CLINICAL OUTCOME AFTER SURGERY FOR FRACTURE-RELATED INFECTION IS DEPENDENT ON BOTH MICROBIOLOGY AND THE HOST INFLAMMATORY RESPONSE

Introduction Microbiological culture and histology of deep tissue specimens are independent diagnostic criteria in fracture-related infection (FRI). However, the association between these tests has rarely been investigated, particularly in relation to clinical outcome after surgical treatment. Method Patients having surgery for International Consensus-confirmed FRI were included. All had ≥5 tissue specimens taken for microbiological culture and 2-3 for histology. The correlation between cultured pathogen, histological positivity (defined as ≥5 polymorphonuclear neutrophils/high power field), and outcome at one year after surgery was explored. Results FRI was confirmed in 430 patients (mean age 51.7 years), predominantly in the tibia(194), femur(111), upper limb(70) and ankle(40). 321(74.7%) were culture positive and 334(77.7%) histology positive. 265(61.6%) were positive for both tests. Staphylococcus aureus was cultured in 169(42.5%), CoNs in 61(15.3%) and Gram negatives in 145(36.3%) cases. Virulent organism FRI was strongly associated with positive histology (Odds ratio 2.72;95%CI 1.61-4.58) but not clinical failure (OR 1.74;0.59-5.14). S. aureus was strongly associated with positive histology compared to all other organisms (OR 2.21;1.27-3.87). Surgery succeeded in 390(90.7%) patients. Failure was weakly associated with positive microbiology alone (OR 2.03;0.83-4.96) or positive histology alone (OR 2.13;0.81-5.6). Combined positive culture and histology was strongly associated with clinical failure (OR 2.3;1.06-4.96). There was no difference in outcome between virulent and non-virulent bacteria when histology was positive, but both have higher failure rates compared to patients with negative culture or histology. Conclusion A pronounced inflammatory response, seen in histology, is a feature of virulent bacterial FRI. However, the presence of a virulent pathogen alone does not dictate clinical outcome without marked inflammation. This suggests that outcome is at least as much related to the host response as to the bacterium. When the pathological response is prominent, this may lead to tissue necrosis, further bacterial invasion of adjacent tissues, osteolysis and loss of fracture stability, contributing to treatment failure. This deserves further study to understand the mechanisms behind this interplay and clinical outcome.