Infectious Arthritis

Clinical Presentation

Acute septic arthritis can occur at any age, but young children and elderly adults are most susceptible, especially if they have an already abnormal joint from previous trauma or from conditions such as hemophilia, osteoarthritis, or rheumatoid arthritis. Immune compromise for any reason and diseases such as cancer, diabetes, alcoholism, cirrhosis, and uremia increase the risk for infection. Usually, predisposing conditions are associated with particular types of causative organisms ( Table 22.1 ); a thorough history and physical examination should be done.

Septic arthritis occurs most frequently in adults; however, the most serious sequelae from infection occur in children, especially if a hip joint is involved and treatment has been delayed. Age-dependent anatomic variables may be responsible for the serious complications in children, such as destruction of the epiphysis and associated osteonecrosis from increased intracapsular pressure and septic effusion. Using immature avian models, Alderson et al. provided evidence that transepiphyseal vessels do exist and provide a direct connection between the physis and epiphyseal cartilage, supplying a route for bacteria to spread from an osteomyelitic focus in the metaphysis to the epiphysis and subsequently to the joint space.

The lower extremity weight-bearing joints are predominantly affected (61% to 79%); however, any joint can be involved, and multiple joint infections do occur. A thorough examination to determine if there is monoarticular or polyarticular infection is necessary before treatment is initiated. Inflammation of a single joint can be caused by numerous diseases ( Box 22.1 ). Joint sepsis should be an early consideration, however, because failure to diagnose this condition promptly may result in irreversible joint damage or death.

Acute septic arthritis can be difficult to diagnose in neonates because the inflammatory response is blunted and signs such as fever, swelling, erythema, and pain may be minimal or lacking. The only finding in a neonate may be infection at another site (e.g., the umbilical catheter), irritability, failure to thrive, asymmetry of limb position, or displeasure at being handled.

Clinical predictor algorithms have been used to differentiate transient synovitis from septic arthritis in children. A C-reactive protein greater than 20 mg/L and inability to bear weight yielded a 74% probability of septic arthritis, and patients with neither predictor had a less than 1% probability of septic arthritis ( Table 22.2 ).

Synovial Fluid Studies

The importance of early diagnosis and treatment of septic arthritis prior to destruction of an affected joint is well established. The organisms and/or the cytotoxins they may produce can irreversibly damage cartilage and subchondral bone within only a few days. An estimated 25% to 50% of patients with septic arthritis end up with irreversible loss of joint function. Therefore, early diagnosis is the most important factor for functional prognosis in patients with septic arthritis.

The current benchmark for identification of the causative microorganism and ascertaining its antimicrobial susceptibility is the conventional culture of aspirated joint fluid. The synovial fluid obtained should be sent for immediate Gram staining, culture, cell counts, and crystal analysis. Measuring erythrocyte sedimentation rates or C-reactive protein levels may be helpful in following the treatment course. A C-reactive protein of more than 10.5 mg/dL is predictive of infection. A leukocyte count greater than 2000/mm ³ points toward inflammation, whereas lower counts suggest a mechanical disorder. However, features classically taken to indicate septic arthritis include cloudy or turbid joint fluid, a nucleate cell count greater than 50,000/mm, or a high percentage of neutrophils; leukocyte counts of 28,000/mm or less also have been implicated, especially in immunocompromised patients. In addition to the total leukocyte count, the proportion of polymorphonuclear neutrophils (PMNs), if greater than 90%, indicates infection.

Crystal-induced arthritis is the main differential diagnosis in septic arthritis, as the clinical manifestations may be similar and the joint fluid findings comparable. Furthermore, infection developing concomitantly with crystal-induced arthritis may be overlooked when the joint fluid examination shows microcrystals. Combining standard joint fluid markers with the absence of microcrystals improves diagnosis of septic arthritis compared to standard cytologic features used alone.

Although synovial fluid culture is useful in the diagnosis of septic arthritis, both false negatives and false positives can occur. Cultures can be negative in up to 75% of patients with septic arthritis. The use of empirical antibiotics may obscure results. Studies are being conducted to develop tests for rapid and accurate diagnosis of septic arthritis. Molecular methods such as polymerase chain reaction (PCR) technique are used to identify bacteria by their genotypes. The advantage of automated multiplex PCR (MPCR) lies in the rapid identification of causative pathogens and their antibiotic sensitivities (5 hours compared with several days in standard tissue culture techniques). Results appear promising for the diagnosis of periprosthetic joint infection in sonication and synovial fluid as well as septic arthritis in synovial fluid of native joints.

Imaging Studies

Numerous imaging techniques are available to help detect joint infections, and although they can help confirm the suspicion of septic arthritis, they are not diagnostic. In the first few days of infection, radiographs usually are normal; however, they may be helpful in that they may show soft-tissue swelling, displacement of the fat pad, or joint space widening from localized edema. As the infection progresses, joint space narrowing from the destruction of cartilage may become evident. Radiographs may be used to monitor the response to treatment and to detect inadequately treated stages of the disease, such as generalized joint destruction, osteomyelitis, osteoarthritis, joint fusion, or bone loss.

Ultrasonography, in contrast to radiographs, can be used to detect even small collections of fluid deep in the joints. Non-echo-free effusions from clotted hemorrhagic collections are characteristic of a septic joint. Ultrasonography can be used to guide initial joint aspiration and drainage and to monitor the status of intraarticular compartments, joint capsules, bone surface, or adjacent soft tissues. It is noninvasive, inexpensive, and easy to use but is heavily operator-dependent. Ultrasound-guided percutaneous needle aspiration has also proved useful in obtaining biopsy and to decompress the infection; in cases of acute arthritis when joint aspiration is not possible or in the absence of synovial fluid, the biopsy can confirm the diagnosis by determining the causal pathogen or can improve it when there is perivascular PMN infiltration .

Computed tomography (CT), magnetic resonance imaging (MRI), and bone scans also may be obtained to diagnose septic arthritis; however, these tests are not always necessary. CT, which is more sensitive than radiography, has limited use in the early stages of infection. It can show soft-tissue swelling, joint effusion, and abscess formation and can be used to guide joint aspiration, monitor therapy, and help select operative approaches. MRI can detect infection, and the extent of infection, and is particularly useful in diagnosing infections that are difficult to access. MRI has a greater resolution than CT and shows better anatomic detail than bone scans, making it useful in differentiating between bone and soft-tissue infections and showing joint effusion. In addition, patients are not exposed to radiation. MRI is costly, has limited value in the presence of metal implants, and has a lower resolution than CT in calcified bone structures and cortices. Similar to other imaging techniques, MRI is nonspecific and cannot differentiate between infectious and noninfectious inflammatory arthropathies.

Radionuclide bone scans often can detect localized areas of inflammation. Although the technetium-99m ( ^99m Tc)-methylene diphosphonate scan shows increases in isotope accumulation in areas of osteoblasts and increased vascularity, it may be normal in the early stages of septic arthritis. Other radiopharmaceuticals, including gallium citrate and indium-111 ( ¹¹¹ In) chloride, are more specific and sensitive in the detection of active infection than ^99m Tc-methylene diphosphonate, but they do not show bone or joint detail well, and it is often difficult to distinguish between bone, joint, or soft-tissue inflammation. ¹¹¹ In-labeled leukocytes localize in the areas of acute infection, but although this scan is positive in approximately 60% of patients with septic arthritis, false-positive results may occur in patients with osteoarthritis.

Pathogenesis

Hematogenous infection of a joint begins with a systemic bacteremia that ultimately invades the synovial cartilaginous junction from the intravascular space and spreads throughout the synovium and synovial fluid. Why joints are affected and other vulnerable organs are not is unclear; however, collagen receptors found on Staphylococcus aureus (the most common nongonococcal infecting cause of hematogenous septic arthritis) may play a role. Also, the lack of a limiting basement membrane in the capillaries of synovia may allow intravascular bacteria to reach the extravascular space of synovial tissue through gaps between capillary endothelial cells. In addition, synovial fibroblasts inhibit phagocytosis of bacteria.

Soon after the synovium has been infected it becomes hyperemic and infiltrated with polymorphonuclear leukocytes that rapidly increase over the next several days. Histologically, the appearance changes from acute to chronic inflammation with an increase in mononuclear leukocytes and lymphocytes, which become the predominant inflammatory cells by 3 weeks.

Destruction of the articular cartilage, which results from degradation of ground substance, is apparent 4 to 6 days after infection. Depletion of ground substance, according to Perry, begins approximately 2 days after inoculation and is caused by activation of enzymes from the acute inflammatory response, production of toxins and enzymes by bacteria, and stimulation of T lymphocytes during the delayed immune response. Bacterial antigens deposited in the synovium and specific toxins, such as staphylococcal enterotoxin, produced by bacteria stimulate proliferation of T lymphocytes. As the T lymphocytes increase and degrade the ground substance, collagen is exposed to collagenases, altering the mechanical properties of the articular cartilage and increasing its susceptibility to wear. Complete destruction of articular cartilage occurs at approximately 4 weeks. Joint dislocation or subluxation and osteomyelitis also may occur.

Microbiology

Age is an important factor in determining the causative agent in bacterial infection. S. aureus is the leading cause in all ages followed by group A . streptococcus and Enterobacter . Until the development of a vaccine, Haemophilus influenzae was the main pathogen in infants and toddlers and is still recognized in the literature as such. S. aureus (including methicillin-resistant strains) is the most common pathogen of septic arthritis in hospitalized neonates. Intravenous catheters and hyperalimentation have been implicated in the transmission of this organism.

Kingella kingae , an organism difficult to recover by joint cultures on solid media, may be a more common cause of septic arthritis than previously recognized , especially in children between the ages of 6 and 36 months. Infections caused by K. kingae are milder, and fever may even be absent. This organism is susceptible to most penicillins and cephalosporins.

Neisseria gonorrhoeae causes approximately 75% of septic arthritis cases in healthy, sexually active young adults, although a septic joint develops in less than 3% of patients infected with N. gonorrhoeae. This infection has a slightly different presentation than other types of infectious arthritis. Often the infection is polyarticular and may be associated with a papular rash. Joint cultures often are negative, but cultures from the pharynx or urethra may be positive. PCR may help identify N. gonorrhoeae in culture-negative synovial fluid. Gonococcal arthritis generally has a favorable outcome if treated with appropriate antibiotics, and drainage usually is unnecessary.

There has been a noted increase in the number of community-acquired methicillin-resistant S. aureus infections in the pediatric population in several large communities. This includes osteomyelitis and septic arthritis. At our pediatric institution, community-acquired methicillin-resistant S. aureus infection accounted for 26% of all septic arthritis and acute hematogenous osteomyelitis admissions. The community-acquired methicillin-resistant S. aureus infections are more likely to have positive blood cultures after appropriate treatment has been initiated and to require multiple surgical procedures.

In older adults with nongonococcal disease, S. aureus infections cause about half of the cases of septic arthritis, and streptococci and gram-negative bacilli are responsible for the other half. Polyarticular sepsis caused by S. aureus is extremely serious in patients with rheumatoid arthritis, hemophilia, or immunosuppression, and mortality rates have been reported to be 56%. Hallmark findings in acute septic arthritis, such as pain with passive motion, swelling, and erythema, also may be difficult to interpret in patients with rheumatoid arthritis. Adults with systemic lupus erythematosus have an increased likelihood of Salmonella infection, and individuals with a history of intravenous drug use are predisposed to gram-negative infections, including those caused by Pseudomonas organisms.

Treatment

The principles in the management of acute septic arthritis include (1) adequate drainage of the joint and resection of infected tissue, (2) antibiotics to diminish the systemic effects of sepsis, and (3) resting the joint in a stable position. Prompt drainage and evaluation of purulent joint fluid is crucial for preservation of articular cartilage and for resolution of the infection.

If a joint is suspected of being infected, aspiration with a large-bore needle should be done before antibiotic therapy is initiated. Treatment in children should be aggressive whether or not a causative organism is identified.

Empirical antibiotic treatment is based on the patient’s age and risk factors ( Table 22.3 ). Empirical antibiotic therapy should be given until culture and sensitivity results are available, at which time definitive treatment is initiated ( Table 22.4 ). If no organism is isolated, empirical therapy should be continued. In general, the decision regarding duration of therapy is left up to the physician and depends on the type of infecting organism, the condition of the patient, and the response to therapy. Infections caused by H. influenzae type b, Neisseria, or Streptococcus generally respond rapidly to appropriate antibacterial management, and duration of therapy can be brief (<2 weeks). Infections caused by staphylococci and gram-negative bacilli respond more slowly, however, often requiring 4 to 6 weeks of treatment. A longer period of therapy is required if the hip or shoulder is involved, if the patient is immunocompromised, or if the response to treatment has been poor. In general, if laboratory findings do not improve after treatment, an infectious disease consultation is warranted.

There is strong evidence to support the beneficial effect of corticosteroids as adjunctive therapy with antibiotics in the treatment of septic arthritis. This is consistent with data that show improved outcomes associated with steroid use in severe sepsis, severe pneumonia, bacterial meningitis, and acute pyelonephritis.

We believe that if the diagnosis is made early and the involved joint is superficial, such as the elbow or ankle, aspiration should be performed and repeated if necessary. Appropriate antibiotics should be administered, and the joint should be splinted in a position of function. The patient should be observed for a decrease in pain, swelling, and temperature and for improved joint mobility. Infections caused by less virulent organisms usually respond promptly to treatment. If the response is not favorable and repeat aspiration does not show a decrease in the synovial leukocyte count within 24 to 48 hours, drainage is necessary. If purulent material is deeply situated in a joint, such as the shoulder or hip, open surgical drainage may be necessary. Arthroscopic drainage is a good alternative to open drainage in many instances. Clinical comparisons of aspiration with arthroscopy or through open procedures have shown comparable results. Stutz et al. reported a 91% cure rate in 78 joints with arthroscopic irrigation and debridement, with only 4% requiring open procedures. The efficacy of treatment, however, was dependent on the stage of the joint infection. Numerous other studies have reported lower infection rates, fewer repeat debridements, shorter hospital stays, and better functional results in patients who have had arthroscopic irrigation and debridement compared to open drainage of the knee, hip, shoulder, wrist, elbow, and ankle. In one study of 161 patients with septic arthritis of the knee, 71% in the open drainage group required repeat drainage compared with 50% in the arthroscopic group. Jiang et al. analyzed 7145 cases from a nationwide database of septic arthritis of the shoulder and found higher incidences of septicemia and urinary tract infections perioperatively in arthroscopically treated patients than in those treated with an open procedure, but they had significantly fewer incidences of osteomyelitis. These authors noted that the patients in the arthroscopic group had substantially more preexisting conditions than those in the open group, which was likely a contributing factor. As Abdel et al. pointed out in their study of 50 patients with native shoulder sepsis, most patients were elderly, with 57% being immunocompromised. One in three required additional intervention. Kang and Lee agreed that host factors, age, comorbidities, and causative organisms are all related to outcome of arthroscopic debridement in septic arthritis. A systematic review of septic arthritis of the hip showed that arthroscopic native hip irrigation was safe and effective in select patients (those without deformity, with no bacterial infection, or those who were immunocompromised). The most important factor for a good outcome remains early diagnosis and treatment and patient selection. Patients with advanced arthroscopic findings have been shown to have worse outcomes.

As the infection resolves, therapy to restore normal joint function is begun, including functional splinting initially to prevent deformity, isometric muscle strengthening, and active range-of-motion exercises. Patients being treated for infectious arthritis often have varying degrees of deformity, and treatment with traction, dynamic splints, serial casting, and passive exercises may be useful. In the residual stage, the infection has completely subsided but the joint or joints involved are left with deformity or limitation of motion, so treatment is directed at correction and functional restoration of the joint. The possibility of reactivating the infection should be considered, however, when any necessary procedure is undertaken at this stage.

Tarsal Joints

Primary septic arthritis of the tarsal joints is rare. An uncontrolled infection in the tarsal joints requires wide surgical drainage.

Ankle

Aspiration

Swelling around the ankle often makes fluctuation difficult to locate. To avoid injuring important structures, the needle is inserted 2.5 cm proximal and 1.3 cm anterior to the tip of the lateral malleolus. This is just lateral to the peroneus tertius tendon ( Fig. 22.1 ).

Knee

The incidence of bacterial infections in the large joints is 2 in 100,000 per year. The knee joint is the most frequently affected.

Aspiration

Because the knee is a superficial joint, it can be aspirated easily. The needle is inserted on the lateral side at the level of the superior pole of the patella. It is advanced through the lateral retinaculum and into the joint ( Fig. 22.2 ).

HIP

Acute septic arthritis of the hip is a more serious disease in children than in adults, and severe complications are much more common in children. In many cases, infection begins first in the metaphysis or epiphysis and is carried into the joint. As a result of the peculiar circulation of the femoral head, a septic hip places the femoral head at high risk for osteonecrosis. Epiphyseal separation also has been reported as a complication of septic arthritis of the hip in children. If a septic hip goes undiagnosed in an infant, a pathologic dislocation may occur. After an infected hip in an infant or child has been surgically drained, the hip should be supported in abduction to reduce the risk of pathologic dislocation. Bilateral septic arthritis is seen more often in the hip than in other joints and occasionally is associated with spinal infection. Independent risk factors for repeat surgical procedures in children include a C-reactive protein greater than 10 mg/L, an erythrocyte sedimentation rate of greater than 40 mm/hr, osteomyelitis, and methicillin-resistant S. aureus infection.