Infectious and Inflammatory Arthritis

Epidemiology

The incidence of pyogenic (bacterial) arthritis is less than that of transient synovitis and varies substantially (i.e., 1–37 cases per 100,000 children) depending on the population studied. ^, Although pyogenic arthritis occurs in all age groups, the peak incidence of disease is in children <3 years of age.

A history of trauma temporally related to the onset of arthritis caused by Staphylococcus aureus is common. ^, Upper respiratory tract infection frequently precedes pyogenic arthritis caused by Haemophilus influenzae type b (Hib) and Kingella kingae. ^{, ,} Gastroenteritis and aphthous stomatitis also can precede arthritis caused by K. kingae. Although most children have no underlying disorder, risk factors for pyogenic arthritis include immunodeficiency, hemoglobinopathy, diabetes, intravenous drug abuse, and rheumatoid arthritis.

Pathophysiology

Most cases of pyogenic arthritis in childhood follow the hematogenous spread of organisms to the vascular synovium of the joint space. Animal models of Hib bacterial arthritis illustrate possible mechanisms of articular damage. Bacterial endotoxin in the joint space stimulates release of tumor necrosis factor and interleukin-1. ^, These cytokines stimulate production of proteinases by synovial cells and chondrocytes, enhancing leukocyte migration. Neutrophil elastases augment destruction of cartilage matrix in the joint.

Bacteria can spread to joints from contiguous osteomyelitis. The transphyseal blood vessels in the child <18 months facilitate spread of infection from the metaphysis across the growth plate to the epiphysis and adjacent joint space. The joint capsule of the hip and shoulder overlies the bony metaphysis of the femur and humerus, allowing direct extension of bone infection into these joint spaces. Except for pyogenic arthritis involving the hips or shoulders, that occurring in neonates, or caused by K. kingae , primary pyogenic arthritis rarely extends from the joint space into the bone. Joints can be secondarily infected from penetrating wounds, intra-articular injections of medications, arthroscopy, and prosthetic joint surgery.

Etiology

Age is the most important predictor of the cause of pyogenic arthritis. ^{, ,} S. aureus, enteric gram-negative organisms, and group B Streptococcus (GBS) are the most frequent causes of pyogenic arthritis among neonates. Methicillin-susceptible S. aureus (MSSA), community-associated methicillin-resistant S. aureus (CA-MRSA), K. kingae, Streptococcus pyogenes, and Streptococcus pneumoniae cause pyogenic arthritis in children <5 years of age. In one series, K. kingae was the most common cause of pyogenic arthritis in children <36 months of age. K. kingae is being reported with increasing frequency as a cause of pyogenic arthritis in the US. Hib infection has become rare among immunized immunocompetent children.

Although S. pneumoniae caused approximately 6% of cases of pyogenic arthritis before universal pneumococcal conjugate vaccination, cases of invasive disease due to vaccine serotypes have decreased dramatically. ^, S. aureus and S. pyogenes are the most common causes of pyogenic arthritis in children >5 years of age. CA-MRSA osteoarticular infections are common, are more aggressive than MSSA infections, and can involve multiple bones and joints. Infections caused by CA-MRSA also can be complicated by venous thrombosis and septic pulmonary disease. ^,

Other organisms reported to cause pyogenic arthritis in children include Salmonella spp., β-hemolytic streptococci other than serogroups A or B, and, rarely, anaerobic bacteria. ^,

Joint infections caused by Pseudomonas aeruginosa and Candida spp. are reported in intravenous drug abusers. Brucella spp. infection should be considered if a history of travel to endemic areas, contact with livestock, or consumption of unpasteurized dairy products is elicted. ^, Arthritis related to Bartonella henselae infection has been reported.

Clinical Manifestations

Fever, malaise, poor appetite, and irritability are heralding systemic symptoms. Pain in the affected joint usually occurs early in the course of the illness. As infection progresses, the joint becomes swollen and the overlying skin red. Limp or refusal to walk occurs with infection of a lower extremity. If the affected joint is in the upper extremity, pseudoparalysis or refusal to use the affected joint may be seen, and manipulation causes pain. The infected joint is swollen, red, warm, and tender on palpation. The range of joint motion is decreased.

The joints of the lower extremities, especially the knees, are the most common sites of pyogenic arthritis ( Table 76.1 ). ^, More than 90% of cases of pyogenic arthritis are monoarticular. ^, However, multiple joints can be involved, particularly with infections caused by Neisseria gonorrhoeae, N. meningitidis, and Salmonella spp.

The diagnosis of pyogenic arthritis of the hip can be difficult because there is often no obvious joint swelling, and signs and symptoms are nonspecific, especially in infants and young children. Infants with pyogenic arthritis of the hip are irritable when the hip is moved (e.g., during diaper changes). Soft tissue swelling around the hip joint occasionally occurs and can extend to involve the entire leg. The affected hip often is held in a flexed, externally rotated and abducted position. Older children with pyogenic arthritis of the hip limp or refuse to walk and complain of pain, which sometimes is referred to the knee. Range of motion of the hip joint is decreased markedly.

Diagnosis

The erythrocyte sedimentation rate (ESR) is >20 mm/hr (mean, 44–65 mm/hr) for most patients with pyogenic arthritis. ^, Similarly, the level of C-reactive protein (CRP) often is increased (mean, 8.5 mg/dL). ^, A normal CRP level has a good negative predictive value (NPV) for pyogenic arthritis. In one study, an NPV of CRP <1.0 mg/dL was 87% for pyogenic arthritis.

Excellent sensitivity for the diagnosis of osteoarticular infections is obtained by using both the ESR and CRP. However, children with pyogenic arthritis caused by K. kingae can lack fever or elevation of inflammatory markers or both.

Analysis of joint fluid helps to differentiate bacterial from other causes of arthritis ( Table 76.2 ). Joint fluid in bacterial arthritis typically has a cloudy appearance. A synovial fluid white blood cell (WBC) count of >50,000 cells/μL with a predominance of neutrophils strongly suggests bacterial infection, even if the culture result for the joint fluid is negative ^, ; however, a count of <50,000/μL can occur in cases of bacterial arthritis, particularly when infection is caused by K. kingae . ^{, , , ,} A fluid WBC count of >50,000/μL with neutrophil predominance can occur in children with juvenile idiopathic arthritis and Lyme disease. ^, Synovial fluid glucose and protein levels do not differentiate reliably among most infectious and inflammatory processes and have limited value.

Blood culture should be obtained and synovial fluid sent for Gram stain, culture, and WBC count. Isolation of K. kingae is enhanced when synovial fluid is inoculated directly into fluid blood culture medium. ^, Use of real-time polymerase chain reaction (rt-PCR) testing increases detection of K. kingae in culture-negative joint fluid in young children. ^, Similarly, serologic testing for Borrelia burgdorferi aids in the diagnosis of Lyme arthritis. Advanced molecular techniques such as 16S rRNA gene sequencing or shotgun metagenomic sequencing of joint fluid may be useful in identifying pathogens in culture-negative cases. ^, For adolescents, specimens should be obtained from the cervix or urethra, throat, skin lesions, and rectum for isolation of N. gonorrhoeae. Alternatively, nucleic acid testing of urine can be performed.

When appropriate cultures are obtained, the bacterial cause is confirmed in 60%–70% of cases of pyogenic arthritis. ^, Blood cultures are positive in 40% ^, of cases, and joint fluid culture is positive in 50%–60%. ^{, , ,}

Imaging Studies

Children with suspected pyogenic arthritis should have plain radiographic studies to exclude osteomyelitis or other osseous abnormalities. Soft tissue swelling and widening of the joint can be observed in children with pyogenic arthritis. Erosion of subchondral bone may be evident 2–4 weeks after onset of infection.

Swelling of the hip capsule and lateral displacement or obliteration of the gluteal fat planes are early radiographic findings for pyogenic arthritis of the hip. With continued swelling of the hip capsule, the femoral head is displaced upward and outward, and lateral subluxation can occur. Concomitant osteomyelitis of the femur may occur. These findings are particularly common in infants, although in this age group, radiographic findings are difficult to interpret because of minimal ossification of the proximal femur. A plain radiograph sometimes is normal for children with proven pyogenic arthritis of the hip.

Ultrasonography (US) should be performed in suspected pyogenic arthritis of the hip. If fluid is detected in the joint, a diagnostic aspiration should be performed under US guidance. ^, False-negative US results reported for children subsequently diagnosed with pyogenic arthritis are related to inadequate imaging or imaging performed very early (<24 hours) after the onset of symptoms.

Magnetic resonance imaging (MRI) is highly sensitive for the early detection of inflamed or infected joints. Abnormal MRI findings in pyogenic arthritis include periarticular high-intensity signal and periarticular abscesses in some cases. MRI can delineate abnormalities of adjacent bone and soft tissue and the extent of cartilage destruction. Compared with patients with transient synovitis, those with pyogenic arthritis are more likely to have MRI findings of high-intensity signals in the bone marrow and decreased signal in the femoral epiphysis on fat-suppressed, gadolinium-enhanced, T1-weighted images.

Management

The European, Canadian, and Australian clinical practice guidelines provide consensus-based recommendations on the management of pediatric bone and joint infections. Children with pyogenic arthritis should be managed in consultation with an orthopedic surgeon experienced in treating children. Goals of therapy include decompression, sterilization of the joint space, and removal of inflammatory debris.

All children with pyogenic arthritis of the hip require prompt surgical drainage and irrigation of the joint space. Delay in drainage increases the likelihood of permanent damage because increased intra-articular pressure can compromise blood supply, resulting in avascular necrosis of the femoral head. Muscle spasms also can occur, predisposing the patient to dislocation. Open surgical drainage of joints other than the hip usually is not required. However, aspiration must be performed promptly to decompress the joint and obtain synovial fluid for analysis. Repeated aspirations often are necessary when fluid reaccumulates. Concurrent osteomyelitis can be associated with the need for repeated débridement of the joint. Débridement by arthroscopy has been undertaken in some cases of pyogenic arthritis of the knee and hip. ^,

The initial choice of antibiotics is based on age, clinical history, and physical examination of the patient. Adequate penetration into the joint is essential. Nafcillin, dicloxacillin, first- and third-generation cephalosporins, clindamycin, vancomycin, and aminoglycosides attain acceptable concentrations in joint fluid after intravenous or intramuscular administration. Agents that are well absorbed from the gastrointestinal tract attain adequate joint fluid concentrations after oral administration. ^{, ,} Because antibiotics achieve high synovial fluid-to-serum ratios, there is no role for intra-articular instillation of antibiotics, which can produce chemical irritation and inflammation.

Parenterally administered therapy is used initially ( Table 76.3 ). Antistaphylococcal therapy should be given for a child of any age and the adolescent. Infants <3 months of age should be treated with antibiotics active against S. aureus, gram-negative enteric organisms, and GBS. Children 3 months to 5 years of age should receive empiric therapy for S. aureus, K. kingae, S. pneumoniae, and S. pyogenes. Although Hib infection is uncommon in immunized children, other serotypes of H. influenzae occasionally cause pyogenic arthritis. Children >5 years are treated for the most likely pathogens, S. aureus and streptococci. Empiric therapy for N. gonorrhoeae should be considered in sexually active adolescents.

Agents that cover MSSA such as parenterally administered nafcillin, clindamycin (depending on local antibiogram), or a first-generation cephalosporin are usual choices for empiric therapy. Empiric therapy with vancomycin or clindamycin is indicated when CA-MRSA isolates, in the region, exceed 10%–15%. Resistance to clindamycin may preclude its use as empiric therapy in some communities. Most S. pyogenes and S. pneumoniae isolates are susceptible to vancomycin and clindamycin, although susceptibility testing should be performed for both organisms. Neither drug is effective in treating infection caused by K. kingae. Most β-lactam antibiotics, including ampicillin, ampicillin-sulbactam, and second- and third-generation cephalosporins, have activity against K. kingae.

Antibiotic therapy should be modified according to results of culture and susceptibility testing. Nafcillin, oxacillin, or a first-generation cephalosporin remains the drug of choice if MSSA is isolated. Choices of antibiotic for MRSA infection in children is limited. Vancomycin is effective, but no absorbable oral formulation exists. If MRSA is susceptible to clindamycin (including absence of inducible resistance), clindamycin is an excellent choice and has excellent oral absorption. Linezolid and daptomycin have been used in some patients with serious MRSA infection, although data regarding daptomycin use in children are limited. Linezolid has excellent oral bioavailability, but thrombocytopenia, anemia, and leukopenia can occur after 2 or 3 weeks; and lactic acidosis has been reported. Longer-term use can be associated with peripheral and optic neuropathy. Linezolid is a weak, reversible monoamine oxidase inhibitor, and serotonin syndrome has occurred in children who also are receiving a serotonin-receptor inhibitor. The Infectious Diseases Society of America has published guidelines for the management of MRSA infections. Ceftaroline has excellent activity against MRSA, but its use in osteoarticular infections and in pediatric patients is limited.

Therapy based on culture results and susceptibility testing is continued parenterally until the child is afebrile; joint pain, swelling, and erythema have decreased; joint mobility has increased and markers of the acute phase response are decreasing. Because the CRP level normalizes more quickly than the ESR level (which typically is elevated for up to 2 weeks) when infection is controlled, CRP is used to monitor the early response to therapy. Open drainage of any joint (with lysis and irrigation of loculated collections) should be undertaken when aspirations yield samples that are persistently positive on culture.

Orally administered antibiotic therapy can be substituted for parenteral treatment after adequate control of infection and inflammation has been achieved if an oral antibiotic with appropriate coverage is available and if adherence and careful monitoring can be ensured ^, ( Table 76.4 ). Use of clinical practice guidelines has been associated with decrease in duration of parenteral antibiotics and hospitalization without increasing complications or sequelae.

For children in whom oral therapy is not feasible, outpatient parenteral antibiotic therapy administered through a tunneled central venous catheter or a peripherally inserted central catheter has been successful. Catheter-related mechanical and infectious complications can occur, and the risk versus benefit of prolonged central venous access should be considered.

Duration of antimicrobial therapy remains a subject of debate. Duration usually is determined by the specific pathogen, clinical and laboratory response, and whether adjacent osteomyelitis exists. Joint infections caused by S. aureus and gram-negative enteric organisms usually are treated for at least 3–4 weeks. Arthritis caused by H. influenzae, S. pneumoniae, S. pyogenes, and K. kingae is treated for 2–3 weeks, depending on the clinical response. Total therapy for <2 weeks for children in Finland with culture-positive arthritis was successful in most cases. None of the children in this study had MRSA infection, and several children with hip infections required longer courses of treatment. A larger, controlled, prospective study is necessary to better evaluate the adequacy of short-course antibiotic therapy.

Prognosis

Sequelae of pyogenic arthritis in children include abnormalities of bone growth, limitation of joint mobility, unstable articulation, and chronic dislocation of the joint. Joint dysfunction may not become apparent for months to years after infection. An estimated 10%–25% of children with pyogenic arthritis have residual dysfunction. ^,

Several risk factors for the development of sequelae have been identified, including age <6 months ; infection of the adjacent bone, which is evident in 10%–16% of children with pyogenic arthritis and increases the likelihood of sequelae to approximately 50% ^{, , , ,} ; infection of the hip or shoulder ^{, ,} ; a delay of 4 days or more before decompression and antibiotic therapy ^{, ,} ; and prolonged time to sterilization of synovial fluid. Staphylococcal and gram-negative bacillary infections carry a high risk of sequelae, whereas meningococcal and gonococcal infections carry a low risk.