Osteomyelitis is an inflammation of the bone caused by infections involving the bone and/or bone marrow with bacteria or other organisms. In pediatric age groups, the incidence is 1:5000. Children <5 years old account for approximately 50% of the pediatric cases of osteomyelitis. In neonates, the incidence ranges from 1 to 3 cases for every 1000 hospital admissions.[1,2] Most of the reports have been case studies. Knudsen and Hoffman reported 34 cases of osteomyelitis in neonates and Hu and Chen reported 7 cases of neonatal osteomyelitis. Although uncommon, it might result in severe sequelae such as joint destruction and growth failure if it is not diagnosed and treated early. The diagnosis of osteomyelitis in neonates in early stages of the disease is challenging because of its nonspecific clinical presentation. As a result, the treatment is often delayed. Traditionally, the treatment of osteomyelitis in pediatric clinics is based on prolonged intravenous anti-infective therapy. Recently, more and more results from clinical trials suggest that a short course of intravenous antibiotics followed by oral therapy is safe and effective for children with uncomplicated osteomyelitis.[6–8] However, these findings do not apply to the neonates. In this paper, we describe 17 cases of neonatal osteomyelitis, including their presentation, laboratory tests and treatment, and discuss the diagnosis and management.
2 Materials and methods
The cases of 17 neonates diagnosed with osteomyelitis between January 2009 and September 2016 at the Children's Hospital, Zhejiang University, School of Medicine, were retrospectively reviewed. The diagnosis was confirmed by a positive blood culture, by the finding of pus at surgery or by radiographic changes such as metaphyseal rarefaction or periosteal reaction. Medical treatment included the use of antibiotics and supportive care. Surgery was performed in cases with subperiosteal abscess (>1 cm) or clinical deterioration despite antibiotic therapy, and ranged from percutaneous drainage to wound debridement and irrigation. Resolution of osteomyelitis was based on improvement in the clinical signs and inflammatory markers. The clinical records of all these patients, including age, sex, manifestations, bacterial culture, management, laboratory test results, and imaging examination results, were collected retrospectively. All neonates were followed up as out-patients. Our department follows up with patients every 6 months after discharge. All parents or guardians of pediatric patients provided written informed consent. The study was approved by the Ethics Committee of the Children's Hospital of Zhejiang University School of Medicine. The study was conducted in accordance with the Declaration of Helsinki for human subjects.
All statistical analyses were performed using SPSS 16.0 software (SPSS, Chicago, IL) for Windows. The data are expressed as mean (range). Categorical data were analyzed using the chi-squared (χ2) test. Differences between the surgical group and non-surgical group were evaluated by t test in the case of normally distributed variables or by the Mann–Whitney U test in the case of nonparametric variables. A P value <.05 was considered statistically significant.
3.1 Patients’ characteristics
The characteristics of the patients are shown in Table 1. All of the patients (11 men and 6 women) were term neonates. The median age was 16 days (range 1–28 days). The most common symptoms were redness or swelling around the affected bone (n = 17) and fever (n = 13). However, 1 patient was hospitalized with seizure. The most common infectious sites were femur (29.4%) and humerus (23.5%). Regarding the additional diagnosis, septic arthritis (n = 2), or cellulitis (n = 14) was also found, while none of the patient was diagnosed as myositis. For predisposing factors, there were 2 patients with chorioamnionitis, and 1 patient was delivered by vacuum extraction resulting in clavicle fracture. One case of skull osteomyelitis was caused by an infectious cephal hematoma. Most of the patients (n = 14) underwent x-ray examination. Furthermore, ultrasound (US) (n = 6), computed tomography (CT) (n = 5) or magnetic resonance imaging (MRI) (n = 7) were also performed. Typical changes included metaphyseal rarefaction, periosteal reaction, bone destruction, soft tissue enhancement, or abscess. However, 3 patients in our study presented with only slight swelling or excessive crying without fever, and there were no pathological findings on x-ray at disease onset. However, typical changes in x-ray or MRI were found after 1 to 2 weeks. The average time from onset to diagnosis was 12 days for the all patients (range: 2–24 days). Thus, diagnosis and treatment were delayed.
3.2 Laboratory examination and bacteriology
Routine blood and C-reactive protein (CRP) tests showed that most of the neonates had higher levels of white blood cells (WBC) (average: 20.53 × 109/L, range: 11.08–36.62 × 109/L), platelets (average: 506 × 109/L, range: 79–963 × 109/L), and CRP (average: 77.6 mg/L, range: 8–160 mg/L). Additionally, 7 cases (41.2%) had elevated liver enzymes, but only 1 case (5.9%) with thrombocytopenia was observed.
A blood or pus sample from all patients was sent for bacterial culture. There were 14 (82.35%) cases with positive cultures; Staphylococcus accounted for 71.43% (10 cases). The other bacteria were Salmonella (1 case), Streptococcus pneumoniae (1 case), Klebsiella pneumoniae (1 case), and Escherichia coli (1 case). Of the staphylococcal infections, 3 (30%) cases were resistant to oxacillin but sensitive to vancomycin (Table 1).
All patients were started with intravenous combined antibiotics after admission to the hospital. Antibiotic therapy was adjusted later according to the culture and the sensitivity results. Twelve cases were treated with vancomycin or oxacillin and antibiotics for Gram-negative bacteria. Third-generation cephalosporin or meropenem were used for the neonate infected with K pneumoniae, E coli, and Salmonella (Table 1).
There were 11 cases (64.71%) treated with surgical drainage. Compared with patients in the non-surgical treatment group, the WBC count of patients who underwent surgery was higher (23.39 ± 8.66 × 109/L vs 15.28 ± 3.95 × 109/L, P = .04). In addition, there was a tendency that patients undergoing surgery had shorter hospital stays than did patients without surgical treatment (21.73 ± 10.20 days vs 31.33 ± 7.05 days, P = .06). Other clinicopathological characteristics in the two cohorts showed no significant differences, including age, sex, platelets, CRP, hospital stays, predisposing factors, location, and culture (Table 2).
With antibiotic therapy alone or combined with surgical intervention, all patients recovered and were discharged. There were no obvious sequelae and no recurrence with a mean follow-up period of 49.47 ± 23.43 months.
3.4 Comparison of osteomyelitis findings, therapies, and outcomes in the reports by Knudsen and Hu and by our group
To further analyze the clinical features and treatment of osteomyelitis, we compared osteomyelitis findings, therapies, and outcomes in the reports by Knudsen and Hoffman and Hu and Chen and by our group (Table 3). We found that prematurity or chorioamnionitis was the main predisposing factors for osteomyelitis. The most common infection sites were the femur and humerus. S aureus was the most common bacterial agent. X-rays might be normal at disease onset, but the positive rate for detecting osteomyelitis by MRI was 100%. All patients received anti-infective treatment or surgical drainage. Eight patients had different degrees of sequelae in Knudsen paper, while there were none in the report by Hu and coworkers and the patients in our series.
Osteomyelitis is uncommon in neonates. The risk factors included chorioamnionitis, perinatal asphyxia, and prolonged rupture of membranes. Neonatal osteomyelitis can lead to permanent joint disability and disturbances in bone growth. To a certain degree, the outcome is dependent on rapid diagnosis and immediate initiation of treatment. Thus, early diagnosis of neonatal osteomyelitis is very important. Late diagnosis by only 4 days is a risk factor for long-term sequelae. However, the diagnosis of neonatal osteomyelitis is difficult and is often delayed because of nonspecific clinical presentation and normal plain radiographs in the early period. The manifestation of typical osteomyelitis includes redness or swelling at local sites or limited activity and fever. However, 3 patients in our study presented with only slight swelling or excessive crying without fever, and there were no pathological findings of x-ray at disease onset. Moreover, typical changes in x-ray or MRI were found after 1 to 2 weeks. The average time from onset to diagnosis was 12 days. Consequently, diagnosis and effective antibiotic treatment were delayed. Osteomyelitis should be considered in the differential diagnosis when neonatal sepsis is present, especially for neonates with local redness and swelling. Neonates whose x-rays are normal should be followed up closely.
Early changes on x-ray include periosteal reaction, lytic lesions, joint effusions, and destructive bone changes that would appear approximately 7 to 14 days after disease onset.[10,11] Similar to other reports, we found that the most common sites for neonatal osteomyelitis were the femur and humerus. There are also case reports of osteomyelitis of the mandible, iliac, rib, and vertebral bones in neonates.[12–15] Moreover, MRI is more sensitive for the early diagnosis of osteomyelitis, complicated abscesses, and related infection of soft tissues. Ultrasound evaluation is a useful diagnostic tool for identifying soft tissues swelling, periosteal elevation, and fluid collection adjacent to the affected bone.[16–18] Additionally, laboratory tests could provide more information to determine the condition of the patients. In our study, it was demonstrated that higher WBC counts were found in patients requiring surgery, indicating that we should be aware of the increased risk of surgery in neonates with high WBC count.
It has been reported that the most frequently isolated organism in neonates with osteomyelitis is S aureus, which is responsible for up to 70% to 90% of the cases.[8,16] Other microorganisms, including Streptococcus agalactiae, E coli, Klebsiella, and Candida albicans, are recognized as potential pathogens. In our study, we also found that the most common bacteria were in the Staphylococcus genus, including S aureus and methicillin-resistant Staphylococcus aureus (MRSA), accounting for 71.43% of the cases. In addition, Salmonella was isolated from 1 patient, which has rarely been reported in neonates. Depending on the neonatal osteomyelitis pathogen, if neonatal osteomyelitis is suspected, combined antibiotic therapy for Staphylococcus and Gram-negative bacteria is recommended, while oxacillin or vancomycin are recommended for use against methicillin-resistant S aureus (MRSA). Castellazzi et al suggested that empirical treatment for neonatal osteomyelitis should include oxacillin with the addition of gentamicin.
There is no doubt that acute osteomyelitis requires prompt anti-infective treatment, starting with intravenous antibiotics. Traditionally, children with acute osteomyelitis receive intravenous antibiotic therapy for 4 to 6 weeks. The duration and routes of administration of antibiotics is currently under debate. Some authors have suggested that a short course of intravenous therapy with a subsequent shift to oral administration in uncomplicated cases appears to be effective and safe.[6,7] However, there has been few randomized trials about neonatal osteomyelitis. In the randomized trial of Kaplan et al including 23 children aged 1 month to 15 years diagnosed with acute osteomyelitis, they addressed that clindamycin administered intravenous until the patient was afebrile for 3 consecutive days and then orally for approximately 4 weeks was an alternative to nafcillin/methicillin in the therapy of S aureus osteomyelitis in children. More studies are needed to confirm these results. In our study, most of the patients underwent intravenous antibiotic therapy for >3 weeks. Of course, the duration of antibiotics should be individualized to treat bacteria in different regions and should consider patient age, clinical status, and inflammatory markers.
In addition to anti-infective treatment, surgery plays a key role in the treatment of acute osteomyelitis. Biological samples that are useful for identifying the etiologic agent can be obtained at surgery. Chou and Mahadev found that early surgical intervention for acute bacterial osteomyelitis in children increased the diagnostic yield with cultures. Moreover, in cases with abscess or joint involvement, surgeons can remove the demineralized bone and clean the surrounding soft tissue; surgery can also improve the penetration of the antibiotics at the site of infection and reduce the bacterial load in surgical drainage, thereby reducing the risk of permanent damage and severe sequelae. Surgery is performed in up to 50% of osteomyelitis in children. In our study, 64.7% (11/17) of patients received surgical intervention combined with antibiotics. Hospital stays for these patients were shorter than those of neonates without surgery. We believe surgical intervention is effective for patients with joint involvement or abscesses.
Our study had some limitations that must be considered. First, the present study was retrospective in nature and a single-center experience. Second, due to the limited number of patients, external validation was not performed. Therefore, future studies should be performed to evaluate the benefits of surgery and different treatment modalities for neonatal osteomyelitis.
In conclusion, awareness of osteomyelitis and knowledge of the clinical and radiographic findings at presentation allow for early diagnosis. Combination of antibiotic therapy and adequate surgical intervention is suggested for neonatal osteomyelitis, especially for patients with a high WBC count.
Conceptualization: Canyang Zhan.
Data curation: Canyang Zhan.
Formal analysis: Bo Zhou, Jing Du, Lihua Chen, Canyang Zhan.
Funding acquisition: Canyang Zhan, Bo Zhou.
Investigation: Bo Zhou, Lihua Chen, Canyang Zhan, Jing Du.
Methodology: Bo Zhou, Lihua Chen.
Project administration: Bo Zhou.
Resources: Jing Du.
Software: Jing Du, Lihua Chen.
Supervision: Canyang Zhan, Jing Du, Lihua Chen.
Validation: Lihua Chen.
Visualization: Lihua Chen.
Writing – original draft: Jing Du, Bo Zhou, Lihua Chen.
Writing – review & editing: Canyang Zhan.
. De Boeck H. Osteomyelitis
and septic arthritis in children. Acta Orthop Belg 2005;71:505–15.
. Asmar BI. Osteomyelitis
in the neonate
. Infect Dis Clin North Am 1992;6:117–32.
. Knudsen CJ, Hoffman EB. Neonatal osteomyelitis
. J Bone Joint Surg Br 1990;72:846–51.
. Hu Y, Chen QX. Clinical analysis of 7 cases of neonatal hematogenous osteomyelitis
. Zhongguo Dang Dai Er Ke Za Zhi 2013;15:785–7.
. IIharreborde B. Sequelae
of pediatric osteoarticular infection. Orthop Traumatol Surg Res 2015;101:S129–37.
. Chiappini E, Mastrangelo G, Lazzeri S. A case of acute osteomyelitis
: an update on diagnosis and treatment. Int J Environ Res Public Health 2016;13:E539.
. Howard-Jones AR, Isaacs D. Systematic review of duration and choice of systemic antibiotic therapy for acute haematogenous bacterial osteomyelitis
in children. J Paediatr Child Health 2013;49:760–8.
. Castellazzi L, Mantero M, Esposito S. Update on the management of pediatric acute osteomyelitis
and septic arthritis. Int J Mol Sci 2016;17:E855.
. Liao SL, Lai SH, Lin TY, et al. Premature rupture of the membranes: a cause for neonatal osteomyelitis
? Am J Perinato 2005;22:63–6.
. Kothari NA, Pelchovitz DJ, Meyer JS. Imaging of musculoskeletal infections. Radiol Clin North Am 2001;39:653–71.
. Blickman JG, van Die CE, de Rooy JW. Current imaging concepts in pediatric osteomyelitis
. Eur Radiol 2004;14:L55–64.
. Martini S, Tumietto F, Sciutti R, et al. Methicillin-resistant Staphylococcus aureus mandibular osteomyelitis
in an extremely low birth weight preterm infant. Ital J Pediatr 2015;41:54.
. Bulbul A, Okan F, Yekeler E, et al. Acute osteomyelitis
of the iliac bone presenting with gluteal syndrome in a newborn. Eur J Pediatr 2009;168:1529–32.
. Ono S, Fujimoto H, Kawamoto Y. A rare full-term newborn case of rib osteomyelitis
with suspected preceding fracture. Am J Perinatol Rep 2016;6:e104–7.
. Felipe Villalobos A, Sanz Marcos N, Ventura Gómez N, et al. Neonatal vertebral osteomyelitis
with disc involvement. An Pediatr (Barc) 2010;73:368–9.
. Thomsen I, Creech CB. Advances in the diagnosis and management of pediatric osteomyelitis
. Curr Infect Dis Rep 2011;13:451–60.
. Pineda C, Espinosa R, Pena A. Radiographic imaging in osteomyelitis
: the role of plain radiography, computed tomography, ultrasonography, magnetic resonance imaging, and scintigraphy. Semin Plast Surg 2009;23:80–9.
. Malcius D, Jonkus M, Kuprionis G, et al. The accuracy of different imaging techniques in diagnosis of acute hematogenous osteomyelitis
. Medicina (Kaunas) 2009;45:624–31.
. Suranigi SM, Joshi M, Deniese PN, et al. Chronic osteomyelitis
of clavicle in a neonate
: report of morbid complication of adjoining MRSA abscess. Case Rep Pediatr 2016;2016:3032518.
. Harik NS, Smeltzer MS. Management of acute hematogenous osteomyelitis
in children. Expert Rev Anti Infect Ther 2010;8:175–81.
. Kaplan SL, Mason EO Jr, Feigin RD. Clindamycin versus nafcillin or methicillin in the treatment of Staphylococcus aureusosteomyelitis in children. South Med J 1982;75:138–42.
. Chou AC, Mahadev A. Acute bacterial osteomyelitis
in children. J Orthop Surg (Hong Kong) 2016;24:250–2.