Mechanism of bone loss in rheumatic diseases

Abstract

Osteoporosis and fragility fractures are recognized complications of inflammatory rheumatic diseases. This is thought to result from the effects of chronic inflammation, relative immobility and corticosteroid use. A rare syndrome of osteoporosis in a patient with coeliac disease has been described which results from production of neutralizing antibodies to the bone protective protein osteoprotegerin (OPG). The aim of my thesis is to evaluate prevalence and clinical predictors of osteoporosis in a contemporary cohort of patients with rheumatoid arthritis (RA) and to investigate the role of OPG autoantibodies in the pathogenesis of osteoporosis in rheumatic diseases. In a retrospective cohort study, I found that the overall prevalence of osteoporosis in patients with RA was 29.9% which is in keeping with older reports that recorded a prevalence rate between 17% and 36%. In our contemporary cohort osteoporosis was significantly more common than in a gender and age matched control cohort (17.4%). Further analysis showed that only age and BMI were independent predictors of osteoporosis in RA. A predictive tool based on age and BMI was developed which had 91.4% sensitivity for the detection of osteoporosis in an independent RA population. I went on to screen for the presence of autoantibodies to OPG in patients with various rheumatic diseases. In a study of 75 patients with rheumatoid arthritis and 199 healthy controls OPG autoantibodies were detected in two controls (1%) compared with seven patients with RA (9.3%). The RA patients with detectable OPG antibodies had a longer disease duration, higher DAS28 scores and higher levels of the bone resorption marker CTX than RA patients who did not have autoantibodies. Purified IgG from patients with high levels of OPG antibodies blocked the ability of recombinant OPG to inhibit RANKL induced NFκB activation in a HEK293 cell based assay indicating that they were functional. In a further study of 134 patients with ankylosing spondylitis (AS), 16 patients (11.9%) had detectable OPG antibodies. The presence of OPG-Ab was independently associated with reduced hip bone mineral density and an increased risk of fractures in this population. In patients with a longer disease duration we have also observed that there was a higher discrepancy between spinal and hip BMD in OPG-Ab positive patients compared with OPG ab negative patients (p=0.003). In order to investigate if OPG antibodies affected measurement of serum RANKL concentrations as detected by ELISA using OPG as the capture reagent, I measured OPG ab and free RANKL concentrations in 55 rheumatic disease patients. Surprisingly there was a significant positive correlation between free RANKL and OPG Ab concentrations (r=0.430, p=0.001) which was the opposite to what I had expected. These findings reject the hypothesis that OPG ab block binding of synthetic OPG to RANKL in the ELISA. In conclusion, I have shown that osteoporosis is a common complication in RA and I have developed a new risk prediction tool for the use in clinical practice. I have also found that OPG antibodies are produced more commonly in patients with RA and AS than in healthy controls and that antibody levels correlate with bone resorption markers in RA and bone mineral density in AS patients. In vitro studies have shown that some OPG antibodies have functional effects on RANKL signalling. These findings raise the possibility that OPG antibodies may contribute to the pathogenesis of local and systemic bone loss in rheumatic diseases and signal the need to study the relationship between these antibodies and bone disease in large-scale longitudinal studies.