acromion to unequivocal signs, such as proximal humeral migration,
may be evident on radiographs. Certain specialized views, including the
supraspinatus outlet (Figure 12-30) (10 to 15°
caudal tilt lateral scapular view) and Zanca (10 to 15° cephalic tilt
AP coronal view), may also be obtained to look specifically for outlet
narrowing and AC joint degeneration.

Because information about the integrity of the rotator
cuff can only be inferred from plain films, additional imaging
modalities are crucial in the evaluation of rotator cuff disease.
Arthrography has been considered the gold standard for diagnosis of
rotator cuff tears. Though highly accurate, it remains an
uncomfortable, invasive test with less utility in determining the size
of full-thickness tears or the presence of partial tears. For these
reasons, it has been largely supplanted by other techniques.

Ultrasonography has been gaining more interest in the
diagnosis of rotator cuff disorders. This noninvasive and relatively
inexpensive technology has been shown to be accurate in detecting and
determining the size of larger full-thickness tears. The problems have
included a high dependence on operator experience and detection of
smaller tears. However, modern equipment and addition of dynamic images
have increased the accuracy.

Magnetic resonance imaging (MRI) is the mainstay of
radiographic diagnosis of rotator cuff pathology. The advantages of MRI
are numerous and include noninvasiveness; capacity to detect
full-thickness, partial-thickness, and intrasubstance tears; and
ability to measure tear size and extent of retraction (Figure 12-31).
In addition, MRI can evaluate other associated abnormal conditions such
as degenerative acromioclavicular or glenohumeral changes, muscle
atrophy and fatty degeneration, capsular and labral pathology, biceps
rupture or dislocation, and ganglion cysts (Figure 12-32).
Through compression of the suprascapular nerve, these ganglion cysts
may produce signs and symptoms mimicking those of a rotator cuff tear.
Experience of the reader and the quality of the equipment limit the
study’s accuracy. Furthermore, long study times and the confined space
in the machine may be intolerable to some patients.

Cadaveric studies have estimated a prevalence of
full-thickness rotator cuff tears ranging from 7 to 40%. Radiographic
imaging studies of asymptomatic patients have shown that rotator cuff
tears are an age-related phenomenon with a significantly higher
prevalence in individuals older than 60 years of age. Of course, the
tears of clinical significance are those that become symptomatic.

Although the reasons why some asymptomatic tears develop
symptoms remain unclear, recent study has suggested that many
asymptomatic tears will become symptomatic as well as demonstrate
progression of tear size (Figure 12-33).
Furthermore, studies have shown that rotator cuff tears do not heal
spontaneously. Over time, irreversible tendon and muscle degeneration
can occur with profound effects on the likelihood of success with
treatment.

Therapeutic decision making in the treatment of rotator
cuff disorders should be based on knowledge of the natural history of
the disease and how a particular intervention may alter its course.

Patients presenting early with outlet impingement and
intact cuffs with tendinosis or partial-thickness tears are almost
universally treated with an initial course of nonoperative therapy.
Inflammation is controlled with activity modification, oral
anti-inflammatory medication, judicious use of subacromial
corticosteroid injections, and physical therapy aimed at regaining
motion and strengthening of the rotator cuff and periscapular
musculature. Although studies have suggested that most
partial-thickness tears will continue to enlarge with time, the role of
operative treatment in modifying disease in this stage is unclear.
There is no universal agreement that surgical debridement of a
partially torn cuff relieves pain or stimulates a sufficient healing
response. Furthermore, the permanent changes in tendon and muscle
quality associated with larger, more chronic tears are not present.
Thus, although nonoperative care runs a risk of prolonged pain and
progression to larger tears, irreversible injury due to delayed
surgical intervention is unlikely.

Surgical care should be considered with failure of
nonoperative treatment in these early stages after a course of
approximately 6 to 12 months. Treatment includes anterior acromioplasty
for impingement lesions and debridement or tear completion and repair
of partial tears. The rationale for removal of the anterior acromial
spur is to eliminate the pathologic compression of the cuff. As
previously mentioned, tear debridement has been met with mixed results.
Tear completion and subsequent repair has been advocated in more
extensive partial tears (more than 50% of the cuff thickness) with
significantly better results reported as compared to debridement alone.

Treatment of symptomatic full-thickness tears is based
on several factors including severity and duration of symptoms, patient
activity level and goals, and tear size, chronicity, and location.
Although a trial of nonoperative treatment is not unreasonable with
smaller, minimally symptomatic tears or massive, chronic tears with
irreversible tendon and muscle degeneration, anterior acromioplasty and
rotator cuff repair remain the mainstays of treatment for most
symptomatic full-thickness lesions (Figure 12-34).
Results have been shown to be good or excellent in as many as 95% of
patients with preoperative tear size as the primary predictor of
outcome. Pain relief is a reliable goal with less consistent return of
function with larger tear sizes. Some authors have recommended the use
of tendon transfers in the presence of larger, chronic, retracted tears
with associated fatty degeneration and muscle atrophy. Transfers of the
latissimus dorsi or teres major for infraspinatus and supraspinatus
loss and pectoralis major transfer for irreparable subscapularis
ruptures have been described for massive tears and in those who have
undergone a failed primary repair with local tissue.

Although conventional open treatment for rotator cuff
repair and acromioplasty remains the standard, less invasive techniques
are becoming more widely used. Arthroscopic subacromial decompression
in conjunction with arthroscopically assisted or mini-open repair or
all-arthroscopic rotator cuff repair offer distinct advantages over
traditional methods with indications identical to those for open
techniques. These advantages include smaller incisions, avoidance of
deltoid detachment, shorter hospital stays, and ability to inspect the
glenohumeral joint and treat pathologic entities such as
articular-surface partial cuff tears, labral tears, biceps lesions, and
arthritic lesions of the glenoid and humeral head. These methods are
technically demanding but early results are promising with patient
satisfaction approaching that or equal to formal open procedures (Figure 12-35).

Postoperative rehabilitation is extremely important to
the success of all of the surgical treatment options. Early passive
motion is instituted and continued for approximately 6 weeks at which
time patients begin active range-of-motion exercises. Strengthening
exercises are added at about 3 months from surgery and continue until 1
year when the patient can expect maximal improvement.

Because of the proximity of the biceps tendon to the
rotator cuff, degeneration of the biceps tendon is thought to occur
much in the same way as mechanical

impingement
of the cuff beneath the coracoacromial arch. The role of other factors
such as failure from tensile overload or vascular insufficiency is
unclear although pathologic findings of the distal tendon have been
found with normal segments proximally where mechanical degeneration
would be expected (Figure 12-36).

Disorders of the biceps origin have gained more
attention with increasing recognition of this entity using arthroscopy.
The term SLAP lesion (superior labrum anterior and posterior lesions) was introduced by Snyder et al. to characterize injuries of the superior labrum and biceps anchor (Figure 12-37).
These lesions have been postulated to occur in various ways including
sudden deceleration of the flexed elbow in the follow-through stages of
throwing and shearing of the glenoid rim from a fall on the
outstretched hand. Disruption of this complex has been shown to
increase humeral head translation and secondarily compromise important
static restraints to shoulder stability.

Pain is typically the presenting complaint of biceps
tendinopathy. Anterior shoulder pain at the bicipital groove may be
found while associated cuff disease or glenohumeral instability may
produce a vaguer, diffuse pain. Painful clicks with shoulder motion may
indicate disorders of the labrum and biceps origin.

Provocative testing may help isolate symptoms to the
biceps tendon. Speed’s test is performed by elevation of the extended,
supinated arm against resistance. Yergason’s test may elicit symptoms
by having the patient supinate the forearm against

resistance
with the elbow flexed to 90°. O’Brien’s test (active compression test)
is utilized to isolate SLAP lesion-related pain from other sources. The
shoulder is internally rotated, flexed to 90° and adducted 10 to 15°.
Pain that occurs with resisted elevation in this position and
diminishes with testing in external rotation signifies a positive test.

As degeneration of the tendon progresses, rupture of the
long head of the biceps may occur with retraction of the distal segment
through the bicipital groove. The loss amounts to little more than a
cosmetic deformity in the arm (Popeye muscle) and a minimal loss in
supination and elbow flexion strength (Figure 12-38).
In fact, removal of the diseased intra-articular portion of the tendon,
either through spontaneous rupture or surgical resection, generally
results in improvement of pain.

Biceps pathology rarely occurs in isolation, and as
previously mentioned, may be associated with rotator cuff disease or
glenohumeral instability. Initial conservative treatment should address
these underlying disorders in conjunction with alleviating the acute
symptoms. Oral nonsteroidal anti-inflammatory medication and avoidance
of exacerbating activities followed by a physical therapy regimen
emphasizing shoulder and periscapular strengthening may provide relief.

If symptoms persist, surgical options may be considered.
Options for operative care entail debridement, tenotomy, or tenodesis
of the biceps tendon. Tenotomy simply involves transection of the
intra-articular portion of the tendon, either arthroscopically or
during an open procedure for

concomitant
pathology. This approach is particularly attractive for older, lower
demand patients who are willing to accept the resultant deformity and
prefer a faster return to activity, although tenodesis may be
preferable in the younger, more active individual.

Several methods of tenodesis have been described from
suturing the tendon back on itself or to adjacent soft tissue
structures to locking the distal segment into the humeral head using a
bone tunnel and interference screw fixation. The effect of tenotomy on
the purported stabilizing function of the biceps is unknown. Isolated
biceps tendinopathy with no evidence of coexisting shoulder pathology,
such as impingement of glenohumeral instability, is rare. Results of
isolated biceps release (tenotomy) or tenodesis have been mixed
implying that there are various mechanisms of biceps-related pathology.

Disorders of the biceps origin and labrum are treated
arthroscopically. Simple fraying of the labrum with a stable biceps
anchor is typically treated with debridement. Detachment of the
superior labrum and biceps origin is best treated with reattachment to
the glenoid rim with bioabsorbable tacks or suture anchors to restore
the stabilizing effect on the glenohumeral joint. Tears of the biceps
tendon extending from the labrum are debrided, but may require repair
or tenodesis if a significant portion of the tendon is torn or
degenerated.

The idiopathic form of shoulder motion loss due to capsular contractures and adhesions was termed adhesive capsulitis
by Neviaser based on the observed pathoanatomy. The pathogenesis of
this disease remains unclear with several systemic disorders, such as
diabetes mellitus, cardiovascular disease, mastectomy, or other
operations about the chest and shoulder, found in association with this
form of stiffness. Patients with diabetes mellitus, in particular, are
more prone to bilateral involvement that is extremely resistant to all
forms of treatment.

Capsular contracture may also result secondary to an
underlying disorder such as voluntary immobilization from rotator cuff
tendonitis or tear or bicipital tenosynovitis. These forms of secondary
stiffness and pain are often difficult, if not impossible, to
clinically discern from the primary or idiopathic type.

In contrast, shoulder stiffness and pain due to trauma
or surgery will typically result in scarring of the extra-articular
tissues and tissue planes. Fractures of the proximal humerus, treated
operatively or nonoperatively, or surgery for instability, for example,
may cause shortening or contracture of the pericapsular tissues and
present with a clinical picture very similar to other etiologies.

Demographically, patients with idiopathic stiffness are
usually females between the ages of 40 and 60 years with the
nondominant arm more commonly involved. As mentioned previously,
patients with diabetes mellitus tend toward bilateral involvement.
Motion loss must be carefully documented to assess severity and track
progress with both active and passive motion deficits found on
examination. Patients may exhibit compensatory increases in
scapulothoracic motion or trunk lean and should be carefully identified
to accurately measure pure glenohumeral motion loss (Figure 12-39).

Detailed history is essential and may reveal concomitant
disorders or potential etiologic sources. History of systemic disease,
trauma or previous surgery, or a long-standing rotator cuff tear with
pain and normal motion at a prior office visit may provide a likely
cause for the current stiffness. Similarly, in patients with a
noncontributory history, symptom patterns may suggest or mimic other
underlying disorders such as rotator cuff disease, although the
stiffness and associated pain often may mask other clinical findings.

Patterns of motion loss may vary with different causes
of stiffness. Adhesive capsulitis or capsular stiffness secondary to
other processes typically result in global loss of motion. Postsurgical
or posttraumatic

stiffness
may present with motion loss in all planes or with stiffness in some
planes although sparing other ones. Discriminating between these types
will help guide specific treatment.

Plain radiographs of the shoulder are typically not
helpful in diagnosis or determination of cause in frozen shoulder but
may identify other associated abnormalities such as fracture,
arthritis, metal implants, or neglected dislocations, which may
contribute to the clinical findings.

Historically, arthrography has been used to confirm a
diagnosis of adhesive capsulitis with obliteration of the axillary fold
seen with contrast infusion. However, little correlation has been
demonstrated between arthrographic findings and motion loss. In
addition, the invasive nature of the test has decreased its utility.

MRI may be used in cases when other associated disorders
are suspected. Rotator cuff or labral tears can be readily detected.
Use of MR imaging to define specific pathologic findings consistent
with capsular contracture such as capsular or synovial thickening and
decrease in capsular volume, however, has been inconclusive.

The natural history of adhesive capsulitis has been
described as self-limiting with resolution of symptoms occurring over 1
year. However, most patients are unwilling to endure the discomfort for
this period of time without intervention. The initial management of
most cases of frozen shoulder consists of physical therapy emphasizing
range of motion and gentle stretching every day with the majority of
patients responding to nonoperative treatment (Figure 12-40).

For those refractory cases demonstrating little or no
improvement in symptoms over several months of therapy, manipulation
and arthroscopic capsular release should be considered. Traditionally,
manipulation alone with anesthesia has been performed to rupture the
adhesions. The role of arthroscopy as an adjunct to manipulation in the
treatment of frozen shoulder has been recently described. Some authors
have adamantly discouraged its use in the treatment of this disorder
although others have had encouraging results, noting the additional
benefits of being able to release rather than tear the capsule, treat
coexisting pathology, and document the results of manipulation.
Nonetheless, essential to any operative form of intervention is the use
of adequate postoperative analgesia and aggressive therapy. Use of
interscalene brachial plexus block regional anesthesia has been used to
facilitate motion in the immediate postoperative period.

Patients with severe osteopenia, those who have
undergone tendon or fracture repair in the previous 3 to 6 months, or
have suspected or known extra-articular contractures (such as after
previous Putti-Platt or Bristow procedures for instability) should
undergo open release.

Crucial to all forms of treatment is early recognition
and realization that several months may elapse before observing
significant gains in motion.

Multiple structures are involved in maintaining
stability of the shoulder. The balance between stability and permitting
a wide range of motion is provided by the interaction of dynamic and
static factors. The static stabilizers include the glenoid, labrum,
capsule, glenohumeral ligaments, and the rotator interval. The role of
the biceps tendon as a static stabilizer is unclear but is also thought
to contribute to glenohumeral joint stability.

The glenoid provides a small, shallow surface to
articulate with the humeral head and provides little constraint for the
glenohumeral joint. The fibrocartilaginous labrum attaches to the
glenoid rim and increases its effective depth and surface area.
Isolated labral deficiency has been shown not to allow glenohumeral
dislocation without associated injury to the capsule, emphasizing the
crucial role of the capsuloligamentous structures in maintaining
stability.

The three major glenohumeral ligaments function as
“check-reins” toward the extremes of motion while remaining relatively
lax in the midrange to allow normal joint translation. The superior
glenohumeral ligament, coracohumeral ligament, and the rotator interval
(between the leading edge of the supraspinatus and the superior edge of
the subscapularis) restrain anterior humeral head translation in 0° of
abduction and external rotation. With increasing abduction to 45°, the
middle glenohumeral ligament provides the primary anterior restraint.
Finally, the inferior glenohumeral ligament tightens and becomes the
prime anterior stabilizer at 90° of abduction and 90° of external
rotation.

The rotator cuff and scapular stabilizers serve as
dynamic restraints in normal shoulder biomechanics. A primary role of
the rotator cuff is to resist translational forces on the joint through
compression of the humeral head into the glenoid cavity. Scapular
winging—an imbalance of the scapular stabilizing musculature—has been
implicated in pain and instability of the glenohumeral joint. Operative
intervention addressing scapulothoracic dysfunction may lead to
elimination of symptoms in select cases.

Critical to the evaluation of glenohumeral instability
is a careful history and physical examination. The nature of the injury
surrounding the onset of symptoms should be determined and are
particularly useful in identifying the type of instability. Position of
the arm at the time of injury or circumstances that provoke symptoms
often indicates the direction of instability. Reproduction of a
patient’s symptoms in a position of abduction, external rotation, and
extension suggests anterior instability. Flexion, internal rotation,
and adduction, in contrast, would more likely point to posterior
instability. Voluntary control of instability must be carefully sought
out as this may change the ultimate course of treatment. Patients with
psychiatric disorders may utilize a concomitant ability to dislocate
the shoulder for secondary gain. Although operative intervention in
this situation would likely fail, treatment options exist for other
forms of voluntary subluxation. Surgery may benefit patients who can
subluxate the shoulder by placing the arm in provocative positions.
Biofeedback techniques, however, may help those patients who sublux
through selective muscular activation.

In determining the degree of instability, the history
should ascertain whether the initial and any subsequent episodes of
instability were elicited by high-energy trauma (such as violent
twisting or fall), minimal repeated trauma (such as throwing a ball),
or no trauma (such as reaching a high shelf). The type of reduction
required (i.e., was the shoulder self-reduced or did it require
manipulation by another person?) may also provide additional
information about the extent of joint laxity.

Detailed record of prior treatment should also be
obtained, including the type and duration of immobilization,
rehabilitative efforts, and previous surgeries. Knowledge of failed
interventions will help guide future treatment in the recurrent
dislocator.

Pain as an isolated symptom will not typically reveal
much useful information. Anterior shoulder pain may indicate anterior
instability as well as other common disorders including subacromial
impingement. Similarly, posterior shoulder pain is nonspecific and may
represent a range of pathology from instability to cervical spine
disorders. Location of the pain in combination with provoking arm
positions and activities, however, may aid in making a diagnosis of
instability. Altered glenohumeral kinematics in throwers, for example,
may result in posterior shoulder pain during late cocking (posterior
internal impingement). Patients may also report other symptoms
consistent with subtle shoulder instability. Rowe and Zarins described
a phenomenon termed the “dead-arm syndrome,” in which paralyzing pain
and loss of control of the extremity occurs with abduction and external
rotation of the shoulder.

Finally, determining the patient’s level of impairment
and activity level is important prior to formulating a therapeutic
plan. The different expectations of a sedentary patient with minimal
functional loss versus the high-performance athlete with pain and
apprehension may affect the type of prescribed treatment.

A thorough physical examination is equally essential in
making an accurate diagnosis and recommending the appropriate
intervention. Both shoulders should be adequately exposed and examined
for deformity, range of motion, strength, and laxity. Unreduced
dislocations may produce an abnormal contour to the shoulder in
addition to restricting active and passive motion (Figure 12-41). Demonstration of scapular winging may accompany

instability, particularly of the posterior type, and should be
considered a potential cause of symptoms. Generalized ligamentous
laxity may also contribute to instability and can be easily elicited
with the thumb-to-forearm and metacarpophalangeal hyperextension tests (Figure 12-42).
Operative reports and evidence of healed anterior or posterior scars
from previous instability repairs indicate what has been done and may
provide a rationale for the patient’s current symptoms. Limitation of
external rotation may accompany a prior anterior tightening procedure
(i.e., Putti-Platt or Magnuson-Stack).

Typically, there is a full range of motion with the
exception of guarding at the extremes as the shoulder approaches
unstable positions. Clinical suspicion should be raised, however, in
the patient older than 40 years of age who is unable to actively abduct
the arm after a primary anterior dislocation. It has been shown that a
high percentage of these patients have a concurrent rupture of the
rotator cuff with restoration of stability following repair.

Various basic provocative tests can be used to reproduce
the patient’s symptoms and confirm the diagnosis. In order to minimize
the effects of muscle guarding, these maneuvers should be performed
first on the unaffected side and then in succession of increasing
discomfort. The sulcus test evaluates inferior translation of the
humeral head with the arm at the side in neutral and in external
rotation. The scapula is stabilized with one hand while the other
applies a mild compressive force to

the
head in conjunction with a downward force to the arm. Significant
findings would include an increased palpable gap between the acromion
and humeral head compared to the opposite side as well as translation
below the glenoid rim. Performing the test in external rotation will
not reduce the gap when there is incompetence of the rotator interval (Figure 12-43).

Laxity can be further evaluated by anterior and
posterior drawer or load-and-shift tests. The proximal humerus is
shifted in each direction while grasped between the thumb and index
fingers. Alternatively, with the patient supine, the scapula is
stabilized while the humeral head is axially loaded and translated
anteriorly and posteriorly. Translation greater than the opposite
shoulder or translation over the glenoid rim indicates significant
laxity (Figure 12-44). Only translations that reproduce the patient’s symptoms are considered to demonstrate instability.

The anterior apprehension test is performed by
externally rotating, abducting, and extending the affected shoulder
while stabilizing the scapula or providing an anteriorly directed force
to the humeral head with the other hand. Significant findings include a
sense of impending subluxation or dislocation, or guarding and
resistance to further rotation secondary to apprehension. Pain as an
isolated finding is nonspecific and may indicate other pathology such
as rotator cuff disease. Jobe’s relocation test is done in the supine
position, usually accompanying the apprehension test. As symptoms are
elicited with progressive external rotation, the examiner applies a
posteriorly directed force to the humeral head. A positive test is
signified by alleviation of symptoms (Figure 12-45).

Posterior instability can be elicited with the posterior
stress test. As one hand stabilizes the scapula, a posteriorly directed
axial force is applied to the arm with the shoulder in 90° of flexion,
abduction, and internal rotation. Positive findings produce posterior
subluxation or symptoms of instability experienced by the patient.

Though the history and physical examination are the key
elements in patient evaluation, a series of radiographic studies may be
helpful in confirming the diagnosis and defining associated pathology.
AP radiographs in internal and external rotation, a lateral view in the
scapular plane (scapular-Y view), and an axillary view should be
obtained in the initial evaluation. A Hill-Sachs lesion (posterolateral
impression fracture) of the humeral head is best seen on the AP
radiograph in internal rotation (Figure 12-46)
or on specialized views such as the Stryker Notch or West Point
Axillary. Fractures or erosions of the glenoid rim can be detected on
an axillary or apical oblique view (Garth).

Other more specialized imaging studies are not routinely
obtained in the initial evaluation of instability but may be useful in
a preoperative workup. Computed tomography can assist in further
assessment of fractures and glenoid defects as well as detect subtle
subluxation of the humeral head. MRI and

MR arthrography can identify associated pathology of the labrum, glenohumeral ligaments, and the rotator cuff (Figure 12-47).
More recent radiographic modalities such as dynamic MR-imaging
currently have no defined indications but may become a useful adjunct
in evaluating glenohumeral instability.

Although the results vary with age and associated bone
and soft tissue injury, nonoperative treatment consisting of a period
of immobilization followed by rehabilitation is typically successful in
managing the majority of patients with glenohumeral instability. Early
studies of young (less than 20 years old), athletic patients, however,
found a recurrence rate as high as 90% after a primary dislocation.
Although subsequent studies have reported lower numbers, clearly the
risk for subsequent dislocations is higher with earlier onset of
instability.

The length and type of immobilization remains a matter
of debate. Several published series have advocated immobilization for a
few days to several weeks. However, studies by Hovelius and Simonet and Cofield
have found no difference in outcome from either the type or length of
immobilization. In general, younger patients (less than 30 years of
age) sustaining a primary dislocation are preferably immobilized for
approximately 3 to 4 weeks. Older patients, who have a smaller risk of
recurrent instability but a higher susceptibility to stiffness, may be
immobilized for shorter periods.

Rehabilitation efforts are aimed at strengthening the
dynamic stabilizers and regaining motion. Progressive resistive
exercises of the rotator cuff, deltoid, and scapular stabilizers are
recommended. Stress on the static restraints (i.e., capsuloligamentous
structures) should be prevented in the immediate postinjury period by
avoidance of vigorous stretching and provocative arm positions.

Failure of conservative management for glenohumeral
instability is an indication for proceeding with operative
intervention. Open procedures are currently the gold standard for
repair of the disrupted soft tissue shoulder stabilizers. Historically,
many of these procedures treated anterior instability through
nonphysiologic shortening or transfer of muscle or bone resulting in
significant loss of motion, postoperative glenohumeral arthritis, and
complications of loose or prominent hardware in the joint.

Modern techniques emphasize anatomic restoration of the
soft tissue structures. Based on the work of Perthes in 1906, Bankart
in 1923 popularized repair of the capsule to the anterior glenoid
without shortening of the overlying subscapularis. After modifications
to his original description, reconstruction of the avulsed capsule and
labrum to the glenoid lip is commonly referred to today as the Bankart
repair. Several capsulorrhaphy procedures have also been described to
address capsular laxity and the increase in joint volume. These
procedures allow tightening of the anterior capsule in combination with
reattachment of a capsulolabral avulsion.

Open treatment of posterior instability most commonly
addresses the capsular laxity seen with this problem. Similar to the
anterior procedure, a posterior capsulorrhaphy functions to reduce
capsular laxity and address associated posterior labral injury.
Difficulty in treatment of multidirectional instability of the shoulder
is, in part, due to improper recognition. Procedures involving anterior
capsular tightening fail to address the inferior and posterior
components of instability seen in this disorder. The inferior capsular
shift procedure, described by Neer and Foster, treats multidirectional
instability by simultaneously reducing volume in the anterior,
posterior, and inferior capsule.

Good results have been achieved with most open
techniques to treat these various problems. Successful outcomes greater
than 80 to 90% can be expected; however, results have been generally
less predictable for posterior and multidirectional instability
compared to anterior instability procedures.

In selected cases, current trends are moving toward
arthroscopic intervention. The goals of restoring the capsulolabral
anatomy and reduction of capsular volume remain the same as in open
procedures. Reduced morbidity from smaller incisions, elimination of
subscapularis detachment, avoidance of scarred and distorted tissue
planes in revision situations, and the ability to simultaneously
address multiple factors contributing to instability have made these
minimally invasive techniques more appealing.

Arthroscopic anterior instability repairs remain
controversial with high (up to 70%) failure rates reported for
transglenoid suture fixation. Improvements in patient selection,
technique, and implant and equipment design have decreased the
recurrence rate. Contraindications to arthroscopic repair include
patients with glenoid bone defects, engaging Hill-Sachs lesions, and
attenuated capsulolabral tissue. Athletes involved in contact sports
have been a relative contraindication, although some surgeons have
modified the exclusion criteria to include this subset of patients.
Improved surgeon technique with bone preparation and soft tissue
tensioning and the use of suture anchors have led to improved results (Figure 12-48).

Thermal energy has been used in arthroscopic instability
repairs, in part, due to its easy application. Thermal capsulorrhaphy
may be done alone or to tighten residual capsular laxity following
capsulolabral repair or capsular plication with sutures. Complications
following its use have included neurologic injury to the axillary
nerve, capsular necrosis and loss of capsular and glenohumeral ligament
integrity, and stiffness and loss of motion. Although proponents of
thermal capsulorrhaphy have found this technique a useful adjunct for
treating instability of the shoulder, long-term follow-up studies
advocating its routine use are not yet available. Rehabilitation
following open and arthroscopic instability repairs is identical. The
shoulder is generally protected with 4 to 6 weeks of sling
immobilization in internal rotation for anterior repairs and in slight
abduction and neutral rotation following posterior and multidirectional
instability procedures. Active hand, wrist, and elbow motion is
permitted followed by progressive range of motion exercises. Resistive
exercises are started between 6 to 8 weeks with return to sports at up
to 9 to 12 months postoperatively. Although there are no strict
guidelines, the postoperative therapy regimen must be tailored to each
particular patient accounting for the quality of the tissues, the
durability of the repair, patient reliability, and future demands on
the shoulder.