MICROBIOLOGY AND RISK FACTORS OF LOW-GRADE INDOLENT INFECTION OF LONG BONE NONUNIONS AFTER OPEN FRACTURE

Aim Non-union following a long bone open fracture may be associated with an underlying low-grade indolent infection. A systematic analysis of bone/tissue samples obtained at the time of revision surgery is recommended, in association with a wide-spectrum antibiotic therapy. Our primary aim was to describe the microbiology of long bone non-unions after open fracture. Then, we sought to determine the clinical and biological factors associated with the risk of positive intraoperative tissue cultures and the outcome after non-union surgery. Method We retrospectively selected patients > 18 years, who had a primary planned surgical treatment of long bone non-union after diaphyseal open fracture, with no previous history of infection. From 2005 to 2021, 90 patients (69 males, 21 females) were included, with a mean age of 36.8 years (18-74 years) at the time of nonunion surgery. Non-union localizations were tibial (63 cases), femoral (21) or humeral (6). The mean time from the index surgical procedure to the time of non-union surgery was 21.3 months (5-170 months). Infection was confirmed when growth of the same microorganism was observed from at least 2 bone/tissue samples obtained at the time of non-union surgery. Postoperatively, all patients received a wide-spectrum antibiotic (AB) therapy from 10 to 14 days until microbiological analysis was completed. Then, the AB therapy was adapted accordingly or interrupted. Results Cultures were positive in 31 cases (34.4%), and revealed a mixed growth of microorganisms, with Staphylococcus aureus (n=13 cases) and coagulase-negative Staphylococcus species (n=12) being the most prevalent, followed by Cutibacterium acnes (n=9), and Pseudomonas aeruginosa (n=3). Factors associated with the risk of septic non-union were the stage of open fracture according to Cauchoix (p=0.03), a traumatic bone defect (p=0.01), and the presence of a preoperative local inflammation or a sinus drainage (p=0.04). Postoperatively, the risk of septic complication was higher (p=0.03) when cultures were positive at the time of non-union surgery. Conclusions In patients with presumed aseptic long bone non-union following an open fracture, an underlying low-grade indolent infection was found in a third of cases. The risk of infection correlated with the complexity of the index trauma. In these cases, the risk of septic failure was almost three times higher as compared to patients with negative cultures. According to our findings, we do not recommend a systematic AB therapy when non-union follows a type 1 open fracture, with limited bone defect, and in the absence of local inflammation.