Abstract
Aseptic loosening of implants or loosening due to persistent bacterial infection remains a severe complication in orthopaedic surgery. To investigate underlying cellular and molecular mechanisms, particularly with regard to bone loss, tissue samples of patients requiring surgery were examined. By histological methods and by quantitative RT-PCR, respectively, infiltration of leukocytes, expression of osteoclast-typical genes and of proinflammatory cytokines was determined. Samples were taken directly from osteolytic sites and for comparison from adjacent sites, distant sites and from muscle. At osteolytic sites, cathepsin K and the metalloproteinases MMP1 and MMP9 were found, as was expression of inflammation-related cytokines, particularly of interleukin (IL)-1β, CXCL8, S100A9 and a very moderate expression of receptor activator of NfκB ligand (RANKL) and tumour necrosis factor (TNF) a. Of note, expression of these parameters gradually decreased from sites of osteolysis to adjacent tissue, to distant tissue to muscle. In patients with infection and osteolysis, expression of cytokines, notably of CXCL8, was markedly enhanced, especially in adjacent and distant tissues, where expression was 10- to 20-fold higher compared to tissue of aseptic patients. A possible source of CXCL8 could be infiltrated cells, particularly neutrophils, because they were found in infected tissue only. Histological examination of the biopsies revealed an additional CXCL8 source, namely endothelial cells of small blood vessels. In conclusion, aseptic loosening and implant-associated infection are associated with osteoclast generation and a local inflammatory response. The proinflammatory environment could promote the differentiation of precursor cells to osteoclasts, thereby linking inflammation to bone resorption. The higher expression of cytokines, particularly of CXCL8 in tissue of patients with bacterial infection, could explain the accelerated time course of bone resorption as it occurs in infection compared to aseptic loosening.