In this study, we investigated the incidence and predictors of fragility fractures in postmenopausal patients with RA who were prescribed oral BP. The cumulative incidence of fragility fracture was 10.5% at a median follow-up of 28 months. We found that a history of prior vertebral fracture and methotrexate dose were positive and negative predictors, respectively, of the development of future fragility fractures. The cutoff dose of methotrexate for future fragility fracture was 4 mg/week. The cumulative incidence of fragility fracture in the patients with both risk factors, a history of prior vertebral fracture and ≤ 4 mg/week dose of methotrexate, was significantly higher than the patients without risk factors.
There are only scarce studies investigating the mid- to long-term incidence of fractures in patients with RA receiving anti-osteoporotic medication (including oral BP). Kwon et al. reported a 3-year fracture incidence of 17.4% in Asian patients with RA aged 50 years or older from a nationwide claim database, with no mention of osteoporosis treatment [30]. Katayama et al. reported that 12.5% of patients older than 50 years old with RA who were prescribed alendronate developed vertebral or nonvertebral fracture during an average follow-up period of 32.8 months [31]. Taking into consideration our result of a 10.5% cumulative incidence of fracture at 28 months, we observed a certain degree of fracture prevention effect, however, some patients with RA are still at risk, even if they are receiving oral BP. Other new anti-osteoporotic drugs have recently become available, such as denosumab and teriparatide, which probably have better fracture prevention effects in patients with RA [10–12]. However, initiating treatment with these recent anti-osteoporotic drugs in all RA patients is very likely to be too costly; therefore, it is important to select patients at risk despite the prescription of oral BP.
In the present study, we identified a history of prior vertebral fracture as a positive predictor of fragility fractures. A systematic review on osteoporosis concluded that a history of prior fracture was an important risk factor for future fractures [32]. Furthermore, several previous studies found that in patients with RA, a history of prior fracture was a risk factor for incident fractures [2, 16, 33] with or without anti-osteoporotic medications. Our results suggest that in postmenopausal RA patients, a history of vertebral fracture is a strong predictor of fragility fractures, even among those who are taking oral BP medication.
In the present study, we found that the dose of methotrexate was a negative predictor and a dose of ≤ 4 mg/week dose was a risk factor for future fragility fractures. We suggest that there are two possible reasons why the outcome of methotrexate dose is a negative predictor. First, one of the reasons for restricting the dose of methotrexate is associated renal dysfunction that increases the incidence of any fracture risk [34]. Renal dysfunction in patients receiving no or low-dose methotrexate might affect the development of fragility fractures; however, we did not include an assessment of renal function in the present study. Second, previous studies reported that a combination of methotrexate and BP prevented systemic bone loss in an in vivo animal model of arthritis [35, 36] as well as in patients with RA [37]. There is a need for future clinical investigations to determine whether the combination of methotrexate and BP can decrease the development of fragility fractures.
The present study had several limitations. First, this was a retrospective longitudinal study limited by the use of data collected from patient charts. We did not evaluate any factors that might have affected the development of fragility fracture, such as renal dysfunction, smoking, alcohol intake or the Fracture Risk Assessment Tool score. However, we included almost every factor correlated with fragility fractures in patients with RA, as reported in previous studies. Second, in the present study, we treated alendronate, risedronate, and minodronate identically to oral BP. Several previous studies reported that these medications had different efficacy in patients with RA [31, 38, 39]. Thus, the type of oral BP might affect the outcome. However, in the univariate Cox regression analysis, we did not find the type of oral BP to be a statistically significant factor. Third, we took several measures to ensure medication adherence, but no specific instrument was used to assess adherence to oral BP. Finally, this was a single-center study with a small population of patients. Thus, our results on the predictors for fragility fractures should be interpreted with caution. Larger, multicenter studies are warranted to obtain greater accuracy and generalizability of the study findings.
The present study investigated the incidence and predictors of fragility fractures in postmenopausal women with RA associated with osteoporosis who is receiving oral BP. We found a cumulative incidence of any fragility fractures of 10.5% at a median follow-up of 28 months. A history of prior vertebral fractures was a positive predictor, whereas the dose of methotrexate was a negative predictor of the development of fragility fractures. Practitioners should consider selecting another anti-osteoporotic drug in patients with RA who remain at risk despite receiving oral BP.