Background/Purpose:

Chronic inflammation in rheumatoid arthritis (RA) may interfere with bone remodelling. Inflammation mediators such as TNF-α stimulate osteoclast formation which favors bone destruction. Small prospective studies have suggested biologic DMARDs preserve bone mineral density at 6-12 months, but impact on fracture risk is yet to be characterized. Our objective was to determine the risk of non-vertebral osteoporotic fractures in RA subjects aged ≥50, comparing outcomes in patients who were exposed or unexposed to biologic DMARDs.

Methods: A population based, retrospective, nested case-control study from January 1 2002 to December 31 2008 was conducted, using Quebec physician billing and hospital discharge data. RA subjects were identified from ICD-9/10 codes in billing (2 RA codes, ≥2 weeks apart but within 2 years) and hospitalization (any RA code) data. Subjects were followed until the earliest of: 1) non-vertebral osteoporotic fracture (index date), 2) death, or 3) end of study period. A validated algorithm identified non-vertebral osteoporotic fractures from physician claims. Controls were selected with incidence density sampling, matched to cases (4 to 1 ratio) on age, sex, and date of study entry. Biologic DMARD exposure was defined as being on treatment for ≥180 days pre-index. Conditional logistic regression was used, adjusting for: RA duration, comorbidity, history of previous fracture/arthrosplasty, use of traditional DMARDs, osteoporosis medication, hormone replacement, corticosteroids, proton-pump inhibitors, anticonvulsants, opioids, non-steroidal anti-inflammatories, antidepressants, anxiolytics, antipsychotics, medical visits and hospitalizations.

Results: The cohort included 29,666 RA subjects aged ≥50. Over the study period, 1,963 fractures were captured (incidence rate: 13.4 per 1000 person-years). Of these, 160 were excluded because they either occurred during the first 180 days of follow-up (where by definition, one could not be exposed to biologic) or controls could not be found. The final number of cases was 1,803 (with 7,175 controls). The most frequent fracture site was hip/femur (43.7%). Cases and controls were mainly older women and most (n=1,233, 68.4%) had a RA duration ≥5 years at the index date. In total, 190 subjects (53 cases, 137 controls) were exposed to biologic DMARDs for ≥180 days. We were unable to demonstrate a statistically significant association between biologic DMARDs and fracture risk (Odds Ratio, OR [95% Confidence Interval, CI]: 1.16 [0.51-2.62]). RA duration had the strongest impact on fracture risk; for subjects of RA duration ≥10 years (vs. <5 years), the OR was 6.40 (95% CI 3.57-11.46), while those with RA duration 5-10 years (vs. <5 years) had an OR of 3.05 (95%CI 1.90-4.89). The inability to detect an effect remained when varying exposure definition, induction period, and definition of corticosteroid use, and when restricting by history of fracture, RA duration, or follow-up time.

Conclusion: Despite the positive impact of biologic DMARDs on bone remodelling observed in small prospective studies, we were unable to demonstrate a reduction in the risk of non-vertebral osteoporotic fractures in older adults with RA.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/biologic-disease-modifying-anti-rheumatic-drugs-and-the-risk-of-non-vertebral-osteoporotic-fractures-in-patients-with-rheumatoid-arthritis-aged-50-years-and-over/