A Clinical guide to pediatric infectious disease
Osteomyelitis and Septic Arthritis
Mechanisms of Infection
There are three basic mechanisms of infection: (a) direct inoculation, (b) hematogenous spread, and (c) contiguous spread from an adjacent area. The hallmark of pediatric infectious disease is infection by hematogenous spread.
Children are colonized with a variety of bacteria; a culture of the nasopharynx of an asymptomatic child could yield any number of bacteria, including Staphylococcus aureus and Streptococcus pneumoniae. Usually, these organisms reside on body surfaces with no ill effects. However, by a process not always well defined, these colonizing bacteria enter the bloodstream.
Once the bacteria enter the bloodstream, numerous things can occur (Fig. 2.1). Bacteremia can be transient and resolve without sequelae; this is often the case with viridans streptococci. Bacteremia, by its very presence in the systemic circulation, can cause overwhelming sepsis, as is often the case with Neisseria meningitidis. Bacteria can also be deposited in secondary sites, such as the cerebrospinal fluid or bone.
The bones are a frequent site of secondary infection because the blood supply takes a hairpin turn at the metaphyses of long bones, increasing the chance of the bacteria being deposited. This secondary seeding of bones from the blood is the major mechanism of pediatric osteomyelitis. This is in contrast to adults, who usually acquire osteomyelitis from direct inoculation following trauma or surgical procedures.
Pyogenic arthritis develops in a fashion similar to osteomyelitis, whereby blood-borne organisms are deposited in the synovium of the joint space. Similar to the long bones of children, the joint space is highly vascularized and is an area where bacteremic organisms are readily deposited. Bacterial arthritis can also spread from a contiguous osteomyelitis; blood vessels can deposit infection from the metaphysis
into the joint space. The organisms of septic arthritis are similar to those of osteomyelitis. S. aureus is the most common organisms, followed by S. pneumoniae, Kingella kingae, and group A streptococcus.
FIG. 2.1. Spread of bacteria from the bloodstream
The hallmark of pediatric osteomyelitis or septic arthritis is fever and localized pain. In the toddler whose verbal skills may not be sufficiently developed, the presentation may be simply fever and refusal to walk. It is for this reason that for many clinicians, fever and refusal to walk in a child indicates osteomyelitis or septic arthritis until proven otherwise. Septic arthritis of the hip is of particular concern to pediatricians because this joint space cannot be visualized directly on physical examination.
Clinical Clues to Diagnosis of Septic Arthritis of the Hip
The classic picture of a child with septic arthritis of the hip is a child holding the hip flexed and in external rotation. Often, the clinical picture can be subtle, with the major clinical clues being decreased leg motion and crying with diaper changes. There has been great interest in developing a clinical model for predicting that the child with fever, limp, and hip pain has either toxic synovitis or septic arthritis of the hip. One frequently quoted study indicated four major clinical variables: peripheral white blood count greater than 12,000/m3, sedimentation rate greater than 40 mm/h, fever, and non–weight bearing. When none of the clinical variables are present, the probability of pyogenic arthritis is less than 0.2%; the probability of septic arthritis of the hip increased to 93% and 99.6% when three and four clinical variables, respectively, are seen.
When facing the clinical condition of the febrile child who is not walking, it is necessary to pursue a logical clinical and laboratory evaluation. A complete physical examination is mandatory. Point tenderness should be sought in an attempt to localize potential infected areas. A careful examination should always include full range of motion of the hips.
Rapid Diagnosis of Septic Arthritis of the Hip
Rapid diagnosis of septic arthritis of the hip is particularly important because the tight confines of the hip joint in the setting of a purulent infection can rapidly compromise the arterial blood supply. If there is any question about the existence of septic arthritis of the hip, an ultrasound of the hip should be obtained. If fluid is seen by this study, one cannot easily determine whether this fluid is purulent or secondary to a transient viral infection (toxic synovitis).
1. Child with fever, refusal to walk
2. Physical examination: assess for point tenderness, hip pain, and decreased internal rotation of hip.
3. Widening of joint space suggests septic arthritis (Fig. 2.2). Plain films can rule out trauma and slipped epiphyses.
4. Laboratory evaluations: sedimentation rate, blood culture, complete blood count
5. Ultrasound of hips: rule out effusion
6. If any effusion is present, aspiration by orthopedic surgeon or interventional radiology is indicated.
All joints that appear swollen and erythematous should be aspirated with fluid sent for Gram stain, culture, and white blood cell count. If purulent fluid is present or an aspirate reveals greater than 100,000 white blood cells/m3 or positive Gram stain, the diagnosis of septic arthritis is made.
If point tenderness is elicited, plain films of the area should be done. Changes on plain films from osteomyelitis are often not apparent for at least 14 days; the real purpose of plain films is to rule out any other reason for the clinical presentation, such as an occult fracture or foreign body. Complete blood counts, sedimentation rate, and blood cultures are also useful. Although only 50% of patients with osteomyelitis have elevations in their white blood cell count, 90% of patients with osteomyelitis or septic arthritis have elevation in the sedimentation rate or C-reactive protein. Given the mechanism of the disease in pediatric osteomyelitis, blood cultures should be obtained and are positive in up to 30% of cases.
Several radiographic studies can be used in the further evaluation of pediatric osteomyelitis. Radionucleotide bone scan has been a traditional exam that shows increased uptake around infected bone (Fig. 2.3). There is increasing experience using magnetic resonance imaging (MRI) in visualizing infected bone and bone marrow for the diagnosis of osteomyelitis. In centers in which there is expertise in the use of MRI, it is often a front-line study. It should be noted that MRI is sensitive, but not specific. Although it is very helpful in documenting abnormalities in bone, bone marrow, or soft tissue, it is not specific in determining the etiology of these changes. Appearance of bone and bone marrow in infection, trauma, or even infarction following sickle cell crisis can appear similar. Interpretation of MRI findings should always be done with the clinical context in mind (Figs. 2.4 and 2.5).
FIG. 2.2. Plain film revealing widened joint space and osteopenia consistent with septic arthritis of hip.
FIG. 2.3. Bone scan showing increased uptake at left distal femur.
FIG. 2.4. Magnetic resonance image revealing large abscess in distal tibia.
FIG. 2.5. Magnetic resonance image revealing osteomyelitis of left femoral head.
All Patients with Bacterial Arthritis of the Hip Require Prompt Surgical Drainage and Irrigation
The presence of fluid as seen by ultrasound in the correct clinical setting necessitates immediate aspiration of that fluid. This can be done by either orthopedic surgery or interventional radiology under fluoroscopy. A joint aspirate that reveals greater than 100,000 white blood cells/m3 is strongly correlated with septic arthritis. If the fluid is thought to be consistent with septic arthritis, the child should proceed immediately to the operating room for drainage. Because Gram stain and cultures are positive in only a small fraction of hip aspirates (25% to 30%), the decision to proceed to surgical drainage is determined by the cell count profile of the aspirated fluid.
Antibiotic Therapy for Pediatric Osteomyelitis
After the diagnosis of pediatric osteomyelitis is made, empiric therapy is begun. The major organism for pediatric osteomyelitis is S. aureus. It is increasingly appreciated that a large proportion of community-acquired S. aureus is methicillin resistant. Once seen only as a nosocomial infection, some communities report that up to 70% of S. aureus
infections are resistant to methicillin. When confronted with S. aureus disease, the pediatrician can no longer assume that the organism will be sensitive to traditional antistaphylococcal medications such as nafcillin or first-generation cephalosporins. Community-acquired methicillin-resistant S. aureus (MRSA) is often sensitive to clindamycin and trimethoprim-sulfamethoxazole. Some clinicians are now empirically using clindamycin for initial treatment of community-acquired S. aureus disease; this usually covers both methicillin-susceptible and methicillin-resistant strains.
Although community-acquired MRSA is often initially susceptible to clindamycin, it has been noted that MRSA sensitive to clindamycin but resistant to erythromycin has the potential to develop clindamycin resistance. The specific test for the presence of inducible clindamycin resistance is the erythromycin induction (D) test. Although it is known that D-testing can detect inducible clindamycin resistance in a large percentage of MRSA isolates, it is not clear whether this in vitro test predicts clinical failure of clindamycin. There are scattered case reports of actual clinical failure in patients with a positive D test in whom clindamycin was used; it is advised that long-term clindamycin treatment for MRSA, such as that given for osteomyelitis, be approached with caution. Alternatives for long-term antibiotic therapy for MRSA infections include vancomycin and linezolid.
Before the development of the Haemophilus influenzae vaccine, this organism was also a frequent cause of pediatric osteomyelitis and septic arthritis. Despite the decline in H. influenzae disease, gram-negative organisms still play a role in pediatric osteomyelitis. K. kingae, a fastidious hemolytic gram-negative bacilli, has emerged in recent years as an invasive pathogen in children. Osteomyelitis and septic arthritis are the most common presentations of invasive K. kingae infections in children. Recent studies have suggested that about 20% of septic arthritis and osteomyelitis may be due to this organism. Some series have reported that K. kingae is the most common cause of septic arthritis in children younger than 2 years, being the causative agent in almost one half of cases. Thought to be a normal part of the oral flora in children, this pathogen gains access to the bloodstream in a manner similar to S. aureus. It is postulated that disruption of the respiratory or oral mucosa allows colonizing bacteria to enter the bloodstream. Preceding stomatitis is thought to play a role in the development of bacteremia and subsequent infection. Bone infection caused by K. kingae can be present in unusual locations, such as metatarsal bones and the epiphysis of long bones. Kingella is a fastidious aerobic pathogen that may not grow on standard agar; direct inoculation of an osteoarticular aspirate into blood culture bottles has been reported to improve the yield of cultures. Polymerase chain reaction amplification of synovial fluid has also been employed successfully in identifying the organism. K. kingae remains highly susceptible to many antibiotics, including third-generation cephalosporins.
Salmonella species are other gram-negative organisms that can cause osteomyelitis, particularly in patients with sickle cell anemia. It is for this reason that therapy with clindamycin, nafcillin, or a first-generation cephalosporin, combined with a third-generation cephalosporin (for optimal gram-negative coverage), is often used as empiric treatment of osteomyelitis until culture results are available.
Duration of Therapy
Early studies pointed to a higher relapse rate in patients treated for 3 weeks or less. Chronic infection has also been reported to develop more frequently in patients receiving only 3 weeks of therapy as compared with patients receiving therapy for 4 weeks or longer. Many clinicians believe that the minimum duration of treatment is 4 weeks and often continue treatment for as long as 6 weeks.
The monitoring of therapy using the sedimentation rate and C-reactive protein has been advocated. In children with septic arthritis, it is thought that the serum C-reactive protein peaks within 48 hours after treatment and normalizes in about 1 week. In contrast, the sedimentation rate may continue to increase despite effective treatment until day 5 and may remain elevated for more than 1 month. It has been recommended that the C-reactive protein be measured about 2 days after treatment is begun. Normalization suggests effective therapy.
Treatment is often continued until the sedimentation rate returns to normal values; this usually coincides with about a 1-month duration of therapy. An increasing C-reactive protein level or persistently elevated sedimentation rate can herald the need for surgical drainage. If, after 1 month of therapy, the repeated sedimentation rate is greater than 30 mm/h, a repeat MRI can be obtained to determine the need for surgery. Antibiotics can then continue for an additional 3 weeks, with repeat MRI and sedimentation rate done at that time.
Intravenous versus Oral Antibiotics
Traditionally, it was thought that serious bacterial infections such as osteomyelitis required intravenous antibiotics. In the early 1980s, studies examined the efficacy of oral treatment of pediatric osteomyelitis. These studies used the serum bactericidal titer (SBT) or Schlichter's test. This test is a modification of the MIC test. Patients with proven S. aureus osteomyelitis are given high-dose oral therapy, often 100 mg/kg per day of oral cephalexin. Serial dilutions are made from peak and trough serum samples. To these serial dilutions, an aliquot of the patient's infecting organism is added; the dilutions remaining clear after 24 hours of incubation is then plated on agar plates. The dilution at which bacteria fail to grow on agar plates is the peak and trough bactericidal titer (Fig. 2.6).
Prospective studies have determined that children with acute hematogenous osteomyelitis with a peak SBT of greater than or equal to 1:16 and a trough of greater than or equal to 1:2 achieve bacteriologic and clinical cure. Following these studies, many physicians advocate the use of oral antibiotics if the following conditions are met: (a) an organism is isolated, (b) adequate peak and trough SBTs can be obtained on oral therapy, and (c) compliance with oral therapy can be assured.
During the past 5 years, numerous reports have questioned the necessity of SBTs during oral therapy for osteomyelitis. Small series have reported good outcome in patients treated with oral antibiotics without confirmatory SBT measurement. To this day, there is great variation in clinical practice; some institutions will not use oral therapy unless a pathogen is isolated and adequate SBTs can be documented.
Some clinicians use oral therapy only if they are sure of the infecting organism, but do not think that SBTs are necessary. It is likely that the pediatrician caring for a child with osteomyelitis will have to make a case-by-case decision regarding what is the best mode of treatment for that particular child.
FIG. 2.6. Serum bactericidal titer (Schlichter's test) evaluating oral antibiotic treatment of Staphylococcus aureus osteomyelitis.
Antibiotic Therapy for Pediatric Septic Arthritis
Septic arthritis in joints other than the hip may not require surgical intervention, although repeated aspirations are sometimes used if fluid reaccumulates. Failure of septic arthritis to respond to appropriate antibiotic therapy should always lead to the consideration for surgical drainage. As in osteomyelitis, antibiotic therapy is traditionally given for at least 4 weeks. As in osteomyelitis, there is increasing experience in switching patients to oral antibiotic therapy after their physical exam has normalized. Many centers use an SBT with a peak titer of 1:16 as the level correlated with bacteriologic eradication and clinical cure.
Although the increased duration of effective therapy has reduced the risk for chronic osteomyelitis, chronic infection can occur and can be difficult to treat. Organisms
can survive in bone, and the accompanying bone necrosis may limit antibiotic penetration into infected areas. Although acute hematogenous osteomyelitis is typically treated with medical therapy alone, chronic disease often requires both medical and surgical treatment. Aggressive surgical techniques, which involve complete débridement of necrotic bone as well as the creation of muscle flaps for revascularization, are often required.
There remains no consensus on the duration of antibiotic therapy for chronic osteomyelitis. If at all possible, medical therapy should be dictated by the results of culture obtained at the time of surgery. Treatment courses of 3 to 6 months are reported to be generally effective.
Management of Pediatric Osteomyelitis and Septic Arthritis
1. Staphylococcus aureus (including methicillin-resistant strains)
2. Salmonella species (found in patients with sickle cell anemia)
3. Kingella kingae
2. Initial treatment
1. Clindamycin, 40 mg/kg per day in three divided doses, combined with Cefotaxime, 100 mg/kg per day; or ceftriaxone, 50 to 75 mg/kg per day
2. For methicillin-sensitive S. aureus, consider switching to high-dose cephalexin (100 mg/kg per day) once the patient is afebrile and clinically improved. Demonstration of peak SBT of more than 1:16 correlates with clinical and microbiologic cure.
Kallio MJ, Unkila-Kallio L, Aalto K, et al. Serum C-reactive protein, erythrocyte sedimentation rate and white blood cell count in septic arthritis of children. Pediatr Infect Dis J 1997;(16)4:411–413.
Karwowska A, Davies HD, Jadarji T. Epidemiology and outcome of osteomyelitis in the era of sequential intravenous: oral therapy. Pediatr Infect Dis J 1998;(17)11:1021–1026.
Kocher MS, Zurakowski D, Kasser JR. Differentiating between septic arthritis and transient synovitis of the hip in children: an evidence-based clinical prediction algorithm. J Bone Joint Surg Am 1999;81(12):1662–1670.
Martinez-Aguilar G, Hammerman W, Mason E, et al. Clindamycin treatment of invasive infections caused by community acquired, methicillin-resistant and methicillin-susceptible Staphylococcus aureus in children. Pediatr Infect Dis J 2003;22(7):593–599.
Moumile K, Merckx J, Glorion C, et al. Osteoarticular infections caused by Kingella kingae in children: contribution of polymerase chain reaction to the microbiologic diagnosis. Pediatr Infect Dis J 2003;22(9):837–839.
Newton PO, Ballock RT, Bradley JS. Oral antibiotic therapy of bacterial arthritis. Pediatr Infect Dis J 1999;(18)12:1102–1103.
Prober CG, Yeager AS. Use of the serum bactericidal titer to assess the adequacy of oral antibiotic therapy in the treatment of acute hematogenous osteomyelitis. J Pediatr 1979;95(1):131–135.
Yagupsky P, Dagan R. Kingella kingae: an emerging cause of invasive infections in children. Clin Infect Dis 1997;24(5):860–866.