Abstract
Background
The process of osteolysis is well studied both in vivo and in vitro. Although multiple pathways have been implicated in osteolytic change and animal models have been developed there are few human tissue studies. There are no extensive human tissue studies comparing osteoarthritic hips to well fixed and loose prostheses.
Methods
We have investigated 96 genes previously implicated in the osteolytic pathway. Genes were included based on previous implication in osteolysis in basic science studies. Candidates included cytokines, growth factors, apoptotic factors, matrix proteinases, interleukins, apoptotic proteins and macrophage activators.
Results
One hundred patients were enrolled into the study and had intraoperative hip tissue removed after ethics approval. Patients were recruited from three cohorts, those undergoing primary hip replacement, revision of a well fixed prosthesis and revision for osteolytic change. A low density Taqman array method was used to determine gene expression for the 96 candidate genes. Expression of five housekeeping genes was measured and expression normalised between the samples. Statistical analysis was undertaken using significance testing and ROC analysis. There were seven candidate genes that were statistically significantly linked to aseptic loosening (p< 0.05) and strongly associated (AUC >0.77); these were BMP4, Frizzled related protein, fibroblast growth factor 18, IL8, IRAK 3, osteoprotegrin and PTGS2. There were a further nine genes which were highly predictive of osteolytic change (AUC >0.77), but did not reach significance (p>0.05): VEGFB, SFRP, TLR3, TLR5, TP53, IGF1, CTSK, CHIT 1 and CCL 18.
Conclusion
We have been able to distinguish for the first time between factors which are associated with osteolytic change, those associated with exposure to wear debris in a well fixed prosthesis and those associated with the process of osteoarthritis.
