Background Patients with rheumatoid arthritis (RA) are at increased risk of osteoporosis. Bone mineral density (BMD) changes related to disease activity, use of steroids or immobility have been described in RA. Tumor necrosis factor alpha (TNFα) inhibitors (TNFi) are effective in controlling inflammation and joint bone erosion. TNF-α plays a role in bone metabolism, directly through Wnt pathway and through RANKL/osteoprotegerin balance and indirectly through inflammation. In several prospective open label studies, TNFi treatment seems to have a beneficial effect on BMD.
Objectives To evaluate the impact of TNFi on BMD and on markers of bone remodelling in patients with RA.
Methods We perform a systematic review and a meta-analysis using Medline, Embase and Cochrane databases and ACR 2012 abstracts. We included two types of studies. Studies of RA patients treated with TNFi and a control group of RA patients without the treatment of interest were included in the meta-analysis. Studies assessing the evolution of BMD or bone remodelling markers under TNFi were included in the systematic review. Reports describing the effect of TNFi on BMD (g/cm2 ) at hip and at lumbar spine, serum osteocalcin (OTC) and serum crosslaps (CTX) were included. The search retrieved 340 articles. 6 articles were controlled of which 2 were analysed for hip and lumbar BMD (229 patients; n=108 and n=121) and 2 for bone remodelling markers (219 patients in OTC trials; 229 in CTX trials). Statistical analysis of difference between each group was performed by comprehensive meta-analysis. The heterogeneity between studies was assessed using the Cochran's Q-test and the I2 value. All the analyses were realized with the RevMan software 5.1 version. A significant statistical threshold of 0.05 was used. 8 articles were included in the review of literature concerning BMD, 4 articles concerning CTX and 5 concerning OTC.
Results Concerning the meta-analysis on controlled studies, patients with TNFi had similar BMD modifications compared to patients without TNFi. The mean BMD difference was 0.00 (-0.03,0.04;95%CI);I2=0% and 0.00(-0.004,0.004;95%CI);I2=0% at lumbar spine and hip respectively. OTC mean difference was 3.33 (-12.94,19.60; 95%CI);I2=98% and CTX mean difference -0.22 (-0.61,0.17; 95%CI);I2=0% Studies duration was heterogeneous (6 to 12 months). Concerning the systematic review of literature, 3 studies showed an increase, 4 a decrease and 1 a stability of hip BMD: largest variation and minimal variation were +13% and -2.6% respectively. 1 study show statistically significance. Concerning lumbar BMD, 5 studies showed an increase, 3 studies a decrease: maximal and minimal variation were +3% and -3.3% respectively. Maximal and minimal CTX decrease were -52% and -2.5%; Maximal and minimal OTC differences were +80.6% and -16.3% respectively.
Conclusions Only few controlled trials have studied effects of TNFi on bone metabolism and there are heterogeneous by way of patients included, disease duration or follow up. For the others biologics drugs, a recent open label prospective study show that BMD remains stable during therapy with TCZ in patients with active RA. TNFi in RA does not seem to significantly affect BMD and bone remodelling markers. However, only few data are available. Long term trials are required to better evaluate the effect of TNFi on BMD in RA.
Disclosure of Interest None declared