The knee is the joint most commonly involved with septic arthritis in children and adults.6,18 Early diagnosis and management of septic arthritis is crucial because the infectious process can result in irreversible cartilage damage if left untreated.13,14,17 Management includes systemic antibiotic administration and drainage of the joint by a surgical procedure or repeated aspirations.
However, the infectious process may persist despite treatment in 1.7% to 7.8% of patients.7,18 Persistence of infection may result from adjacent osteomyelitis undiag-nosed at the time of presentation and not treated surgically, leading to repeat seeding of the joint with microorganisms.5
Septic arthritis with adjacent osteomyelitis has been widely described in pediatric patients, but very few adult patients have been reported.1,9 Marsh et al described four patients with chronic osteomyelitis resulting in involvement of the adjacent joint. All patients had a history of previous trauma and/or surgery of the affected bone.9 Only one case report described the potential relationship of acute septic arthritis with adjacent hematogenous osteomyelitis in adult patients.1 Jackson et al5 reported seven of 16 children with combined septic arthritis and adjacent osteomyelitis responded slowly to the management of septic arthritis and had persistent pain, fever, or extremity swelling. However, it remains unclear whether adjacent osteomyelitis can also explain the poor response to treatment in adult patients with septic arthritis.
We asked whether adjacent osteomyelitis was present in patients with septic knee arthritis who did not respond favorably to initial management. We wondered whether comorbid conditions were risk factors for development of adjacent osteomyelitis and if so determine the responsible microorganisms.
MATERIALS AND METHODS
We retrospectively reviewed all patients presenting with septic arthritis of the knee treated at our institution from 1993 to 2004. We included patients who fulfilled all of the following criteria: adult patients (> 18 years), septic arthritis (diagnosed by positive joint fluid cultures or/and aspiration of pus from the knee joint), and persistence or recurrence of symptoms (knee pain, decreased range of motion [ROM], and effusion) despite treatment. Patients were excluded if they had previous surgery at the involved knee, distal femur, or proximal tibia, or if they had penetrating or soft tissue trauma to these areas. Using this strategy we identified 147 adult patients (151 knees) with septic arthritis, which was considered to be hematogenous due to the absence of trauma or surgery. The median duration of symptoms prior to diagnosis of septic arthritis was 5 days (range, 1-21 days).
Episodes of septic arthritis were treated surgically. Arthrotomy of the knee joint was performed using a midline skin incision and a medial parapatellar approach. The purulent material was evacuated and partial synovectomy of inflamed and hypertrophic synovial tissue was performed. Specimens were sent for aerobic, anaerobic, mycobacterial and fungal cultures. The joint was copiously irrigated with 10 liters of saline, adding polymyxin and bacitracin in the last liter of fluid. The knee joint was closed over a drain. Culture-specific antimicrobial therapy was administered for three weeks.
Twenty-nine of these patients (33 knees) demonstrated persistence or recurrence of symptoms despite surgical drainage and antibiotic therapy. Persistence was defined as the lack of resolution of knee symptoms by two weeks postoperatively (22 patients). Recurrence was defined as emergence of knee symptoms at a later time following a symptom-free interval (7 patients).
The prevalence of persistence or recurrence of septic knee arthritis was 22% (33 of 151 knees). There were 22 men (76%) and seven women (24%) patients with a mean age of 49 years (range, 27-72 years). The diagnosis of adjacent osteomyelitis was based on magnetic resonance imaging (MRI) (Fig 1) and/or on the presence of soft metaphyseal bone and positive bone cultures at the repeat débridement procedure (bone cultures were positive in 27 of 31 knees with adjacent osteomyelitis). None of the patients in the current study had MRI as part of the initial diagnostic evaluation. Patients with adjacent osteomyelitis were treated with a repeat débridement and irrigation procedure and 6 weeks of antibiotic therapy.
Proportions of patients with comorbid conditions in the groups of patients with and without adjacent osteomyelitis were compared with the chi-square test or the Fisher's exact test as appropriate. All tests were two tailed and significance was set at p < 0.05.
Adjacent osteomyelitis was present in 31 of 151 (21%) of knees with hematogenous septic arthritis and diagnosed in 31 of 33 (94%) knees (in 27 of 29 patients) that did not respond favorably to appropriate treatment and demonstrated persistence or recurrence of infection. Adjacent osteomyelitis involved the distal femur in 11 of 31 knees (35%), the proximal tibia in four (13%) knees, and both locations in 16 (52%) knees. The median time interval between surgery/antimicrobial therapy and diagnosis of adjacent osteomyelitis was 7 days (range, 3-86 days).
Patients with adjacent osteomyelitis had a higher (p = 0.002) rate of comorbid conditions compared to patients without adjacent osteomyelitis (Table 1). Twenty-three (85%) of the 27 patients with adjacent osteomyelitis were compromised patients by having one or more comorbid condition, compared to 64 (53%) of the 120 patients without osteomyelitis. The most common comorbid conditions in patients with adjacent osteomyelitis were diabetes mellitus in 10 patients (37%) and intravenous drug use in eight patients (30%). Seven (26%) patients reported chronic alcohol use, five (19%) patients had chronic viral hepatitis or other liver disease, two (7%) patients were receiving corticosteroids, and two (7%) patients had acquired immunodeficiency syndrome (AIDS). Diabetes mellitus was present in a higher (p = 0.02) proportion of patients with adjacent osteomyelitis compared to patients without (37% vs 17%).
Staphylococcus aureus was identified in 19 of 31 knees (61%), eight (26% of total cases) of which were methicillin-resistant. Overall, 25 infections were monomicrobial and two were polymicrobial. The infecting organism causing osteomyelitis was the same organism isolated during the initial diagnosis of septic arthritis. Four patients, who were on antimicrobial therapy at the time cultures were obtained, had no growth of any microorganisms and their diagnoses were based on MRI findings (Table 2).
Our data show septic knee arthritis may coexist with adjacent osteomyelitis of the distal femur or proximal tibia in adult patients with persistence of symptoms despite drainage of the knee and antibiotic therapy.
We note several limitations. This study did not determine the overall prevalence of adjacent osteomyelitis in patients with septic arthritis, as MRI was performed only for patients who did not respond to management. It is possible some patients had adjacent osteomyelitis which cleared with the initial treatment. However, our study includes all patients admitted to the musculoskeletal infection ward of a single institution and treated with a consistent protocol implemented by the senior author (MJP).
The development of hematogenous osteomyelitis is infrequent in adult patients, and the combination of septic arthritis with associated hematogenous osteomyelitis is rare in the postantibiotic era.8,16 The literature is limited to a case report by Atcheson and Ward,1 who described one adult patient with a septic knee associated with osteomyelitis of the distal femur. Similarly, septic arthritis has been rarely reported in patients with adjacent chronic osteomyelitis. Marsh et al9 described four patients with chronic osteomyelitis who developed an adjacent septic joint, but all patients had a history of previous trauma and/or surgery of the affected bone.
In pediatric patients by contrast, septic arthritis has been associated with hematogenous adjacent osteomyelitis. Perlman et al12 evaluated 66 pediatric patients diagnosed with hematogenous osteomyelitis and found 22 (33%) patients demonstrated septic involvement of the adjacent joint. Welkon et al17 observed concomitant osteomyelitis was present in 12 of 95 pediatric patients (13%) with septic arthritis, and Jackson et al5 reported coexisting bone involvement in 16 of 96 pediatric patients (16%) who presented with septic arthritis.
The pathogenetic mechanism for the simultaneous presence of septic arthritis and osteomyelitis remains unclear. In pediatric patients, the proposed mechanisms for the coexistence of septic arthritis with hematogenous adjacent osteomyelitis includes the presence of transphyseal vessels in children younger than 18 months, and the intraarticular location of the metaphysis in joints such as the hip, shoulder, and ankle.10,11,15 However, Perlman et al12 observed no difference in the rate of coexisting septic arthritis in patients younger versus older than 18 months, and in intraarticular versus extraarticular locations of the involved metaphysis. The authors hypothesized additional pathogenetic mechanisms may be present, such as subperiosteal spread of infection and hematogenous seeding of the joint from the site of osteomyelitis.12 Atcheson and Ward1 also proposed hematogenous seeding of the synovial lining from the focus of osteomyelitis as the explanation for the association of septic arthritis and osteomyelitis in their adult patient. The vascular connections between the metaphyseal, epiphyseal, and synovial vessels may lead to infection of the synovial lining, which progresses from the lining to the joint fluid.1,3,4
Adjacent osteomyelitis was present in almost all of our patients (94%) who did not respond favorably to treatment for septic knee arthritis. We presume the area of osteomyelitis, which was undiagnosed and untreated at the time of surgical arthrotomy and drainage, prevented a satisfactory clinical response and caused reinfection of the joint. The presence of comorbidities in most of our patients (85%) may have also resulted in a less efficient host response to the involved microorganism(s)2 and a predisposition to hematogenous osteomyelitis and reinfection.
Persistence or recurrence of clinical signs of infection after management of septic knee arthritis in adults should raise the suspicion of adjacent osteomyelitis, especially in patients with diabetes or other compromised patients. The distal femur and proximal tibia can be evaluated by MRI to determine the presence of osteomyelitis. Repeat dé-bridement procedures should include the involved bone. The adverse effect of the adjacent osteomyelitis to the clinical course of pediatric patients with septic arthritis was noted by Jackson et al,5 who reported seven of 16 children with combined septic arthritis and adjacent osteomyelitis responded slowly to the management of septic arthritis and had persistent pain, fever, or extremity swelling. Additional joint decompression and drainage were performed in four patients, but failed in three patients.3 The authors emphasized children with a slow clinical response to therapy should be considered at high risk for osteomyelitis.3 They also suggested additional imaging studies to diagnose adjacent osteomyelitis and, if present, management with bone débridement.5
Slow response to management of septic knee arthritis in adult patients should raise suspicion of adjacent osteomyelitis, especially in compromised patients. Further evaluation and management may be necessary.
1. Atcheson SG, Ward JR. Acute hematogenous osteomyelitis progressing to septic synovitis and eventual pyarthrosis: the vascular pathway. Arthritis Rheum
2. Cierny G III, Mader JT, Penninck H. A clinical staging system for adult osteomyelitis. Contemp Orthop
3. Crock HV. The Blood Supply of the Lower Limb Bones in Man. London: E &S Livingstone, Ltd; 1967.
4. Gardner E. Blood and nerve supply of joints. Stanford Med Bull
5. Jackson MA, Burry VF, Olson LC. Pyogenic arthritis associated with adjacent osteomyelitis: identification of the sequela-prone child. Pediatr Infect Dis J
6. Kaandorp CJ, Dinant HJ, van de Laar MA, Moens HJ, Prins AP, Dijkmans BA. Incidence and sources of native and prosthetic joint infection: a community based prospective survey. Ann Rheum Dis
7. Kaandorp CJ, Krijnen P, Moens HJ, Habbema JD, van Schaardenburg D. The outcome of bacterial arthritis: a prospective community-based study. Arthritis Rheum
8. Lew DP,Waldvogel FA. Osteomyelitis. Lancet
9. Marsh JL, Watson PA, Crouch CA. Septic arthritis caused by chronic osteomyelitis. Iowa Orthop J
10. Ogden JA. Changing patterns of proximal femoral vascularity. J Bone Joint Surg Am
11. Ogden JA. Pediatric osteomyelitis and septic arthritis: the pathology of neonatal disease. Yale J Biol Med
12. Perlman MH, Patzakis MJ, Kumar PJ, Holtom P. The incidence of joint involvement with adjacent osteomyelitis in pediatric patients. J Pediatr Orthop
13. Ross JJ. Septic arthritis. Infect Dis Clin North Am
. 2005;19: 799-817.
14. Shirtliff ME, Mader JT. Acute septic arthritis. Clin Microbiol Rev
15. Trueta J. The three types of acute haematogenous osteomyelitis. J Bone Joint Surg Br. 1959;41:671-680
16. Waldvogel FA, Medoff G, Swartz MN. Osteomyelitis: a review of clinical features, therapeutic considerations and unusual aspects. N Engl J Med
17. Welkon CJ, Long SS, Fisher MC, Alburger PD. Pyogenic arthritis in infants and children: a review of 95 cases. Pediatr Infect Dis
18. Weston VC, Jones AC, Bradbury N, Fawthrop F, Doherty M. Clinical features and outcome of septic arthritis in a single UK Health District 1982-1991. Ann Rheum Dis