Low-grade infections in nonunion of the femur and tibia without clinical suspicion of infection - Incidence, microbiology, treatment, and outcome.

Treatment algorithms for fracture nonunion depend on the presence or absence of bacterial infection. However, it is often impossible to identify infection preoperatively. While some infections may present with clinical signs of infection, low-grade infections lack infection signs and have a clinical presentation similar to aseptic nonunion. The clinical relevance of low-grade infection in nonunion is not entirely clear. Therefore, the aim of this study was to evaluate the role of low-grade infection in the development and management of lower extremity nonunion. A prospective multicenter clinical study enrolled patients with femoral or tibial shaft nonunion and regular healed fractures, scheduled for nonunion revision and routine implant removal, respectively. Preoperatively, serum markers including C-reactive protein (CRP), leukocytes, and procalcitonin were determined, clinical infection signs were recorded, and a suspected septic or aseptic diagnosis was made prior to surgery and further diagnostics. Tissue samples were collected for microbiology and histopathology, and osteosynthesis material for sonication. Nonunion patients were followed for twelve months, during which the definitive diagnosis of "septic" or "aseptic" nonunion was made according to diagnostic criteria for fracture-related infection. One hundred and ten patients with nonunion and 34 patients with regular healed fractures were included. Sixty-two nonunion patients were diagnosed as aseptic, 22 with expected and confirmed infection, and 23 with unexpected low-grade infection. Three patients had an unclear diagnosis. Low-grade infection was detected in 28 % of presumed aseptic nonunion patients. Sensitivity and specificity for the suspected diagnosis were 49 % and 95 %, respectively. The suspected diagnosis had a significant impact on revision strategy. All medians of the preoperative blood values were within the reference ranges except for CRP, which was slightly elevated in the expected and confirmed infected nonunion group. Expected and confirmed septic nonunion and unexpected low-grade infected nonunion demonstrated a similar bacterial spectrum. While 10 % of patients with aseptic nonunion required follow-up surgeries, re-operation rates were higher in patients with low-grade infection and expected and confirmed infection at 30 % and 64 %, respectively. Patients with low-grade infections were treated less frequently with systemic antibiotics and for a shorter duration than patients with expected and confirmed infections, with no significant difference in healing rate which was 83 % in low-grade and 62 % in expected and confirmed infections. The healing rate of aseptic nonunion was 90 %. A limitation of this study is the limited number of tissue samples for microbiological and histopathological diagnostics in the suspected aseptic nonunion cohort, which may have led to an underestimation of the low-grade infection rate. Our findings suggest that unexpected low-grade infection is frequently associated with nonunion. While expected and confirmed infected nonunion differs significantly from aseptic nonunion, low-grade infected nonunion is very similar to aseptic nonunion, except for intraoperative bacterial detection. In addition to antibiotic therapy, surgical nonunion revision with implant exchange and debridement appears to be highly effective in achieving consolidation of low-grade infected nonunion.