Background/Purpose: Although early intensive treatment has improved medication of rheumatoid arthritis (RA), treatment for osteoporosis (OP) in RA patient will be more important. Here, we report 2-year outcome of denosumab (DMB) on OP of RA patients. This was a multiple center, prospective study (TBCR-BONE) to investigate the efficacy of DMB for 2 years on OP of RA patients and the predictors for the efficacy in DMB treatment.

Methods: 74 females completed 24-month (m) DMB treatment and were used for the analysis. Bone mineral density (BMD) of lumbar spine (LS) and total hip (TH) by DEXA, and P1NP and TRACP-5b were measured every 6m.

Results: Mean age was 70.2 years old. Mean RA duration was 17.1 years. Oral PSL was used in 26 cases. Biological agents (BIO) were used in 17 cases. 33 cases had the past history of fractures. The rate of change in LS-BMD (%LS-BMD) significantly increased as 4.4% at 6m, 5.8% at 12m, 7.2% at 18m and 7.6% at 24m. %TH-BMD significantly increased as 2.7% at 6m, 3.3% at 12m, 4.6% at 18m and 4.6% at 24m. %P1NP at 24m was -28.1%. %TRACP-5b at 24m was -20.6%. To confirm the predictors for large increase of %LS-BMD at 24m and %TH-BMD at 24m, all cases were divided into two groups, good outcome group (LS-GO group: n=47, mean %BMD at 24m=11.4% and TH-GO group: n=47, mean %BMD at 24m=6.8%) and no good outcome group (LS-NGO group: n=24, mean %BMD at 24m=0.18% and TH-NGO group: n=24, mean %BMD at 24m=0.21%) with the cutline of %LS-BMD at 24m (4.0%) and %TH-BMD at 24m (3.1%) made of each superior two thirds of all cases. We performed multivariate logistic regression analyses and confirmed large decrease of %TRACP-5b at 12m was associated with large increase of %LS-BMD at 24m [OR (%) 0.97, 95% CI 0.95-0.99], and concomitant use of BIO was associated with large increase of TH-BMD at 24m (OR 5.8, 95% CI 1.10-30.64). The cutoff value of %TRACP-5b at 12m for larger increase of %LS-BMD at 24m was -38.7% (AUC of ROC=0.766). We divided all cases into two groups, large decrease of %TRACP-5b at 12m group (L group: n=36) and small decrease of %TRACP-5b at 12m group (S group: n=37) with this cutoff value. %LS-BMD at 24m was 8.9% vs. 5.9% (p<0.01), %TH-BMD at 24m was 5.0% vs. 4.2% (n.s.), %P1NP at 24m was -59.6% vs. 0.6% and %TRACP-5b at 24m was -48.8% vs. 5.6% (L group vs. S group). At last, to evaluate the efficacy of combination of BIO and DMB, all cases were divided into two groups, BIO used group (B group: n=17) and no BIO used group (NB group: n=57). %LS-BMD at 24m was 9.7% vs. 7.0% (n.s.), %TH-BMD at 24m was 6.5% vs. 4.0% (p<0.01), %P1NP at 24m was -33.8% vs. -26.5% and %TRACP-5b at 24m was -26.7% vs. -18.7% (B group vs. NB group).

Conclusion: DMB was effective in OP of RA patients. Large decrease of %TRACP-5b at 12m was a predictor for large increase of %LS-BMD at 24m. Combination of BIO and DMB was a predictor for large increase of %TH-BMD at 24m. They might suggest that there are different mechanisms associated with increase of LS-BMD and TH-BMD in DMB treatment for OP of RA patients.