Salmonella osteomyelitis is a rare entity, constituting 0.8% of all Salmonella infections and only 0.45% of all types of osteomyelitis. There is a striking association between Salmonella osteomyelitis and children with sickle cell disease. Other risk factors include history of injury, malignancy, the use of immunosuppressants, diabetes, autoimmune diseases, chronic granulomatous disease, immunodeficiencies and HIV infection. However, in otherwise healthy children, it has very rarely been reported.
A review of the literature demonstrates a total of 13 cases of Salmonella osteomyelitis in previously healthy children.1–8 We report a series of 4 children without any predisposing factors, who were admitted to our hospital during a 10-year period (Table 1). This series represents, as far as to our knowledge, the largest single-hospital experience of Salmonella osteomyelitis in previously healthy children.
The first patient (patient 1) in our case series was a 13- year-old boy, who presented with a 10-day history of limping and pain in the right knee joint with a subsequent 3-day fever. Physical examination revealed tenderness over the upper right tibia, slight swelling and heat and normal joint mobility. The patient had an unremarkable history for trauma. The roentgenogram demonstrated an osteolytic lesion at the proximal metaphysis of the right tibia, which was confirmed with a magnetic resonance imaging scan. The presumptive diagnosis of osteomyelitis was made and surgical drainage was undertaken in addition to drilling of the metaphyseal cortex. Aspirates were cultured simultaneously onto selective agar media and also inoculated in blood culture bottles. Although traditional cultures remained negative, the cultures inoculated in blood culture bottles yielded Salmonella enteritidis resistant to ampicillin but susceptible to third-generation cephalosporins, quinolones and trimethoprim-sulfamethoxazole. Blood cultures were sterile and stool cultures were negative. Initial empiric treatment was switched to intravenous (IV) cefotaxime for 7 days, followed by oral trimethoprim-sulfamethoxazole for a total treatment period of 6 weeks.
The rest of our case series included a 10-month-old girl (patient 2), a 15-month-old boy (patient 3) and a 13-year-old boy (patient 4). All of them experienced a prolonged period of febrile illness and decreased limb mobility. One patient had a history of diarrheal disease before fever onset and 1 reported close contact with a pet turtle. As with patient 1, none of them had a previous history of trauma or surgery. Patient 2 presented with tenderness, localized swelling and erythema over the lateral epicondyle of the humerus, whereas patient 3 presented with decreased mobility of the left hip joint and patient 4 with signs of inflammation over the upper left tibia with concurrent motion restriction of the adjacent knee joint. A roentgenogram was performed in every case, revealing osteolytic lesions over the affected bones, except for patient 3, where soft tissue swelling and joint capsular distension, with widening of the joint space was demonstrated. Magnetic resonance imaging scans were performed in all cases confirming the diagnosis of osteomyelitis. Surgical drainage was undertaken in all patients and aspirates were subsequently cultured in both traditional media and blood culture bottles as per case 1. Interestingly, only aspirates inoculated in the blood culture bottles yielded S. enteritidis with regard to patients 2 and 3 and Salmonella corvallis with regard to patient 4. All strains were susceptible to ampicillin, cefotaxime, ciprofloxacin and trimethoprim-sulfamethoxazole and the initial empiric antimicrobial regimen was subsequently switched to third-generation cephalosporins for 2 to 3 weeks. Only patient 3 had Salmonella bacteremia and none had positive stool cultures.
An investigation for possible underlying conditions was performed in all 4 patients, including hemoglobin electrophoresis, sickle cell test, antinuclear antibodies, CH50, C3, C4, immunoglobulin levels, flow cytometry assay using dihydrorhodamine 123 and antibodies against HIV. All tests were negative or within age-specific range.
Although patient 2 had a satisfactory response to therapy, with a smooth transition to oral-administered antibiotics (ampicillin for 4 weeks), patient 3 experienced a clinical relapse with radiographic evidence of deterioration. A repeat surgical drainage was performed and he received ciprofloxacin IV for an additional period of 6 weeks. Follow-up at 1 year showed complete resolution of the lesions and full recovery of limp mobility in all patients.
Osteomyelitis is a known, but relatively uncommon, manifestation of Salmonella infection. In 1876, Paget was the first to report Salmonella osteomyelitis as a complication of typhoid fever. The 4 most common strains of Salmonella causing osteomyelitis in adults are Salmonella typhimurium, S. enteritidis, Salmonella enterica subsp. arizonae and Salmonella typhi.
Salmonella osteomyelitis has no distinguishing features, either clinically or radiologically. Most commonly affected are the long bones, especially the proximal portion of the humerus and the distal portion of the femur, but any bone structure can be involved, with the majority being single-bone involvement. The involvement of vertebral bodies is less frequent, occasionally complicated by a mediastinal or paravertebral abscess. In previously healthy children, Salmonella osteomyelitis is extremely rare. A review of the literature demonstrates a total of 13 cases of Salmonella osteomyelitis in children without predisposing factors.1–8 All but one, caused by nontyphi forms. With regard to our series of patients, S. enteritidis was the dominating serotype because it was isolated in 3 of 4 children.
Locations in the previously reported 13 cases,1–8 included 4 patients with humeral (30.7%), 3 with vertebral (23%), 3 with pelvic (23%), 2 with multiple site (15.3%) and 1 with tibia (7.7%) involvement. Age distribution varied from 3.5 months to 16 years. In almost half, treatment was with antimicrobials only (both IV and oral), whereas the remainder was treated with a combination of surgical irrigation with debridement and antimicrobial treatment. Duration of treatment ranged from 5 weeks to 6 months. It appears that prolonged duration of treatment is associated with a more favorable outcome. Indeed, the relapse of patient 3 in our series, shortly after discontinuation of treatment, can be attributed to its short duration. It is speculated by some authors that a 6-week conservative treatment with IV and oral antimicrobials leads to complete recovery of acute Salmonella osteomyelitis, with decreased danger of recurrence.5 However, there is currently no consensus regarding the treatment of acute or chronic osteomyelitis caused by Salmonella. Most experts treat chronic Salmonella osteomyelitis conventionally by surgical debridement combined with antimicrobials. Carlson and Dobozi9 consider that surgical debridement is not adequate and advocate radical debridement of the lesion.
Despite what has previously been reported in adults, where recurrence rate after an acute episode of Salmonella osteomyelitis is high, this does not seem to be the case in children. All previous healthy children reported in literature, including our series, experienced complete recovery except a female patient with multifocal infection and concurrent bacterial meningitis.1 Although blood cultures have been reported to be frequently positive (71%) in adults with Salmonella osteomyelitis, this does not seem to be the case with children. In only 1 of 4 patients from our series, was Salmonella isolated from blood cultures.
Because recently the Centers for Disease Control and Prevention reported outbreaks of turtle-associated salmonellosis in 18 US states, it is interesting to note that in 2 patients (1 of the previously reported cases and 1 of our series), a previous close contact with a reptile (iguana and turtle accordingly) has been documented.4
Since treatment duration in Salmonella osteomyelitis is expected to be prolonged, the importance of targeted therapy is fundamental. In all of our cases, several pus cultures were taken during surgery and subsequently cultivated with traditional methods and inoculated in blood culture bottles. However, in all patients, the culprit microorganism was isolated only when samples were cultured in blood culture bottles, whereas aspirates cultured directly onto selective agar media remained negative. It appears that the practice of direct inoculation of aspirates into blood culture bottles, suggested by literature in suspected Kingella kingae etiology, could also be useful in the diagnosis of Salmonella osteomyelitis.10 Although Salmonella osteomyelitis in previously healthy children is uncommon, clinicians should suspect this etiology in cases of empiric antimicrobial treatment failure.
1. Adeyokunnu AA, Hendrickse RG.. Salmonella
osteomyelitis in childhood. A report of 63 cases seen in Nigerian children of whom 57 had sickle cell anaemia. Arch Dis Child. 1980;55:175–184
2. Ebrahim GJ, Grech P.. Salmonella
osteomyelitis in infants. J Bone Joint Surg Br. 1966;48:350–353
3. Ingram R, Redding P.. Salmonella
virchow osteomyelitis. A case report. J Bone Joint Surg Br. 1988;70:440–442
4. Nowinski R, Albert R.. Salmonella
osteomyelitis secondary to Iguana exposure. Clin Orthop. 2000;372:250–253
5. Sucato DJ, Gillespie R.. Salmonella
pelvic osteomyelitis in normal children: report of two cases and a review of the literature. J Pediatr Orthop. 1997;17:463–466
6. Bettin D, Schaphorn G, Blasius S, et al. A rare case of Salmonella
osteomyelitis in the humerus as a differential diagnosis to a malignant bone tumor. Arch Orthop Trauma Surg. 2002;122:544–546
7. Canessa C, Trapani S, Campanacci D, Chiappini E, Maglione M, Resti M.. Salmonella
pelvic osteomyelitis in an immunocompetent
child. BMJ Case Reports. 2011;10
8. Miller ME, Fogel GR, Dunham WK.. Salmonella spondylitis. A review and report of two immunologically normal patients. J Bone Joint Surg Am. 1988;70:463–466
9. Carlson D, Dobozi W.. Hematogenous Salmonella typhi
osteomyelitis of the radius: a case report. Clin Orthop. 1994;308:187–191
10. Petti CA, Bhally HS, Weinstein MP, et al. Utility of extended blood culture incubation for isolation of Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella organisms: a retrospective multicenter evaluation. J Clin Microbiol. 2006;44:257–259