Juvenile Idiopathic Arthritis

Oligoarticular onset of JIA is defined as arthritis
affecting one to four joints during the first 6 months of disease.
There are two subcategories: persistent oligoarthritis affects no more
than four joints throughout the entire course of arthritis; extended
oligoarthritis affects a cumulative total of five joints or more after
the first 6 months of disease. A child will be excluded from this
diagnostic category if any of the following five conditions are met:
psoriasis in the patient or a first-degree relative; family history of
medically confirmed HLA-B27–associated disease in at least one
first-degree relative [ankylosing spondylitis, enthesitis-related
arthritis (ERA), sacroiliitis with inflammatory bowel disease (IBD),
Reiter syndrome, or acute anterior uveitis (AAU)]; HLA-B27–positive
male with onset of arthritis after 6 years of age; a positive RF test
on at least two occasions more than 3 months apart; or systemic
arthritis. The characteristics include the age at onset of arthritis.
The patterns of occurrence of the arthritis should also be noted: large
joints only, small joints only, limb predominance (upper, lower, both),
involvement of specific joints, and symmetry of arthritis. Finally, the
occurrence of anterior uveitis, and the presence of antinuclear
antibodies (ANA) or any human leukocyte antigens (HLA) class I and II
showing predisposing or protective alleles, should be taken into
consideration. This diagnostic subgroup will certainly contain some
children with psoriatic arthritis who have not yet developed psoriasis.
It will also exclude the few children with oligoarticular disease and a
positive RF. However, these children are likely to have an early onset
of RF-positive polyarticular arthritis, or at least would be predicted
to have a more prolonged and severe course, and therefore should be
excluded from the oligoarticular group.

Polyarticular-onset (RF-negative) JIA is defined as
arthritis affecting five or more joints during the first 6 months of
disease, with negative RF and no systemic arthritis. Descriptors for
RF-negative polyarticular JIA include age at onset, symmetry of
arthritis, presence of ANA, and occurrence of uveitis. Exclusions
include: psoriasis in the patient or a first-degree relative, patient
is male and HLA-B27 positive with

onset after 6 years of age, HLA-B27–associated disease in a first-degree relative, IgM RF, or systemic JIA.

RF-positive polyarticular JIA is defined as arthritis
affecting five or more joints during the first 6 months of disease,
associated with a positive RF test on two occasions at least 3 months
apart, and the absence of systemic arthritis. Descriptors include age
at onset of arthritis, symmetry, presence of ANA, and immunogenetic
characteristics. Exclusions include: patient is male and
HLA-B27–positive with onset after 6 years of age, psoriasis in the
patient or a first-degree relative, HLA-B27-associated disease in a
first-degree relative, or systemic JIA. This disease is likely to be
the equivalent of early onset adult rheumatoid arthritis.

Systemic JIA is characterized by arthritis with or
preceded by daily fever of at least 2 weeks’ duration, which is
documented to be quotidian for at least 3 days, and accompanied by one
or more of the following: evanescent, nonfixed, erythematous rash;
generalized lymphadenopathy; hepato- or splenomegaly; and serositis.
This type of arthritis can be described in terms of age at onset of
arthritis and the pattern of arthritis during and after the first 6
months: oligoarthritis, polyarthritis, or arthritis present only after
the first 6 months of systemic illness. Other descriptors include the
features of systemic disease after the first 6 months of disease and
the level of C-reactive protein (CRP). Exclusions include: psoriasis in
the patient or a first-degree relative, patient is male, HLA-B27
positive with onset after 6 years of age, HLA-B27–associated disease in
a first-degree relative, or IgM RF.

Psoriatic arthritis is defined as arthritis and
psoriasis or as arthritis and at least two other criteria: dactylitis,
nail abnormalities (pitting or onycholysis), or a family history of
psoriasis in at least one first-degree relative. Exclusions include the
presence of RF and systemic arthritis. Descriptors include: the age at
onset of arthritis or psoriasis, the pattern of involvement of the
joints, oligoarticular or polyarticular course, presence of ANA, and
uveitis. Exclusions include: patient is male, HLA-B27 positive with
onset after 6 years of age, HLA-B27–associated disease in a
first-degree relative, IgM rheumatoid factor, or systemic JIA.

ERA is defined as arthritis and enthesitis (pain or
tenderness at insertion sites of tendons, ligaments, or fascia to
bone), or either arthritis or enthesitis along with at least two of the
following characteristics: sacroiliac joint tenderness and/or
inflammatory spinal pain, presence of HLA-B27, family history of at
least one first-degree relative with a medically confirmed
HLA-B27–associated disease (e.g., ankylosing spondylitis,
enthesitis-related arthritis, sacroiliitis with IBD, Reiter syndrome,
or AAU), or onset of arthritis in a boy after the age of 6 years.
Descriptors for ERA include age at onset of arthritis or enthesitis,
patterns of arthritis, symmetry of arthritis, oligoarticular or
poly-articular disease course, and the presence of IBD. Exclusions
include: psoriasis in the patient or a first-degree relative, IgM
rheumatoid factor, or systemic JIA.

Despite our best attempts to categorize children with
idiopathic arthritis, there will inevitably be children who do not fit
into any known category. This group of children with JIA will be
considered to have undifferentiated, or overlap, arthritis. This
category of other arthritis is defined as children with arthritis of
unknown cause that persists for at least 6 weeks, but that either does
not fulfill criteria for any other category, or fulfills criteria for
more than one of the other categories. The most common cause for the
exclusion from specific categories is the presence of psoriasis in a
first-degree relative (13).

The time of occurrence of the pain is important.
Children with arthritis will frequently have mild-to-moderate pain and
stiffness (gelling) in the mornings, after a nap, after sitting in the
classroom, after a long automobile ride, or other periods of
inactivity. These complaints will generally improve within minutes or
hours of renewed activity. Although there may be some residual limp or
mild complaints of pain or stiffness, many children with JIA will be
entirely asymptomatic on presentation to the physician’s office. The
major exception to this is the child with ERA, who may have morning
pain and stiffness, but who will worsen throughout the day or with
activities, because of repeated stress on inflamed tendons and
entheses. A minority of children with JIA will have significant pain
throughout the day, and the physical exam is likely to reveal severe
arthritis in those who experience increased pain. It would be unusual
for the pain associated with JIA to keep a child from falling asleep or
to wake the child from sleep. When a child has only afternoon, evening,
and/or nighttime pains that resolve in the morning, this is typical for
benign pains of childhood (growing pains). When pain is present both
day and night and interferes with sleep, it is less likely to be due to
JIA, but, depending on the duration and location, may be associated
with malignancy, infection, or psychogenic pain syndromes.

The duration of pain is also important. Severe pain of
new onset, which prompts immediate evaluation, may be due to infection,
malignancy, or trauma. Pain that is slowly worsening or unchanged over
weeks or months is typical of JIA. When pain has been present for many
years, the cause is often psychogenic or mechanical. It is not unusual
for a parent to report that a child has had pain, clumsiness, or
difficulty in walking from the time the child took his or her first
steps. If neurologic abnormalities, including cerebral

palsy, have been excluded, most of these children have only benign mechanical problems.

The location of the pain is an important part of the
history. Children with JIA will rarely have muscular pains. True
weakness should suggest either an inflammatory or congenital myopathy.
Long-bone or periarticular pain is often seen in trauma or
malignancies. When pain is random, intermittent, or migratory,
especially if it comes and goes during the clinic visit, it is often
psychogenic in origin. In JIA, pain in the joints can be additive or
can show spontaneous improvement, but it is usually present
consistently over weeks or months.

The quality or intensity of the pain can be difficult to
determine in children, especially in the preverbal child. The
pain-rating scales that have been developed for children are sensitive
to the cognitive-developmental conceptualizations of children (15).
Children are asked to rate their pain on a visual analog scale (VAS),
in terms of present pain and worst pain intensity for the previous
week. Each VAS is a 10-cm horizontal line with no numbers or
descriptors. The child VAS is anchored with the developmentally
appropriate pain descriptors: happy and sad faces, corresponding to no
pain and severe pain, respectively.

In general, the intensity of pain reported by children with JIA ranges from moderate to absent (16).
Rarely does the child report severe pain. Most children with JIA will
maintain near-normal functioning. Acute arthritis, such as
postinfectious reactive arthritis including acute rheumatic fever
(ARF), is typically much more painful than JIA. Intense pain that is
nonmigratory and present night and day is more likely to represent
infection or malignancy. Severe pain that is intermittent or migratory,
disrupts school attendance, and causes sleep disturbance in a child
with normal growth and normal parameters in laboratory tests is usually
psychogenic in origin. Children with psychogenic pain syndromes are
often unable to perform normal daily activities because of the high
levels of perceived pain.

The age of the patient is important. Most children with JIA have the onset of arthritis before 6 years of age (17). This age group is far less likely to have psychogenic pain syndromes. A frequent complaint on first presentation is that

a fall or injury caused a sudden swelling and pain in a joint. However,
most children younger than 6 years fall frequently, and children of
ages 1 to 4 fall nearly every day. Although young children may have
fractures from even innocent falls, accidental trauma should not be
considered as the most likely cause of true joint swelling in this age
group.

In children in whom it is possible that a significant
injury or repetitive stress has caused joint or tendon pain, resting or
immobilizing the limb is often tried. Although children with arthritis
can have minimal pain when the affected joint is immobilized, they are
uniformly worse when the cast or splint is removed. If the child has
pain in the joint accompanied by effusion and stiffness, prolonged
immobilization will often result in severe restriction in the range of
motion as well as increased pain. Childhood arthritis usually improves
with activities, and every attempt should be made to keep the children
mobile. We do not limit athletic activities, but we encourage them to
engage in noncontact sports. In children with psychogenic pain
amplification syndromes, the condition will frequently be worse after
immobilization, and this may initiate or exacerbate a flare-up of
reflex sympathetic dystrophy (RSD).

Other associated symptoms can often be useful in guiding
the evaluation of joint pain. Fevers, whether continuous or periodic,
may be associated with infection and malignancies or with systemic
inflammatory disease. Weight loss, with laboratory evidence of
inflammation, suggests the presence of a systemic illness, such as
infection, malignancy, IBD, or systemic-onset JIA. Persistent fatigue,
sleep disturbance, and depression are often signs of a psychogenic pain
syndrome. Fatigue should not be confused with true weakness, which
could represent the onset of an inflammatory myositis.

The family history is important in evaluating for
suspected JIA. The presence of HLA-B27-associated diseases, chronic
back pains, or psoriasis could suggest the onset of an associated
disease. Other systemic autoimmune disorders, including adult
rheumatoid arthritis, often cluster in families but have no direct
pattern of inheritance. Children with psychogenic pain syndromes are
often found to have a relative, usually the mother, who has chronic
pain and acts as a role model for the pain behavior.

A comprehensive physical examination, when combined with
a complete history and screening laboratory tests, will often be
sufficient to diagnose the child with arthritis. The musculoskeletal
examination will often enable differentiation between mechanical or
psychogenic causes of pain on the one hand, and inflammatory etiologies
on the other. Children with functional pain syndromes will often
complain of pain that is out of proportion to the examination and are
often migratory or transient, even during the examination.

There are no multitest panels that are appropriate for
evaluating children with pain and swelling of joints. The selection of
specific laboratory evaluations should be guided by the history and
physical examination. For most children, a complete blood count with
differential, CRP, and erythrocyte sedimentation rate (ESR) are
indicated. This will help to identify hematologic abnormalities
suggesting malignancy, and to document the presence or absence of
systemic inflammation. The CRP is an acute-phase protein synthesized in
the liver in response to proinflammatory cytokines. The ESR is an
indirect measure of systemic inflammation and the acute-phase response.
In most children, the ESR is less than 15 mm per hour. The ESR rises in
response to the relative decline in concentration of serum albumin and
increase in acute-phase proteins, including fibrinogen and others. The
ESR may be elevated because of marked anemia or a low serum
concentration of albumin due to decreased production or loss, as in
nephrotic syndrome. Most children with arthritis will have an ESR less
than 100 mm per hour, whereas systemic arthritis, malignancies, and
infections are more likely to be associated with an ESR greater than
100 mm per hour. However, many children with oligoarticular and some
with polyarticular arthritis will have a

normal
ESR and CRP. The addition of a CRP test can be helpful in situations in
which infection is strongly suspected, because the short half-life of
this acute-phase protein results in a rapid decline in concentration
with effective antibiotic treatment, whereas the ESR may even continue
to rise.

The ANA titer is a measure of serum antibodies that can
bind to one of many potential antigens present in the nucleus of normal
human cells. ANA titers are usually considered to be elevated when they
can be identified at a dilution of 1:40, or with an absolute value of
7.5 IU per mL. The presence of an elevated ANA by itself should never
be used for diagnosing arthritis. However, the ANA does have some
utility as a screening test for JIA (18,19).
The frequency of ANA positivity is greatest in younger girls with
oligoarticular disease, and represents an increased risk for anterior
uveitis (20). When arthritis is suspected on
the basis of history and physical examination, the presence of a
positive ANA should prompt an immediate referral to an ophthalmologist
for a slit-lamp examination to evaluate for the presence of uveitis.
Even in the absence of an ANA, children with confirmed arthritis should
have a routine ophthalmologic exam with a slit lamp. However, it is
known that elevated ANA titers may be present in up to 20% of healthy
children (typically, at titers of 1:40 to 1:80), and may be induced by
recent illness or be present in first- or second-degree relatives with
SLE (21,22). Children
who have an elevated ANA, whatever its level, with no evidence of
systemic inflammation and no arthritis on examination by a pediatric
rheumatologist, are extremely unlikely to subsequently develop a
significant autoimmune disease (21,23).

The RF is an autoreactive antibody, usually IgM,
recognizing IgG that has bound to antigen. RF positivity is infrequent
in children with arthritis, and rarely occurs in children younger than
7 years. When present in children with arthritis, the RF signifies a
chronic inflammatory state, and is associated with a higher frequency
of erosive synovitis and poor prognosis (24,25).
Studies in children and adults have demonstrated that a positive RF is
as likely to be present in children with diseases other than JIA as it
is in those with JIA (26,27).
Therefore there is no role for RF testing in the orthopaedic or
pediatric office evaluation of children with possible arthritis.

The presence of HLA-B27 is strongly associated with
transient reactive arthritis, IBD, and ERA. The high familial
occurrence of ankylosing spondylitis is directly related to the
presence of HLA-B27 (28). Although HLA-B27 is
found in approximately 8% of the white population, it can be useful in
the diagnosis of ERA. It is especially important in boys above the age
of 6, where there is a family history of HLA-B27–associated illness, or
sacroiliac joint or spinal inflammatory pain.

On rare occasions and in specific circumstances, children may develop gout (29).
However, there is no utility in obtaining uric acid levels as a
screening test for arthritis. The diagnosis of gout is made by
documentation of the presence of urate crystals in synovial fluid,
irrespective of serum uric acid levels.

Arthrocentesis with synovial fluid analysis and culture
should be performed in all children with an acute febrile
monoarthritis. The possibility of infection should also be considered
when a child with polyarticular arthritis has an acutely swollen and
tender joint, usually accompanied by fever, because this may represent
a secondary septic arthritis. The diagnosis and treatment of septic
arthritis are discussed in detail in Chapter 13.

Synovial fluid analysis in children with JRA usually
shows evidence of inflammation. However, the total white blood cell
count in joints without septic arthritis can range from 150 to greater
than 100,000 cells per mm³ (30, 31, 32), with average counts of between 10,000 and 12,000 cells per mm³ There is often a neutrophil predominance, with a range of 18% to 88% and average of 56% (30).
A synovial biopsy should be carried out if the tuberculin test is
positive, or if the diagnosis of sarcoidosis is being considered.

Plain radiographs are useful in the initial evaluation
of children with pain and/or swelling in the joints, predominantly for
identifying periarticular osteopenia, fractures, or other bony lesions.
In early JIA, there are no pathognomonic radiographic findings. The
diagnosis of JIA is typically made long before bony changes are
apparent. Ultrasound is often a rapid and noninvasive way to identify
an intraarticular effusion. Radionucleotide imaging with Tc 99m (bone
scan) to evaluate for osteomyelitis and malignancy will occasionally
identify other joints with subclinical inflammation, thereby suggesting
a diagnosis of JIA. Although rarely required for diagnosis, magnetic
resonance imaging (MRI) is the most sensitive technique for detecting
early articular changes in JIA (33, 34, 35).

When a child presents with a history of several weeks or
even longer of having swelling in one or more joints, the most likely
diagnosis is juvenile arthritis. Unless there is fever, weight loss or
severe pain, it is unlikely that the child has an infection (other than
tuberculosis or Lyme disease) or a malignancy. Minor trauma does not
ordinarily result in prolonged swelling. Most children with an internal
derangement of a joint will be identified by a history of significant
trauma.

Our typical evaluation includes a comprehensive history,
an ophthalmoscopic examination to rule out asymptomatic anterior
uveitis, and a comprehensive musculoskeletal

examination
of the affected joint(s) as well as all other joints which may be
affected though asymptomatic. We obtain a plain radiograph of the
affected joint(s), complete blood count, ESR, and ANA titer. In endemic
areas, one may consider ordering a PPD to screen for tuberculosis
and/or obtaining Lyme antibody titers. If the clinical presentation is
atypical, a bone scan is often indicated. All patients with continued
joint swelling should be started on a nonsteroidal antiinflammatory
drug, but corticosteroids should be withheld until the diagnosis of
juvenile arthritis has been confirmed. Immobilization of the affected
joint will cause loss of range of movement, and is counterproductive.

Children with an acute inflammatory oligo- or
polyarthritis will often present with a suddenly swollen and/or painful
joint. This is in contrast to most children with JIA, who (except those
with systemic onset) often have a subacute or insidious onset. Children
with injuries can often describe exactly when and where the injury
occurred, whereas children with benign pains will frequently have a
history of pain from the time they could walk. Conversely, children
with psychogenic pain can also frequently describe an event, minor or
major injury, or illness as the initiator of their pain. However, many
children fail to fit the expected profiles, and atypical presentations
often occur. It is important to evaluate children with joint and limb
pain with a broad differential diagnosis, which can be better defined
with a thorough history and comprehensive physical exam. The presence
or absence of fever, the age and sex of the child, and associated signs
and symptoms will aid the consultant in determining the optimal
strategy for the selection of diagnostic testing.

Septic arthritis generally affects a single joint and is
associated with fever, elevated neutrophil count, ESR, and CRP. This is
in contrast to monoarticular JIA, which seldom has significant systemic
inflammatory signs. However, gonococcal arthritis in sexually active
children can present with an oligoarticular, polyarticular, or
migratory pattern, with significant tenosynovitis. There are instances
in which organisms such as Staphylococcus aureus
can present with a subacute arthritis. However, this presentation is
most common for mycobacterial infections or Lyme disease. In most
cases, septic joints are extremely painful, but in JIA, swelling is
often out of proportion to the reported pain.

In early Lyme disease, the signs and symptoms of
infection include fever and migratory arthralgia, with little or no
swelling in the joint. Early localized disease is typically manifested
by the presence of erythema migrans, the classic expanding rash that
occurs most often at the site of the

tick bite and develops within 7 days to 1 month after infection (37).
Lyme arthritis occurs months to years after the initial infection. Many
patients with untreated Lyme disease will complain of migratory
arthralgias or arthritis (38). In a recent retrospective study of 90 children with Lyme arthritis, Gerber et al. (39)
noted that the majority (63%) had monoarticular disease, but no child
had more than four joints involved. The knee was affected most often
(90%), followed by hip (14%), ankle (10%), wrist (9%), and elbow (7%),
whereas small joints were rarely involved. Most children with Lyme
arthritis do not recall a tick bite or erythema migrans (39,40).
This is in contrast to prospective studies in which 90% of children
diagnosed with Lyme disease had a history of erythema migrans. The most
likely reason for this discrepancy is that most children with erythema
migrans are identified and treated with antibiotics, and do not develop
late complications of Lyme disease. Lyme arthritis is typically a
low-grade inflammatory synovitis with a large and relatively painless
joint effusion. The ESR can be normal or elevated, and 25% of the
patients may have values greater than 60 mm per hour (39). In both children and adults, a chronic form of Lyme arthritis can persist after treatment, and is associated with HLA-DR4 and HLA-DR2 alleles (41).
Most children with Lyme arthritis can be effectively treated with a
single, 4-week course of orally administered amoxicillin or doxycycline
(39).

Postinfectious or reactive arthritis results in a
sterile synovitis that is an immune response to a nonarticular
infection. In most children, the reactive arthritis occurs after upper
respiratory or gastrointestinal infections, whereas in adult patients
it is more likely to occur following a genitourinary infection (42, 43, 44).
The classic presentation of reactive arthritis is the triad of
conjunctivitis, urethritis, and arthritis found in Reiter syndrome
(RS). The complete triad of RS is very uncommon in childhood. Children
account for less than 1% of all patients with complete RS, and the
ratio of boys to girls is 4:1 (44,45). A history of sexual activity could suggest infection with Chlamydia (46). Most patients with classic RS and other postinfectious reactive arthritis carry the HLA-B27 allele (44,47).

Transient synovitis of the hip is a self-limiting,
postinfectious, inflammatory arthritis. Transient synovitis of the hip
has a peak incidence, predominantly in boys (70%), at between 3 and 10
years of age. It is an idiopathic disorder often preceded by a
nonspecific upper respiratory tract infection (48).
Trauma has frequently been reported as having occurred prior to
transient synovitis of the hip, and may be a predisposing factor (49).
The onset of pain is often gradual, may be focused in the hip, thigh,
or knee, and may last for an average of 6 days. Occasionally, transient
synovitis of the hip can be bilateral (4%). The child often presents
with inability to walk or with a severe limp. There is a loss of
internal rotation of the hip, and it is usually held in flexion,
abducted, and externally rotated. There is often low-grade fever. The
ESR and white blood cell count are normal to mildly elevated (50).
Plain radiographs often show no abnormality, or may show a small
widening of the joint space. Ultrasound is a sensitive and reliable
method to confirm the presence of an effusion (50).
With rest and non-steroidal antiinflammatory drugs (NSAIDs), most
children will have complete resolution of symptoms within 2 weeks. Most
children with transient synovitis of the hip will have only a single
event, with 4% to 17% having a recurrence usually within the first 6
months after the initial onset (49).

ARF occurs as a postinfectious reaction to infection of the oropharynx with group A β
hemolytic streptococcus. The incidence of ARF has remained relatively
constant at around 1 per 100,000 children between the ages of 5 and 17
years (51). It is very unusual for ARF to occur
in patients younger than the age of 4 years. Although ARF is rare in
developed countries, it remains the most common cause of acquired heart
disease in the developing world. In South Africa, the prevalence of ARF
has been estimated to be 690 per 100,000 (52).

Recently the entity poststreptococcal-reactive arthritis (PSRA) has been described, but many investigators are of the opinion that it is merely a variant of ARF (55, 56, 57).
PSRA typically presents as a nonmigratory oligo- or polyarthritis. It
is differentiated from ARF by the frequent presence of tenosynovitis
and the poor response to aspirin or other nonsteroidal drugs. In
addition to arthritis, other clinical manifestations include erythema
nodosum, livedo reticularis, cutaneous vasculitis, and systemic
polyarteritis nodosa (58,59).
Limited studies have suggested that further episodes of streptococcal
pharyngitis lead to an increased risk for ARF and rheumatic carditis,
and that streptococcal prophylaxis is indicated (55,57).
Among children with PSRA, there was found to be a statistically
significant increase in the frequency of occurrence of HLA-DRB1*01,
whereas those with ARF had a higher occurrence of HLA-DRB1*16, with the
occurrence of HLA-B27 being no different from the controls (60).
The association of PSRA with HLA-DRB1*01, but not with HLA-B27,
suggests that the pathogenesis of PSRA may be more similar to that of
ARF than to that of reactive arthritis. This would again support the
recommendation for prophylaxis for at least 1 year after the onset of
arthritis.

Serum sickness is a clinical syndrome resulting from an
adverse immunologic response to foreign antigens mediated by the
deposition of immune complexes. Although serum sickness was first
described after injection of heterologous serum, today the most common
causes are antibiotics (penicillins and sulfonamides) and viral
infections (61, 62, 63).
Serum sickness is characterized by fever, arthralgia or arthritis,
lymphadenopathy, cutaneous eruptions (urticarial or morbilliform), and
angioedema. Both serum sickness and allergic angioedema can be mistaken
for acute-onset juvenile arthritis. However, most children with serum
sickness will spontaneously improve within a few days to weeks. For
mild disease, removal of the offending antigen and treatment with
antihistamines and non-steroidal antiinflammatory medications is
sufficient. In severe cases, a several-week course of corticosteroids
may be required.

Children with congenital coagulopathies (hemophilia) and
hemoglobinopathies (sickle cell disease) will present with acute pain
and swelling in the joints, resulting from hemarthrosis and localized
ischemia, respectively. A comprehensive discussion of these conditions
is found in Chapter 11.

Leukemia in childhood frequently presents with musculoskeletal pain and arthritis (78,79).
Although effusion can occur in a joint, the pain is usually localized
to the metaphyses of the long bones. The pain in children with
malignancies is typically more severe than in JIA and will frequently
be continuous. Another frequent feature of children with malignancies
is an extreme elevation of the ESR (often >100), whereas in JIA, the
ESR is usually only moderately elevated, and may even be normal.
However, a normal ESR does not exclude malignancy. Plain radiographs
may show subperiosteal elevation, osteolytic reaction, or metaphyseal
rarefaction. In a recent study of 29 children

with
malignancy who were referred to pediatric rheumatologists, features
suggestive of malignancy included nonarticular “bone” pain (68%), back
pain as a major presenting feature (32%), bone tenderness (29%), severe
constitutional symptoms (32%), and atypical clinical features (48%) (80).
Atypical features included night sweats (14%), ecchymoses and bruising
(14%), abnormal neurologic signs (13%), and abnormal masses (7%).
Children with malignancy were more likely to have the combination of an
elevated ESR and lactate dehydrogenase (LDH) with a low platelet count
(28%).

Pigmented villonodular synovitis (PVNS) is a benign
tumor of the synovium. Although PVNS is rare in childhood, it does
frequently lead to recurrent joint swelling (81,82).
This usually results in recurrent effusions that are minimally painful,
with progressive cartilage destruction and erosion of bone. On
aspiration of the joint, a chocolate brown synovial fluid is frequently
found. The diagnosis is often confirmed by synovial biopsy showing
nodular hypertrophy, with proliferating fibroblasts and synovial cells,
and hemosiderin-laden macrophages. Surgical excision can be curative.
However, many patients have recurrences, and occasionally multifocal
disease can occur.

Historically, back pain in children was thought to be unusual (83).
Although back pain may be psychogenic in origin, this symptom,
particularly in young children, has been thought to be the result of a
serious underlying organic disorder. However, there are multiple
potential pathologic causes of back pain in children and adolescents (Table 12.5).
Many of these entities are discussed elsewhere in this text and are
relatively uncommon in young children. However, benign back pain is
increasingly recognized as a common occurrence in pediatric practice.

Recent studies have actually found that back pain in children is a common problem that is infrequently reported to physicians (84). In a study of 648 children with spinal disorders, no organic cause was identified in 57% of those who reported back pain (85).
The overall incidence of benign mechanical back pain seems to be
increasing. A large epidemiologic study found that the prevalence of
back pain in 12-year-old children was 7% and by the ages of 18 years
(females) and 20 years (males) more than 50% had experienced at least
one episode of low back pain (86).

Several factors may be contributing to the increased
incidence of back pain episodes. This may include the relative
deconditioning of today’s young people. Several recent studies have
implicated school backpacks as a cause of increased back pain in
children (87,88). These
authors have identified that a book bag weighing more than 15% to 20%
of the child’s weight is associated with back pain and improper use of
the backpack (one arm) can result in changes of posture and gait.

Most children have benign back pain that can be
diagnosed with a thorough history, comprehensive physical examination,
and selected imaging and laboratory studies. However, the orthopaedic
surgeon should be familiar with the associated signs and symptoms of
both benign and pathologic causes of back pain. Most patients with
benign mechanical back pain have intermittent symptoms that do not
interfere with sleep. These symptoms are typically improved with rest.
The pain is frequently located in the middle to lower back paraspinal
muscles with no limitation of motion. The most common causes of
significant pathologic low back pain include spondylolysis, Scheuermann
disease, and muscle or ligament injury. A history of specific trauma
may suggest a musculoskeletal origin for the pain. Continuous or severe
pain with localized point tenderness, with or without constitutional
symptoms, can be associated with infection or malignancy. Exacerbation
of the pain with a straight leg lift may suggest lumbar disc disease.
Flattening of the lumbar spine on forward flexion or tenderness over
the sacroiliac joints may represent JAS. Limitation of extension of the
cervical spine with morning stiffness and pain is frequently seen in
JIA. Osteoporosis typically causes generalized spine pain that becomes
worse with load-bearing.

If signs and symptoms point to benign back pain,
treatment is conservative with rest and decreased activity. Children
with back pain who do not improve with rest can be further evaluated
with plain radiographs, a complete blood count, and ESR. If these are
normal, a continued conservative approach is indicated. Patients with
suspected

benign
mechanical pain should be counseled on proper techniques for wearing
their backpacks, or advised to use rolling backpacks or to obtain two
sets of books in order to avoid carrying them back and forth between
home and school. Many patients will improve with a physical therapy
back exercise program along with a 2 to 4 week trial of an NSAID. If
muscle spasm is suspected, a trial of rest combined with a muscle
relaxant and NSAID is frequently beneficial.

For patients who do not improve, or who have atypical
findings in their history, physical examination, or laboratory tests,
additional imaging such as computerized tomographic (CT) scan, MRI,
and/or bone scan may be utilized. CT scan is most useful for
identifying abnormalities in vertebrae or erosive changes in joints. If
disc or spinal nerve abnormalities are suspected, MRI is the superior
imaging modality. A bone scan is useful in identifying local areas of
infection, malignant infiltration, or active inflammatory arthritis.

The worldwide incidence of juvenile arthritis ranges from 7 to 20 cases per 100,000 children at risk per year (1,6,103,104).
A study of the epidemiology of juvenile arthritis in Rochester, MN from
1960 to 1993 suggested that the incidence of arthritis in children may
be decreasing (6). However, the improved
recognition of Lyme disease, and the exclusion of psoriatic arthritis
and juvenile spondyloarthropathies from the definition of arthritis,
may be contributing factors in this finding.

The overall prevalence of JRA in the United States has
been estimated to be between 57 and 113 per 100,000 children younger
than 16 years (105). The prevalence rates of
childhood-onset psoriatic arthritis and ERA (spondyloarthropathies)
have not been as well studied. The prevalence of JAS is reported to be
2 to 10 per 100,000, whereas juvenile psoriatic arthritis has been
reported to have a prevalence of 2 to 12 per 100,000 (106).
The incidence of juvenile spondyloarthropathies in whites of Northern
European ancestry is slightly greater than 1 per 100,000 (103,107).
However, spondyloarthropathy is the most common form of juvenile
arthritis in some Mexican and North American Indian children (108,109).

The etiology of JIA is unknown. Chronic arthritis is a
complex disorder characterized by inappropriate immunologic activation,
with failure of self-tolerance, in the setting of multiple genetic and
environmental factors in the host. JIA is a heterogeneous disorder with
multiple ages and patterns of onset, and with a highly variable course.
It is likely that multiple initiating factors are involved, including
infection, trauma, and autoimmunity, all in conjunction with genetic
predilection for arthritis.

There are many reported associations between HLA types
and juvenile arthritis. Apart from HLA-B27, JIA has been associated
mostly with the HLA class II antigens, which are restricted to cells of
lymphoid origin (110). In oligoarticular
arthritis, there is an increased association with HLA-DR8, HLA-DR6, and
HLA-DR5, with relative risks of 2 to 27. This means that a child who
carries one or more of these genes has a 2- to 27-fold increased risk
of developing the disease compared to the population as a whole. The

presence of uveitis is correlated with HLA-DR5, whereas protection from uveitis is correlated with HLA-DR1 (110). Chronic uveitis has also been associated with HLA-DRB1 and HLA-DQA1 (111).
Polyarticular onset with positive RF is associated with HLA-DR4, which
parallels the association with adult rheumatoid arthritis, whereas
HLA-DR7 seems protective. RF-negative polyarticular disease is
associated with HLA-DR8, HLA-DPw3, and HLA-DQw4, with relative risk
factors of 3 to 10. Systemic-onset disease has overlapping risk
factors, showing associations with HLA-DR4, HLA-DR5, and HLA-DR8, with
relative risks ranging from approximately 2 to 7 (112).

JAS and related diseases show a striking familial
incidence. The only immunogenic factor in common in this class of
diseases has been shown to be HLA-B27. Data from multiple immunogenetic
studies have shown that 90% of patients with JAS express the HLA-B27
antigen (113,114).
These data are supported by an animal model in which spontaneous
inflammatory disease of the gastrointestinal tract, peripheral and
vertebral joints, male genital tract, skin, nails, and heart were seen
in transgenic rats that express a functional human HLA-B27 allele (115).

Imbalances in levels of proinflammatory and
antiinflammatory cytokines may be associated with chronic inflammation.
A polymorphism in the IL-1α gene was found to be associated with uveitis and pauciarticular arthritis in Norwegians (116). Children who have an IL-6 genotype, which has a relatively higher transcription rate when stimulated, may be at greater risk for systemic arthritis (117).

Infection has been implicated in both the onset and the exacerbation of chronic arthritis in children (118).
Of all types of JIA, systemic onset has clinical features most
consistent with an infectious process: acute onset, high fever, rash,
lymphadenopathy, and arthritis. However, there has to date been no
convincing laboratory evidence of infection in this relatively
homogeneous disease. Multiple viral and bacterial agents have been
associated with JIA (119). However, no single
or even large group of agents has been convincingly implicated in any
form of JIA. It is more probable that multiple conserved viral and
bacterial antigens, with epitopes that cross-react with human antigens,
may promote an inappropriate autoimmune response. This association is
strongest for the HLA-B27-associated diseases (120) in which arthritogenic peptides from enteric pathogens have generated specific B27-restricted CD8⁺ T lymphocytes; these lymphocytes have been found in arthritic joints (121).

The relative prevalence of each of the subtypes of
juvenile arthritis varies widely, depending on whether the EULAR or ACR
norms are utilized. Over many years, the rates of occurrence of the
various types at onset among children with JRA have been quite
consistent, with approximately 50% of children having oligoarticular
disease, 30% to 40% presenting with polyarticular disease, and 10% to
20% having systemic onset. Only recently has psoriatic arthritis been
separated from the spondyloarthropathies and differentiated from JRA.
The subtypes of JCA show similar figures for oligoarticular (50%) and
systemic onset (11%). However, the prevalence of polyarticular disease
is only 20%, and the remainder is divided among undifferentiated
spondyloarthropathy, JAS, juvenile psoriatic arthritis, and
inflammatory bowel disease–associated arthritis (103) (Table 12.6).
At the time of writing, the current summary of the prevalence of
individual subtypes utilizing the ACR criteria is from the Pediatric
Rheumatology Data Base (102) (Table 12.6). Of 2828 children with arthritis, 11%

had systemic onset, 24% polyarticular (RF-) 1% polyarticular (RF+) 38%
pauciarticular, 24% spondyloarthritis (11% JAS), and 3% psoriatic
arthritis. Each subtype of juvenile arthritis has individual
characteristics, and each type can have widely different courses and
outcomes, thereby further emphasizing the heterogeneity of JIA.

Systemic-onset juvenile arthritis (122) was first completely described by Still in 1897, and is therefore often referred to as Still disease. His classic article was reprinted in 1978 (123). Systemic arthritis is characterized by the presence of daily or twice-daily spiking fevers, usually to 39°C or higher (124).
Children with systemic arthritis frequently appear quite ill while
febrile. The fever often responds poorly to NSAIDs, but will typically
respond well to corticosteroids. In most children, the fever is
accompanied by a characteristic rash (125) (Fig. 12.3).
The rash consists of discrete, erythematous macules, which are
blanching, transient, and frequently nonpruritic. The rash is often
more pronounced on the trunk, but is often present on the extremities
and may occur on the face. Many children with systemic arthritis will
have extraarticular manifestations, including hepatosplenomegaly,
pericarditis, pleuritis, lymphadenopathy, and abdominal pain. The
extraarticular features may be present for weeks, months and,
occasionally, years prior to the onset of arthritis. Usually, the
extraarticular manifestations of systemic arthritis are self-limiting
and will resolve spontaneously or with corticosteroid therapy.
Occasionally, the pericarditis can result in tamponade. Systemic
arthritis can occur at any age, but is slightly more common before the
age of 6 years (103), and can occur rarely in adulthood, when it is referred to as adult-onset Still disease.
The condition occurs in both sexes in equal ratio, and this may support
the premise that there is an infectious trigger for systemic arthritis (126).

The laboratory features of systemic arthritis are
notable for elevated acute-phase reactants. The ESR and CRP are greatly
elevated. The disease is often accompanied by anemia of chronic disease
(127, 128, 129), a leukocytosis, and a marked thrombocytosis, which may exceed 1million per mm³ Clinical experience has shown that when the platelet count remains greater than 500,000 per mm³
after 5 years, remission is unlikely. An elevation in the level of
serum ferritin has been correlated with active inflammation in some
children with systemic arthritis (130).
Patients with systemic arthritis can have coagulation abnormalities,
with generation of fibrin split products that have also been correlated
with active disease (131). Children with systemic arthritis are rarely ANA- or RF-positive.

The differential diagnosis of systemic arthritis is
essentially the same as for fever of unknown origin. Systemic arthritis
often presents the greatest challenge to the clinician during the phase
prior to the onset of arthritis. The diagnostic possibilities that must
be considered include infections, malignancy, IBD, SLE, and
vasculitides (polyarteritis nodosa, Kawasaki disease).

Oligoarthritis is the most common subtype of JIA, and is
characterized by arthritis in four or fewer joints during the first 6
months of disease. These children rarely have complaints of pain, do
not have associated fever, and are not systemically ill. The knee is
the most common joint affected, followed by ankles and elbows. The hips
and the small joints of the hands and feet are seldom affected.
Asymmetric oligoarticular involvement of the small joints, with or
without large-joint arthritis, is most characteristic of psoriatic
arthritis. Most children with oligoarthritis present before 6 years of
age, with girls outnumbering boys in the ratio 4:1. Oligoarthritis can
present in older children, but this late-onset type, in which there is
a male predominance and high incidence of HLA-B27, should now be
classified as ERA.

Most children with oligoarthritis will have a mild
elevation of the ESR (rarely more than 80 mm per hour), but it can be
normal in some patients. The CRP is usually normal or mildly elevated.
ANAs are found in 40% to 80% of children with oligoarthritis, and are
associated with an increased risk for anterior uveitis. An RF is
generally absent in oligoarthritis. However, when an RF is present in
children with chronic oligoarthritis, it has been associated with an
aggressive and erosive disease (141).

The differential diagnosis of a child with monoarticular
arthritis depends on the duration of involvement of the joint. In
children with acute onset of pain and swelling in the joint, infections
(septic arthritis or osteomyelitis), trauma, hematologic causes of
hemarthrosis, and malignancy must be considered. These patients should
have a thorough evaluation, including an arthrocentesis. If the
arthritis is long standing, these causes are less likely. However, both
Lyme disease and mycobacterial infections can produce a prolonged
monoarthritis indistinguishable from JIA.

Polyarticular-onset JIA is characterized by the
insidious, but occasionally acute, onset of a generally symmetric
arthritis in five or more joints. It can involve both large and small
joints, and frequently affects the cervical spine and temporomandibular
joints. Typically, girls outnumber boys 3 to 1. Mild systemic features
may be present in children with polyarthritis. They may have low-grade
fevers, lymphadenopathy, and hepatosplenomegaly. The fevers are not
typically the high quotidian temperature spikes that are diagnostic of
systemic arthritis, and rash is rarely seen (8). There are at least two distinct subgroups of polyarthritis: those with and those without the presence of RF.

RF-negative polyarthritis can occur at any age, with the median age of onset at 6.5 years (103). This subgroup may be ANA positive (40% to 50%), and this is associated with an increased incidence of uveitis (5%) (144).

The second subgroup of polyarthritis includes those who
are positive for RF. This subtype occurs predominantly in older girls
(>8 YEARS) who are HLA-DR4 positive, and is indistinguishable from
adult rheumatoid arthritis. These children are more likely to have a
symmetric small-joint arthritis, rheumatoid nodules, and early erosive
synovitis with a chronic course. However, these children rarely develop
chronic uveitis.

Most patients with active polyarthritis will have an
elevated ESR, typically 20 to 80 mm per hour. The ESR is often a useful
measure of disease activity in children with polyarthritis (19,149).
Children with significant joint disease will often develop anemia of
chronic disease, with hemoglobin in the range of 7 to 10 g per dL,
although this is more marked in systemic arthritis (127,128).

The differential diagnosis of polyarthritis is quite different from that of monoarticular disease. Polyarticular septic

arthritis is unusual, although an asymmetric polyarthritis and tenosynovitis can be caused by Neisseria gonorrheae.
Systemic lupus should be considered, especially in adolescent and
preadolescent girls. Reactive arthritis, inflammatory bowel-related
arthritis, juvenile psoriatic arthritis, and ERA, including JAS, should
be considered. Although juvenile dermatomyositis and scleroderma may
present with polyarthritis, the associated signs and symptoms of these
disorders usually lead to a correct diagnosis.

The incidence of juvenile psoriatic arthritis was thought to be rare in children. Prior to 1982, when Shore and Ansell (152)
published the first large report of 60 children with psoriatic
arthritis, there had been fewer than 80 cases described in the
literature in the English language. The rarity of juvenile psoriatic
arthritis was unusual, given the relatively large number of children
with psoriasis, and the fact that 7% of adults with psoriasis have
arthritis (153). Juvenile psoriatic arthritis
has historically been thought of as a juvenile spondyloarthropathy.
However, recent studies have shown that juvenile psoriatic arthritis is
a distinct entity that has been underdiagnosed, often because of the
long period from onset of arthritis to onset of psoriasis (153,154).

Psoriatic arthritis may account for up to 7% of JIA.
There is a slight preponderance of girls (1.6 to 2.3 times as many) and
the disorder often affects young children, with the median age at onset
being 5.9 to 10.1 years. The arthritis is often an asymmetric oligo- or
polyarthritis affecting both large and small joints. At onset, patients
may have pitting of the nails (67%) (Fig. 12.6)
and a family history of psoriasis (69%) or dactylitis (39%), while less
than one-half of the children have the rash of psoriasis (13% to 43%) (103,153,154). Current criteria do not require the development of psoriasis to confirm a diagnosis of psoriatic arthritis (10) (Table 12.2).

Children with psoriatic arthritis usually do not develop
an RF, but a positive ANA can be seen in 50% of them and is a risk
factor for uveitis. HLA-DR1 and HLA-DR6 are statistically significant
risk factors for development of juvenile psoriatic arthritis (154).
There is a mild, but not statistically significant, increase in the
presence of HLA-B27 in children with psoriatic arthritis, and these
children are more likely to have axial arthritis (152, 153, 154).
In children younger than 5 years, the presentation is often
characterized by the involvement of a small number of fingers or toes
that are relatively asymptomatic, but leading to marked overgrowth of
the digit(s).

The differential diagnosis of psoriatic arthritis is
essentially the same as for polyarthritis. However, the diagnosis of
psoriatic arthritis should be considered in a child with dactylitis,
nail pitting, asymmetric involvement of large and small joints,
arthritis of the distal interphalangeal (DIP) joints (Fig. 12.6),
or a first- or second-degree relative with psoriasis. There is rarely
the kind of fever or systemic illness that may be seen in septic
arthritis caused by Neisseria gonorrheae.

The criteria for classification of ERA describe a group
of arthritides that includes undifferentiated spondyloarthritis,
ankylosing spondylitis, and inflammatory bowel disease–associated
arthritis. At the onset, juvenile spondyloarthropathies are often
undifferentiated, preventing the application of adult-onset criteria
for diagnosis. The addition of other criteria (the presence of HLA-B27,
a family history of HLA-B27–associated disease, and the onset of
arthritis in a boy after 6 years of age) will increase the number of
children included in this category (10).
However, in an effort to better define the group of children who have
psoriatic arthritis, the ILAR norms now exclude even those with
ankylosing spondylitis from the diagnosis of ERA, if they have a
first-degree relative with psoriasis. This contributes a significant
number of children to the “Other Arthritis” category in that they
either fulfill no criteria or fulfill criteria for more than one
category. It is probable that in families with a genetic propensity for
psoriasis, who also carry the HLA-B27
gene, there may be two distinct mechanisms that contribute to the
development of arthritis. These disorders will begin to be categorized
more accurately as the underlying mechanisms are further elucidated by
molecular and genetic research. The current criteria will include many
of the children who have been previously diagnosed with a syndrome of
seronegativity, enthesopathy, and arthropathy (SEA syndrome) who were
shown to be at increased risk for development of classic
spondyloarthritis or JAS (156,157).

ERA is often associated with enthesitis and arthralgias or arthritis long before any axial skeletal involvement is identified (157).
Enthesitis is identified when marked tenderness is noted at the 6, 10,
and 2 o’clock positions on the patella, at the tibial tuberosity, iliac
crest, or the attachments of the Achilles tendon or plantar fascia (72). In some children, the only manifestation of ERA may be severe enthesopathy of the heel(s) (158) (Fig. 12.7).

Laboratory evaluation of children with ERA yields
relatively unremarkable results. There is often systemic inflammation
with thrombocytosis and an elevated ESR. A highly elevated ESR (greater
than 100 mm per hour) is more likely to be associated with inflammatory
bowel disease in a child who meets the criteria for ERA. These patients
uniformly test negative for RF, but ANAs can be present in the same
proportion as in the general population of children (21,22,72).

The primary extraarticular manifestation of ERA is AAU, which can occur in up to 27% of children with ankylosing spondylitis (159). AAU has a strong association with the presence of the HLA-B27 antigen (50%) (160).
It typically presents with an acute, painful, red, photophobic eye. AAU
may resolve with no ocular residua, but some of the children will have
a persistent uveitis that is relatively resistant to therapy and can
result in blindness (161,162).

Children with JAS have often been diagnosed on the basis
of adult criteria that require radiographic evidence of sacroilitis.
JAS most often presents in late childhood or adolescence. Children with
JAS and sacroiliac (SI) involvement are often HLA-B27-positive (82% to
95%), and boys outnumber girls in the ratio 6:1 (106).
Most children who are ultimately diagnosed with JAS will initially have
had an episodic arthritis of large joints of the lower extremities and
the tarsal bones. Regardless of axial disease, the most reliable
predictors to differentiate JAS from oligo- or polyarticular JIA are
the presence of enthesitis and tarsal disease in children who have
arthritis of the lower, but not of the upper, extremities (163).

The presentation of JAS is most remarkable for the
absence of axial involvement. Only 12.8% to 24% of children with JAS
have pain, stiffness, or limitation of motion

of
the sacroiliac or lumbosacral spine at onset. A peripheral arthropathy
or enthesopathy, affecting predominantly the lower limb joints and
entheses, is seen in 79% to 89.4%. These children tend to have fewer
than five joints involved, and rarely more than ten. At presentation,
the pattern of involvement of the joints is usually asymmetric, or even
unilateral (164).
Small joints of the toes are commonly involved in JAS but are seldom
affected in other forms of JIA, with the exception of psoriatic
arthritis.

The diagnosis of JAS is often difficult to make at the
stage of onset. However, the combination of peripheral joint arthritis
with preponderance in the lower extremity, enthesitis, and SI or
lumbosacral disease, would strongly suggest the diagnosis of ERA and
possible JAS. Septic SI joint disease and osteomyelitis may present
with SI pain and limitation of motion.

Examination of the axial skeleton is important in order
to arrive at a diagnosis of JAS. Pain may be elicited over the SI
joints by direct pressure, lateral compression of the pelvis, or
distraction of the SI joints (Patrick test). Quantitation of the normal
lumbar spine flexion, by Macrae and Wright’s modification of the
Schober test (165), can identify children with
limitation of lumbar spine flexion. With the child standing upright, an
anchoring mark is made at the lumbosacral junction (dimples of Venus).
A mark is then made 5 cm below and 10 cm above the lumbosacral
junction. Then, with the child in maximal forward flexion, the distance
between the upper and lower points is measured. In general, a modified
Schober measurement of greater than 21 cm (i.e., an increase of 6 cm)
is within normal limits (72). The measurement
of fingertip to floor distance on forward flexion is not reproducible,
because it reflects both hip and back flexion and does not correlate
with the Schober index. Chest expansion is also not reliable as a test
for spine involvement in JAS (164).

The frequency of occurrence of arthritis in children
with IBD has been reported to be 7% to 21%, and it usually occurs after
the diagnosis of the bowel disease (170, 171, 172). Two different patterns of arthritis are seen (72).
The most common type is oligo- or polyarticular arthritis of the lower
limbs. This group is less likely to meet the criteria for ERA. This
arthritis is often episodic, with exacerbation lasting 4 to 6 weeks or,
rarely, for months. The activity of the peripheral arthritis often
reflects the underlying activity of the IBD, but may also be
independent of it. The less common type of IBD-associated arthritis is
an HLA-B27–associated oligoarticular arthritis of the lower limbs, with
sacroiliitis and enthesitis, and no relationship to bowel inflammation (72).
This form is more likely to persist and progress despite control of the
bowel disease, and the course seems to be identical to that in other
children with ERA.

There are many drugs available to treat arthritis. Often
two or more drug classes must be used simultaneously in order to
achieve disease control. Many recent advances in understanding the
mechanisms of inflammation in arthritis have led to novel therapeutic
strategies. The fundamental purpose of pharmacologic therapy is to
achieve pain control, decrease inflammation, and promote and maintain
remission. The medications used are individualized for each patient,
depending on their subtype of arthritis, degree of inflammation, and
previous response to medications.

The mechanism of action of NSAIDs is through the
inhibition of the biosynthesis of prostaglandins by direct action on
the enzyme cyclooxygenase (COX) (175). The
discovery of a second COX enzyme (COX-2), which is induced during the
proinflammatory cascade, and the differential inhibition of the two COX
isoforms (COX-1 and COX-2) by individual NSAIDs has provided the basis
for the development of safer NSAIDs (176). At
the time of this writing, no COX-2-specific NSAID has been approved for
use in children. In the near future, however, this exciting scientific
advance will likely change the way NSAIDs are used in the treatment of
children.

NSAIDs are the initial therapeutic intervention in most
children with JIA. NSAIDs provide both analgesia and an
antiinflammatory effect. The average time-course for response to NSAIDs
is 4 to 12 weeks (177). Therefore, an NSAID is
usually tried for 4 to 8 weeks before another is substituted if there
has not been sufficient improvement. NSAIDs are generally safe and well
tolerated in most children. Abdominal pain, nausea, and vomiting are
the most common side effects, and gastrointestinal hemorrhage is rare (178). However, gastroduodenal injury is more frequent in children who are receiving high doses, or more than one NSAID at a time (179).
The use of aspirin in JIA is no longer recommended because of the risk
of Reye syndrome, increased hepatotoxicity, bleeding, and
four-times-per-day dosing.

The doses of NSAIDs in children are based on body
weight, and are often proportionally greater than in adult rheumatic
diseases (Table 12.8). Preparations that come in a liquid form and have once- to twice-daily dosing are

preferred. In the United States, the most commonly used NSAID for JIA
is naproxen (10 to 20 mg/kg/day, b.i.d.). In children with fevers and
serositis associated with systemic arthritis and with JAS, indomethacin
is often the most effective NSAID (72).
Children on long-term NSAID therapy should have a complete blood count,
renal and liver function tests, and urine analysis every 6 months.

Nearly two-thirds of children with juvenile arthritis are inadequately treated with NSAIDs alone (180). These children require additional pharmacologic interventions.

Intraarticular corticosteroid injections have been shown to be safe and effective in controlling the synovitis in JIA (181,182).
Triamcinolone hexacetonide (1 mg per kg for large joints and 0.5 mg per
kg for medium joints) is the most commonly used agent and often
provides long-term control of inflammation. The most frequent adverse
consequence of intraarticular corticosteroids is the development of
subcutaneous atrophy at the site of injection. Systemic corticosteroids
can be used for rapid control of severe arthritis. However, long-term
use should be restricted to those children who have severe arthritis or
systemic features that do not respond to other interventions.

Methotrexate is the most commonly used second-line agent
for treatment of juvenile arthritis. It is typically given at a dosage
of 0.5 to 1 mg per kg (with a maximum of 20 to 30 mg) once weekly,
either orally or by subcutaneous injection. It has been shown to be
superior to placebo in poly-articular and extended oligoarticular
arthritis but not in systemic arthritis (183,184), and can produce radiologically evident improvement in erosions (185).
Methotrexate has been shown to decrease the severity of uveitis in
children with JIA who were dependent on topical cortico-steroids (186).

The major side effects of methotrexate are nausea,
diarrhea, and oral ulcers. Supplementation with folic acid (1 mg per
day) can usually prevent gastrointestinal complications. One of the
most significant long-term side effects of methotrexate use is the
development of liver fibrosis and cirrhosis (187, 188, 189).
Serial abnormalities of hepatic enzymes were significantly associated
with liver fibrosis in children taking methotrexate for juvenile
arthritis (190), thereby suggesting that the
current guidelines for patients with rheumatoid arthritis are also
applicable to patients with JIA (191).

Sulfasalazine has been used extensively in Europe, and
increasingly in North America, for treatment of both
spondyloarthropathies and JRA/JCA (192,193).
It is typically given in the enteric-coated form at a dose of
50mg/kg/day in two divided doses. Recently, a randomized, double-blind,
placebo-controlled trial showed that sulfasalazine is both safe and
effective for the treatment of oligo- and polyarticular JCA (194).
Serious side effects have been noted in children with systemic
arthritis, and the routine use of sulfasalazine is not recommended for
this subgroup (195,196).

Although the etiology and pathogenesis of juvenile
arthritis are still unclear, macrophage-derived cytokines, such as
tumor necrosis factor α(TNFα) appear to play a critical role in the
induction and perpetuation of the chronic inflammatory process in
juvenile arthritis. Etanercept (Enbrel) is a protein containing the
extracellular domains of 2p75 human TNF receptors attached to the Fc
portion of a type 1 human immunoglobulin. A multicenter study showed it
to be safe and effective in the treatment of juvenile arthritis (197).
The study showed that 81% of children treated with etanercept reached a
JRA 30% definition of improvement (DOI), whereas 67% had a JRA 70% DOI.
Infliximab (Remicade) is a chimeric, monoclonal anti-TNFα antibody
developed as a therapeutic agent for the treatment of diseases in which
TNFα is thought to mediate chronic inflammation. This antibody is a
recombinant IgG1κ human-murine chimeric monoclonal antibody that
specifically and potently binds to and neutralizes soluble TNFα and its
membrane-bound precursor. Infliximab has been shown to be highly
efficacious in combination with methotrexate for the treatment of
rheumatoid arthritis and, in small trials, for refractory juvenile
arthritis (198) and chronic inflammatory uveitis (199).
The major adverse events associated with the use of anti-TNFα agents
have been infections and an increased risk of contracting tuberculosis
and coccidiomycosis (200). Other
cytokine-specific monoclonal antibodies, including anti-IL6 for the
treatment of systemic juvenile arthritis, are in clinical trials at the
time of writing.

All children with prolonged arthritis should be
evaluated by a physical and/or occupational therapist to provide an
appropriate teaching and treatment program. Most treatment programs for
JIA will include active and passive range-of-motion exercises,
strengthening, and other modalities such as use of hot paraffin for
relief of hand stiffness. Swimming has the advantage of providing
muscle-strengthening and active range of motion without significant
weight bearing. Splinting may be used for maintaining alignment,
providing rest, and reducing flexion contractures. For children with
severe flexion contractures, a dynamic tension splint or serial casting
can be used to correct the contracture. Physical therapy for range of
motion in JAS is primarily to prevent loss of mobility and poor
functional positioning.

For most children with JIA, orthopaedic surgery has a
limited role in the management plan. With early detection and
aggressive medical management, including intraarticular corticosteroid
injections, the majority of children with juvenile arthritis have a
satisfactory outcome without significant disability. However, for those
children with persistent arthritis (treated or untreated), continued
pain, or progressive leg-length discrepancy, there is often significant
benefit from individualized orthopaedic surgical intervention. Many of
the reports of surgery for juvenile arthritis actually refer to adults
who have had arthritis since childhood. Surgical intervention in JIA
presents several problems to the management team. The small size of
children and their growth potential must be taken into consideration.
Also, in the postsurgical period, prolonged immobilization can lead to
decreased strength and range of motion, with or without active
arthritis. After a surgical procedure, intensive physical therapy will
often be required in order to mobilize the child’s joints. There is no
universal agreement about which procedures are indicated for the
treatment of complications of chronic arthritis in childhood. However,
the overall goal is to provide symptomatic relief and improved
functioning.

Synovectomy may be indicated in a few of the children
with JIA for relief of pain, swelling, and impaired motion of the joint
related to synovial hypertrophy. There may be short-term relief from
swelling and pain, but the range of motion may not improve or may even
worsen (201, 202, 203, 204, 205). The greatest benefit has been seen in large joints (206,207). But recurrences are common, and the ultimate outcome for children with JIA is not altered by prophylactic synovectomy (208,209). A recent study has confirmed that

elbow synovectomy in patients with JRA provided complete pain relief in
44% of patients, and 73% had a subjective good to excellent outcome (210). However, there was no significant improvement in range of motion or functional ability.

Soft tissue release may rarely be useful in a child with
a severe contracture of the knee or hip that has been resistant to
splinting or serial casting. Initially there were encouraging reports
regarding soft tissue release (211). However, more recent reports have been less positive, showing only a modest benefit and a tendency to deteriorate (212,213). In most cases, vigorous physical therapy will avoid the need for this frequently unsuccessful operation.

Arthrodesis is indicated for the treatment of severe
joint destruction of the ankle after prolonged synovitis in oligo- or
polyarticular JIA. After puberty, a fixed and painful deformity of the
ankle is best corrected by performing a triple arthrodesis.
Occasionally, in children with isolated damage of the subtalar or
talonavicular joint, a single joint fusion may be appropriate (214). If required, these procedures may later be converted to a triple arthrodesis.

Although many children with JIA have cervical spine
arthritis and atlantoaxial instability, there is no consensus on the
indications for prophylactic fusion. In many cases, a simple cervical
orthosis may stabilize the neck and prevent further subluxation.
However, fusion of the cervical spine (C1-C2) is indicated in children
who have progressive neurologic involvement (215,216).

Osteotomy has only occasional utility in children with
arthritis. In a younger child with a fixed deformity of the knee but
good remaining joint surface and minimal active synovitis, a knee
osteotomy may result in correction of the deformity (217).
Unfortunately, the osteotomy makes a subsequent total knee arthroplasty
more difficult to perform because of the distorted anatomy.

Infrequently, a shortening osteotomy of the radius or
lengthening of the ulna may be carried out to correct length
discrepancies of the forearm bones. Recently, distraction lengthening
of the ulna in six children (eight wrists), with severe destructive
changes in the wrist due to JCA, was reported (218).
This procedure was found to adequately correct the deformity and
improve functioning, and in most of the children no further splinting
was required during a follow-up of an average of 70 months (range 12 to
152 months).

An appropriately timed epiphysiodesis can be successfully used to correct leg-length discrepancies in oligoarticular arthritis (219,220). The discrepancy can be predicted using the method of Moseley (221). Simon et al. (220) reported that 15 such patients were followed up to skeletal maturity and showed satisfactory results.

Total joint arthroplasty is indicated for children with
JIA who have severe destructive joint changes with functional
impairment or disabling pain. The most common joints replaced are the
hip and knee, but there may be indications for shoulder and elbow
arthroplasty.

Typically, total hip replacement is performed when there
has been severe destruction or ankylosis resulting in functional
impairment. Initial series using predominantly cemented hip
replacements showed reduction in pain and improved functional ability,
but with a significant rate of loosening and subsequent revision (223,224).
Subsequent studies have suggested an improved outcome with cementless
arthroplasty of the hip, but poor bone stock remains an indication for
cementing (225,226). In
a recent prospective study of the outcome of total hip arthroplasty in
small-proportioned patients with JRA, it was found that the miniature
anatomic medullary locking femoral component combined with an
uncemented acetabular cup resulted in a good outcome (227).
A recent study has suggested that bipolar hemiarthroplasty of the hip,
with a 79% 10-year survival, may be an alternative to conventional
joint arthroplasty (228).

In JIA the knee is frequently involved, and when there
is significant pain, deformity, and functional incapacity, a total knee
replacement is indicated. Initial results of total knee arthroplasty in
JIA have been encouraging, with few revisions required (229, 230, 231).
A recent long-term follow-up of total knee arthroplasty in young adults
and children with arthritis has been equally encouraging (232,233). Cementless total knee arthroplasty has been used in selected cases (234).
Recent studies have confirmed the efficacy of the procedure by
reporting an overall 99% survival for nonconstrained anatomically
graduated components prosthesis with cementless fixation (233).

Total elbow replacement may be indicated in children
with severe destruction of the elbow joint. In a recent review, Connor
and Morrey (235) evaluated the long-term
outcome for 19 children (23 elbows) who had been managed with total
elbow arthroplasty and followed up for at least 2 years. Only three
(13%) had poor results caused by late complications: aseptic loosening,
instability, and worn bushings.

Shoulder replacement in juvenile arthritis is indicated
when there is prolonged pain, limitation of function, and significant
joint destruction. The available data are not sufficient for evaluating
the efficacy of total shoulder arthroplasty in children. However, in
studies of adults with rheumatoid arthritis, and also in some studies
that included children, the results have been promising (236,237).