Risk of Hip and Spine Fractures in Axial Spondyloarthritis Is Associated with Treatment Class

Devin Driscoll¹, Navya George², S. Reza Jafarzadeh³, Christine Peloquin³, Jean Liew³ and Maureen Dubreuil⁴, ¹Boston Medical Center, Boston, MA, ²Boston University Medical Center Internal Medicine Residency Program, Boston, MA, ³Boston University Chobanian & Avedisian School of Medicine, Boston, MA, ⁴Section of Rheumatology, Boston University School of Medicine, Boston, MA

Meeting: ACR Convergence 2024

Keywords: Fracture, Pharmacoepidemiology, Spondyloarthropathies, TNF-blocking Antibody

Session Information

Date: Sunday, November 17, 2024

Title: SpA Including PsA – Diagnosis, Manifestations, & Outcomes Poster II

Session Type: Poster Session B

Session Time: 10:30AM-12:30PM

Background/Purpose: Individuals with axial spondyloarthritis (axSpA) have a risk of fracture that is nearly double that of the general population, possibly related to chronic inflammation. Vertebral fracture risk may be due to regional osteopenia paired with regions of excess bone stiffness. We assessed the impact of treatment with tumor necrosis factor inhibitors (TNFi) and non-biologic disease-modifying antirheumatic drugs (DMARDs) on hip and spine fractures in axSpA, relative to nonsteroidal anti-inflammatory drugs (NSAIDs).

Methods: We conducted a nested case-control study using 2006-2021 data from the US-based Merative^TM MarketScan® Database (Figure 1). We included adults 18-65 years old with ≥1 inpatient or ≥2 outpatient axSpA ICD-9 or 10 diagnosis codes. The primary outcome was hip and/or spine fracture, defined by diagnosis or procedure codes. For each fracture case, we selected up to 10 controls without fracture. We evaluated medication use hierarchically (TNFi, DMARDs, NSAIDs [referent], or none) using pharmacy claims and procedure codes (for infusions). We assessed the odds of hip or spine fracture with medication class using unconditional logistic regression with adjustment for potential confounders.

Results: Our main analysis included 13,519 individuals with axSpA, comprising 1,229 hip/spine fracture cases and 12,290 controls. Hip/spine fracture occurred in 9% of the study sample. Among fracture cases, 25% were TNFi users, 9% were DMARD users, 18% were NSAID users and 48% were not on any medication. Individuals on TNFi had 29% lower odds of fracture compared to those on NSAIDs (OR 0.71, 95% CI 0.59-0.85), accounting for age, sex, and diagnosis year (Figure 2). Results for TNFi were similar in the fully adjusted model (OR 0.75, 95% CI 0.62-0.91) and when stratified by sex. There was no protective effect of DMARDs.

Conclusion: Using a large US insurance claims database, we observed a protective effect of TNFi on hip/spine fractures in axSpA compared with use of NSAIDs or DMARDs. These findings suggest a beneficial effect of TNFi in bone remodeling in axSpA. By reducing systemic inflammation and slowing local bone remodeling, TNFi use may help preserve both bone density and architecture, thus lowering risk for fractures. Further research investigating the impact of timing of TNFi initiation and alternative axSpA treatment modalities, such as IL-17 inhibitors and Janus kinase (JAK) inhibitors, on fractures will enhance understanding of how to mitigate fracture risk in axSpA.

Figure 1. Continuous enrollment, exposure assessment period, and covariate assessment period in relation to index date and disease date. Index date: fracture date for cases, and random date from the corresponding case’s fracture year for controls. Disease date can be any time prior to index date. Continuous Enrollment (CE) Period was required to include 1 year prior to index date. Exposure Assessment Period was 6 months prior to index date. Covariate Assessment Period was 2 months prior to Exposure assessment period.

Figure 2. Odds of hip or spine fracture associated with axial spondyloarthritis therapeutic class, with NSAIDs as the referent. Tumor necrosis factor inhibitors (TNFi) include etanercept, adalimumab, golimumab, certolizumab and infliximab. DMARDs include apremilast, auranofin, azathioprine, chloroquine, cyclophosphamide, cyclosporine, gold sodium thiomalate, hydroxychloroquine, leflunomide, methotrexate, minocycline, mycophenolate, sulfasalazine. Minimally adjusted model adjusted for age, sex, and diagnosis year. Fully adjusted model adjusted for age, sex, disease year, alcohol use disorder, antiepileptic drug use, BMI, breast cancer, chronic kidney disease, falls, glucocorticoid use, inflammatory bowel disease, osteoporosis, osteoporosis medication use, prostate cancer, tobacco use, ESR/CRP laboratory orders and number of outpatient visits (rheumatology and primary care).

Disclosures: D. Driscoll: None; N. George: None; S. Jafarzadeh: None; C. Peloquin: None; J. Liew: None; M. Dubreuil: Amgen, 2, Pfizer, 5, UCB Pharma, 2.

To cite this abstract in AMA style:

Driscoll D, George N, Jafarzadeh S, Peloquin C, Liew J, Dubreuil M. Risk of Hip and Spine Fractures in Axial Spondyloarthritis Is Associated with Treatment Class [abstract]. Arthritis Rheumatol. 2024; 76 (suppl 9). https://acrabstracts.org/abstract/risk-of-hip-and-spine-fractures-in-axial-spondyloarthritis-is-associated-with-treatment-class/. Accessed .

« Back to ACR Convergence 2024

ACR Meeting Abstracts - https://acrabstracts.org/abstract/risk-of-hip-and-spine-fractures-in-axial-spondyloarthritis-is-associated-with-treatment-class/