Osteoarticular infections caused by Kingella kingae in children: contribution of polymerase chain reaction to the microbiologic diagnosis

Abstract

We report the cases of 18 children with osteoarticular infections caused by Kingella kingae. The bacteriologic diagnosis, mainly based on cultures of articular fluids, was significantly improved by directly inoculating the liquid medium used for hemocultures, yielding 65% of positive cultures vs. 13.5% with standard solid medium. In 5 cases infection of articular fluids was detected only by PCR of 16S ribosomal RNA. Kingella kingae is a Gram-negative bacterial pathogen that is responsible for osteoarticular infections among young children. K. kingae belongs to the Neisseriaceae family. It is a fastidious and a slow-growing bacterium that is rarely isolated from blood cultures. 1 The prevalence of respiratory carriage of K. kingae appears to peak in children <2 years of age, corresponding to the age of invasive infections. 2, 3 The emergence of the osteoarticular infections caused by K. kingae occurred coincidentally with a significant decrease in upper respiratory tract carriage and invasive disease by Haemophilus influenzae type b 2 as a result of implementation of H. influenzae type b vaccine. 1, 4 It is believed that damage to the respiratory mucosa allows hematogenous spread of this organism to joints and bones. Viral infections may facilitate the hematogenous invasion by this bacterium. For example K. kingae was isolated from the blood of 4 of 29 young children with culture-proved herpetic gingivostomatitis. 5 However, the pathogenesis and affinity of this microorganism for joints are not clearly understood. The increase in the number of reported cases of K. kingae infections can also partly be explained by the improvement of isolation methods and by a better recognition of this pathogen. Here we report the bacteriologic and clinical characteristics of K. kingae osteoarticular infections occurring in a French surgical orthopedic unit during a 4-year period. Patients and methods. We analyzed the clinical and bacteriologic records of all patients diagnosed with K. kingae osteoarticular infections in the Necker-Enfants Malades Hospital (Paris, France) during a 4-year period. The clinical data collected included medical history, physical examination, radiologic imaging, antibiotic treatment and clinical outcome. The biologic data collected on admission included white blood cell counts and C-reactive protein concentrations. Osteoarticular fluids and synovial biopsy samples were used to inoculate blood-agar and chocolate-PolyVitex medium (Aventis Pasteur, Marcy l’Etoile, France) in aerobic and anaerobic conditions during 5 days at 35°C. When the samples were sufficiently large, BacT/Alert blood culture bottles were simultaneously seeded and then incubated in aerobic and anaerobic conditions for 21 days at 35°C (Organon Teknika Corp., Durham, NC). Isolates of K. kingae were identified on the basis of colony morphology, alpha-hemolysis, positive oxidase and negative catalase reactions and the biochemical profiles obtained with the rapid NH system (Aventis Pasteur). The disk diffusion method on Mueller-Hinton blood agar medium was used to determine antibiotic susceptibility profiles. When the volume of osteoarticular fluids was sufficient, samples were stored at –20°C until DNA extraction and PCR amplification of 16S ribosomal RNA (rRNA). DNA was extracted from samples with the QIAmp DNA minikit (Qiagen, Courtaboeuf, France). The PCR mixture (final volume, 50 μl) contained 10 μl of DNA, 1 μM (of each) primer, 200 μM (of each) deoxynucleoside triphosphate and 1 unit of Taq gold DNA polymerase (PerkinElmer, Foster City, CA) in 1× amplification buffer [10 mM Tris-HCl (pH 8.3), 50 mM KCl, 3 mM MgCl2]. We used two pairs of universal primers, 27F (5′-AAG AGT TTG ATC CTG GCT CAG), 244 (5′-CCC ACT GCT GCC TCC CGT AG) to amplify a 357-bp fragment and PL06 (5′-GGT TAA GTC CCG CAA CGA GCG C), 1492R (5′- GGT TAC CTT GTT ACG ACT T) to amplify a 425-bp fragment. PCR was performed in a PerkinElmer 9600 thermocycler with an initial step of 7 min at 95°C, followed by 40 cycles of 20 s at 94°C, 30 s at 55°C, 45 s at 72°C and a final extension step of 10 min at 72°C. The amplicons were then purified by filtration through a purification column (Pharmacia Biotech, Uppsala, Sweden), and both strands were sequenced with an ABI-Prism 310 sequencer (PerkinElmer) using the Big Dye Terminator Cycle Sequencing kit (Perkin Elmer). The GenBank database was searched using the resulting sequences for species assignment. Results. From January 1999 to December 2002, we recovered K. kingae from 17 children with suppurative arthritis and one patient with osteomyelitis. The main features of the patients are listed in Table 1. The clinical presentation was characterized by moderate fever in most cases (mean temperature on admission, 37.6°C) and painful joints with inflammation and limited passive movement of the suppurative joint, with no evidence of a previous penetrating injury. One patient was diabetic and eight had previously had an upper respiratory tract infection. The joints most commonly infected were the knee and the hip. Ultrasound revealed joint effusions in 13 of 15 cases. Three patients had normal radiographs. CRP concentrations were mildly elevated in 15 of 18 patients (83%). The mean white blood cell count was 13 300/mm3 with a predominance of polymorphonuclear cells.TABLE 1: Summary of demographic and laboratory data for the 18 children with Kingella kingae osteoarticular infectionsResults of bacterial culture are listed in Table 2. K. kingae was isolated from 13 children, predominantly from articular fluid (66.5%vs. 6% from blood culture and 9% from synovial biopsy). For one patient all osteoarticular samples were culture-negative, and K. kingae grew from one blood culture collected on admission. K. kingae was mostly isolated on liquid medium used for hemoculture (65%vs. 73.5% on standard media, for articular fluids). In one case the liquid medium was not inoculated in addition to standard medium. On average bacterial growth in hemoculture bottles was detected after 3 days. All cultured isolates of K. kingae were susceptible in vitro to amoxicillin. For 5 of 10 patients for whom PCR 16S rRNA was performed on osteoarticular samples, K. kingae was detected only by this method, with the culture remaining negative.TABLE 2: Results of bacterial culture of samples collected from the 18 children with Kingella kingae osteoarticular infectionsThe surgical treatments included aspiration of the purulent fluid, drainage and cast immobilization for all children, associated with arthrotomy for 13 patients. In all patients standard antibiotic therapy for suspected osteoarticular infections was initiated with the intravenous injection of cefotaxime (100 mg/kg/24 h), oxacillin (100 mg/kg/24 h) and gentamicin (5 mg/kg/24 h) for 3 days, followed by the oral administration of antistaphylococcal treatment with both rifampin (20 mg/kg/24 h) and fusidic acid (40 mg/kg/24 h). Based on the results of bacterial cultures, the oral antistaphylococcal therapy was replaced by the association rifampin (20 mg/kg/24 h) and amoxicillin/clavulanic acid (100 mg/kg/24 h) or by rifampin (20 mg/kg/24 h) and amoxicillin (50 to 100 mg/kg/24h) in 10 patients. The mean duration of hospitalization was 6 days (range, 4 to 10 days), and the mean length of antibiotic therapy was 8 weeks. The 18 patients were completely cured and had no functional sequelae during a 6-month follow-up period. Discussion. We report a series of 18 children with osteoarticular infections caused by K. kingae. During the 4-year study period, K. kingae was the most frequently isolated pathogen from osteoarticular infections in children <3 years (32% of the causative pathogens isolated), before S. pneumoniae (18%) and S. aureus (16%). The high isolation rate of K. kingae is consistent with the increased prevalence of this bacterium in young children. 1, 4, 6, 7 Our clinical data confirm that the symptoms of K. kingae osteoarticular infections are mild and nonspecific and that ultrasonography is useful for identifying articular effusions. The main locations affected were the knees and hips, and no cases of underlying illness before the osteoarticular sepsis were identified. 8–12 Our results also suggest that previous episodes of mucosal respiratory inflammations are a risk factor for osteoarticular infection with K. kingae, as previously reported. 3, 5, 8, 10, 13, 14 We found that directly inoculating the liquid medium used for hemoculture with articular fluids strongly enhanced the yield of positive cultures, as mentioned in other studies. 3, 10, 15, 16 Thus it is essential to use this enrichment technique to isolate this bacterium, because it is probably present in low quantities in samples and might be inhibited by factors in joint fluid. 3 Enrichment was insufficient in some cases. In these cases the PCR technique based on the amplification of K. kingae 16S rRNA from the osteoarticular fluids was helpful in detecting the causative pathogen. In our series this PCR method detected K. kingae in all 10 of the articular fluid samples tested, even though 5 of these PCR-positive samples remained negative with the liquid medium culture method. This PCR method has been reported only once in K. kingae suppurative arthritis. 17 All strains of K. kingae isolated in our series were beta-lactamase-negative, as was the case in most other series. 8–10, 13, 18, 19 However, some clinical isolates are resistant to penicillin, as revealed by the nitrocefin disk test. 13, 20 The favorable evolution of all 18 patients probably reflects the low pathogenicity of this organism and its high susceptibility to the antibiotics used for therapy. 8, 10, 13