INVESTIGATING OSTEOLYSIS IN OSTEOMYELITIS: WHAT TO BELIEVE?

Aim The osteolytic process of osteomyelitis is, according to textbooks, caused by increased osteoclast activity due to RANKL production by osteoblasts. However, recent findings contradict this theory. Therefore, the aim was to investigate, in a porcine osteomyelitis model, how osteolysis is affected by massive inflammation and RANKL blocking, respectively. In parallel, patients with chronic osteomyelitis, diabetes, foot osteomyelitis, and fracture related infections (FRI) were included for advanced histological analysis of osteolysis. Methods In pigs, a tibial implant cavity was created and inoculated with 10 4 CFU of Staphylococcus aureu s: Group A (n=7). Group B (n=7); + 1cm 3 spongostan into the cavity. Group C (n=4); + systemic Denosumab treatment. Spongostan was used as an avascular material to support bacterial growth and thus increase the inflammatory response. Denosumab treatment was administrated to suppress osteoclast activity by RANKL inhibition (as in osteoporotic patients). The volume of osteolysis was accessed by CT scans. Immunohistochemistry with antibodies towards Cathepsin K was used to identify osteoclasts within the bone lesions. Briefly, the number of Cathepsin K positive cells, i.e., both precursors and bone resorbing osteoclasts, respectively, were counted in 10 high power fields (400x). In total, 50 bone infection patients were included (Herlev Hospital). From each patient five parried samples were taken for histology and microbiology, respectively. Histopathology, CT osteolysis volume estimation, and molecular expression of osteoclasts and inflammatory markers are ongoing. One FRI patient was osteoporotic and treated with Denosumab for 6 years. Results All pigs were confirmed infected in the implant cavity. The volume (2.41 ± 1.29cm 3 ) of osteolysis was significantly increased in the spongostan group in comparison to Group A (1.24 ± 0.59 cm 3 ) (p=0.04). Thereby, the spongostan group had bacteria deeper into the bone from the inoculation point. Sufficient Denosumab treatment, i.e. reduced serum Ca was seen in 3 pigs. None of the Denosumab treated pigs showed reduced osteolysis in comparison to Group A (1.42 ± 0.63 cm 3 ). The Cathepsin K score of Group C was 17 (15-23 IQR) of precursor osteoclasts and 2 (0-2 IQR) of osteoclasts in Howship lacunae. The Denosumab treated patient showed substantial osteolysis and histological analysis confirmed acute inflammatory. Conclusions Application of spongostan, i.e., bacterial host optimization and massive inflammation promotes osteolysis and local bacterial dissemination. Osteoclast blocking with Denosumab showed no impact on osteolysis. Elucidation of the pathophysiology causing bone loss in osteomyelitis is fundamental. However, the widely accepted osteoclast-based theory might not be the only relevant.