SEPTIC ARTHRITIS IN A 14 MONTHS OLD CHILD

Clinical History

A 14-month-old girl presented to the emergency department with history of low-grade fevers (38.5oC). The child was the product of a full-term delivery and was meeting all milestones. However, the mother reported that the child had started refusing to walk several days before presentation and had been "abnormally fussy" secondary to teething and stomatitis. The right knee was slightly swollen and painful upon movement during physical examination. Fluid was aspirated from the right knee, which showed 30,000 leukocytes with a neutrophil predominance of 80%. C-reactive protein was elevated at 3.0 mg/dL (normal, 0.0-0.9 mg/dL). No organisms were seen on direct Gram stain. The child was started on empiric broad-spectrum antibiotic therapy with ceftriaxone and vancomycin.

Direct culture of the synovial fluid on routine media demonstrated no growth at 4 days. Additional knee fluid was sent for broth enriched culture in aerobic and anaerobic blood culture bottles. Both bottles demonstrated growth of a gram-negative rod after 37 hours' incubation. Subculture of the broth demonstrated very small colonies at 24 hours incubation on blood (Figure 1a) and chocolate agar, but no growth on MacConkey agar. Colonies were faintly beta-hemolytic (Figure 1c). The Gram stain from the solid medium is shown in Figure 2. The organism was oxidase-positive and catalase-negative.

Figure 1. Colonial growth morphology: (a) blood agar plate at 24 hours demonstrating small, clear colonies; (b) blood agar plate at 48 hours demonstrating more luxurious growth of the organism; (c) blood agar plate demonstrating soft beta hemolysis.

Figure 2. Gram stain of the joint fluid from the broth enrichment culture. The Gram-stained preparation demonstrates gram-negative rods in pairs and short chains. The rods measure 0.6-1 uM x 1-3 uM and demonstrate tapered ends. Kingella kingae may under-decolorize when staining and thus be mistaken for gram-positive rods.

What is the diagnosis and what should be included in the differential of septic arthritis in a child under 2 years of age?Differential Diagnosis

Staphylococcus aureus is the most common pathogen implicated in septic arthritis in children, representing approximately 40% of cases.[1,2] Included in the differential are other gram-positive cocci, such as group A and group C streptococci and Streptococcus pneumoniae. Prior to routine vaccination, Haemophilus influenzae type B was a significant pathogen, but its frequency in joint infections has decreased in recent years. Other gram-negative organisms commonly found in joint infections in children include Neisseria species (N meningitidis and N gonorrhoeae), Enterobacteriaceae species, Kingella kingae, and Brucella melitensis. Seen in much lower frequency are other etiologic agents, including yeasts. In a significant proportion of cases (over 33% in most studies), no etiologic agent can be identified by routine culture methods.[1,2] A recent article describing polymerase chain reaction (PCR) for K kingae demonstrated that K kingae DNA could be identified in almost half of the culture-negative joints, making this pathogen the most common organism identified in septic joints of children in that study.[3]

The Gram stain helped us to eliminate the gram-positive organisms, whereas the morphology of the gram-negative rods, growth rates, and growth media allowed us to eliminate Haemophilus influenzae (gram-negative coccobacillary forms, no growth on blood agar), Neisseria species (gram-negative diplococci), and Brucella melitensis (small, gram-negative coccobacilli) from the differential. The positive oxidase and lack of growth on MacConkey agar eliminated the Enterobacteriaceae from consideration. K kingae fit the Gram stain morphology, growth characteristics, and oxidase production. It is also the most likely etiologic agent in the patient's age group (under 2 years of age) when very sensitive methods like PCR are used to identify the infectious agent.Diagnosis

This is a case of joint infection caused by K kingae. K kingae is a fastidious gram-negative rod within the family Neisseriaceae. Its growth characteristics include slow growth on blood agar with the production of soft beta hemolysis. It fails to grow on MacConkey agar. The organism is oxidase-positive and urease-, indole-, and catalase-negative. It is capable of producing acids from glucose and maltose but no other sugars.[4] In this case, a definitive identification was made using the BD Phoenix Automated Microbiology System (BD Diagnostics; Sparks, Maryland).

Another key growth characteristic of this organism is its reluctance to grow in direct culture from septic joint fluids. Dr. Pablo Yagupsky demonstrated that inoculating joint fluid into aerobic blood culture bottles led to an increased incidence of K kingae in the joint fluids of young children.[1] In this study of 216 children with joint infections, synovial cultures were performed by routine culture and/or culture using a broth-enrichment blood culture system. The majority of children demonstrated infection with Staphylococcus aureus (36.5% of infections). The second most common isolate in this series was K kingae, representing 22.2% of isolates. Brucella melitensis, Streptococcus pyogenes, Streptococcus pneumoniae, group C streptococcus, Haemophilus influenzae, and rarely Neisseria meningitidis or Escherichia coli were also seen as joint pathogens. This study demonstrated that the Bactec broth enrichment blood culture system (Organon Teknika; Durham, North Carolina) was comparable to direct culture of the synovial fluid for all pathogens except K kingae. The majority of K kingae infections were detected by the broth-enrichment blood culture system only. Cultures with K kingae became positive in a median of 3.5 days, with the rare culture requiring extended incubation to produce clinical isolates. Following the study by Yagupsky and colleagues, many laboratorians have demonstrated increased detection of this organism when using blood culture broth-enrichment medium bottles from other manufacturers.[5-7]

It is hypothesized that inhibitors in the highly inflamed synovial fluid impede the direct growth of organisms on solid medium. These inhibitors are significantly diluted in blood culture bottles, potentially allowing for the growth of this fastidious organism.[8] Because of these findings, the Centers for Disease Control and Prevention (CDC) has recommended the routine use of an aerobic blood culture bottle for a broth enrichment culture to detect K kingae from synovial fluids.[9] More recently, PCR assays performed on synovial fluids cultured in aerobic blood culture bottles have been used to demonstrate K kingae DNA in the joint fluid of about half of the culture-negative patients in one series, suggesting that the prevalence of this organism in causing septic arthritis is dramatically underestimated.[3]
Disease Processes Associated With Kingella kingae

K kingae is most commonly associated with monoarticular septic arthritis syndrome in previously healthy children under 2 years of age. The joints most commonly infected are the large weight-bearing joints (hip, knee, and ankle), the knees being the most frequently involved. Of the non-weight-bearing joints, the wrist is most commonly involved.[4,10] Osteomyelitis may also be seen, most frequently involving the femur, calcaneus, and talus bones.[4] Infections of the vertebral disc, cellulitis, infections of bursae and tendons, and subcutaneous abscess formation have also been described in young children.[4]

K kingae is also associated with bacteremia and infective endocarditis.[4] It is one of the HACEK organisms (Haemophilus species, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, and Kingella species). The HACEK organisms are mouth flora that may hematogenously spread when there is a disruption of oral mucosal integrity, such as in cases of dental/periodontal disease or teething. These organisms were grouped together because of their fastidious nature and their association with "culture-negative" endocarditis. In the past, culture for these organisms required long incubation periods (2 or more weeks) and blind terminal culture onto solid medium. Because of the optimization of broth-enriched blood culture systems, the HACEK organisms are now often isolated during routine culture with a standard 5-day incubation period.[11] If extended incubation is requested by the ordering physician to rule out endocarditis, Clinical and Laboratory Standards Institute (CLSI) guidelines recommend subculture of all bottles onto chocolate and blood agar plates instead of extended incubation of the bottles themselves. Transient bloodstream infections without endocarditis are seen mostly in young children, whereas endocarditis is most commonly seen in school-aged children, adolescents, and adults.[4] Endocarditis with Kingella species generally occurs in the setting of preexisting heart and valvular abnormalities, including rheumatic heart disease, congenital heart defects, abnormal native valves, and prosthetic valves. Rarely it may occur in individuals whose valves were previously normal.[4]

Of note, bacteremia is only rarely detected in children with joint infections.[3,6,10,12] In this current case, blood cultures were sent, but no growth was detected at 5 days. In response to a request from the ordering physician, the cultures were held for 2 weeks. At 9.6 days, both bottles from one of the cultures demonstrated growth of Propionibacterium acnes.
Epidemiology

Similar to the other HACEK organisms, K kingae is part of the normal oral flora. It has been identified as a colonizer of the oropharynx in asymptomatic populations and in individuals who have developed invasive disease.[13-17] Child-to-child transmission, likely through oral secretions, has been suggested by studies looking at family units and daycare centers.[12,14,16,17] The highest level of colonization is seen in children under 3 years of age, the group in which most invasive disease occurs.[15,17] Although no gender difference exists for colonized children, there is a male predominance (67.8%) for invasive disease.[15] Predisposing features for developing active disease include preceding upper respiratory infections, hand-foot-and-mouth disease, stomatitis, aphthous ulcers of the mouth, tonsillar abscess, other oral lesions, otitis media, and diarrhea.[10,12,15,16,18,19] For infections seen in Israel, an apparent seasonality between July and December has been suggested.Treatment and Prognosis

K kingae isolates have been found to have relatively homogeneous antibiotic susceptibility patterns. Evaluation of 145 isolates from Israel demonstrated universal susceptibility to erythromycin, gentamicin, chloramphenicol, tetracycline, and ciprofloxacin. All of the isolates tested were negative for beta-lactamase production and had low minimum inhibitory concentrations (MICs) to penicillin. Nearly 40% of isolates, however, were resistant to clindamycin.[20] In a 2004 review of the literature, Yagupsky found published evidence of occasional resistance to erythromycin, ciprofloxacin, and co-trimoxazole and complete resistance to vancomycin and trimethoprim. Beta-lactamase had been found in 4 isolates.[4] In an outbreak of infections seen in an Israeli daycare center, all of the isolates were resistant to clindamycin and the MIC for penicillin ranged from 4 µg/mL to 94 µg/mL with a median value of 32 µg/mL.[17] Treatment generally includes 1 or more weeks of intravenous (IV) antibiotics and may be followed with several weeks of oral antibiotic therapy. Some patients' symptoms, however, will resolve without antibiotic therapy.[10]

The prognosis for joint infections with K kingae is generally good. Many of the cases are mild and may resolve without any interventions. Surgery is not indicated in most cases.Clinical Course

At the time of admission, the patient was treated with 625 mg ceftriaxone IV daily (75 mg/kg/day). Upon detection of K kingae in the joint culture, the child was taken to the operating room for an irrigation and debridement of the right knee joint, and an additional single intramuscular dose of 625 mg ceftriaxone was administered. Vancomycin, which was initially given empirically to cover for Staphylococcus aureus, was discontinued once the culture results were known. Knee fluid cultures collected intraoperatively had no growth at 5 days for the aerobic blood culture bottle and no growth at 4 days for culture on solid media. Repeat blood cultures were also negative. After 6 days of hospitalization, the patient was discharged home with a plan to continue IV ceftriaxone daily for a total of 2 weeks of therapy and consider conversion to oral therapy at the time of follow-up visit.