RETHINKING BACTERIAL OSTEOLYSIS: EVIDENCE BEYOND THE RANKL PARADIGM

Background Bone and joint infections (BJIs) are debilitating conditions characterized by intraosseous bacteria, biofilm formation, suppurative inflammation, and osteolysis. Bacteria-associated bone loss is traditionally attributed to RANKL-mediated osteoclast activation, largely based on in vitro and murine studies. However, emerging translational evidence from large-animal models and clinical studies suggests that this paradigm may not fully explain bacteria-associated osteolysis. Methods The translational study applied a porcine implant-associated osteomyelitis (IAO) model and patients with fracture-related infection (FRI) to investigate pathophysiological mechanisms of infectious osteolysis. RANKL signalling was inhibited using Denosumab in vivo, and bone loss was assessed by CT and histology, while osteoclast activity was assessed by immunohistochemistry and qPCR. The FRI patients were categorized as osteolytic (n =14) and non-osteolytic (n =7) based on CT scans. Biopsies collected at revision surgery were assessed by histology to quantify osteoclast numbers and by mRNA in situ hybridization to quantify RANKL, MMP1, and C3 expression. Results In the IAO model, RANKL inhibition with Denosumab did not reduce radiographic osteolysis compared to infected controls; CT-defined bone loss was even increased in treated animals (1.25 vs. 0.63 cm 3 , p=0.005) despite a marked reduction in osteoclast numbers. Furthermore, local RANKL expression did not correlate with bone loss. In FRI patients, osteoclast numbers were not associated with osteolysis severity (ρ=0.34, p=0.21). In contrast, MMP1 expression was significantly increased in patients with osteolytic lesions, whereas RANKL expression remained unchanged between osteolytic and non-osteolytic groups. Spatial analyses demonstrated heterogeneous expression of MMP1 and RANKL, in contrast to uniform C3 expression. Osteolysis severity correlated with MMP1 spatial burden (ρ=0.74, p=0.011). Discussion These findings shows that bacterial-associated osteolysis is not primarily driven by RANKL-mediated osteoclastogenesis but is a multifactorial process shaped by inflammation, proteolysis, neutrophil responses, and impaired osteogenesis. This challenges the prevailing RANKL paradigm and supports a broader osteoimmunological model, or conceptualization, of infectious bone loss.