Patients with axial spondyloarthritis (axSpA) treated with tumor necrosis factor (TNF) inhibitors were at significantly lower risk for hip and spine fractures compared with those receiving only nonsteroidal anti-inflammatory drugs (NSAIDs), a large claims-based study indicated.
In records from the Merative MarketScan Database, which covers nearly everyone in the U.S. with employer-paid insurance, axSpA patients had nearly 30% lower odds for hip and spine fracture if they were on anti-TNF drugs after adjusting for age, sex, and year of diagnosis (OR 0.71, 95% CI 0.59-0.85), according to Maureen Dubreuil, MD, MSc, of Boston University, and colleagues.
Adjustment for additional parameters such as corticosteroid use and comorbid osteoporosis attenuated the risk decrease only slightly (OR 0.75, 95% CI 0.62-0.91), the researchers reported in Arthritis & Rheumatologyopens in a new tab or window.
Conventional disease-modifying antirheumatic drugs (DMARDs), on the other hand, had no apparent effect on fracture risk either in the minimally adjusted analysis (OR 0.96, 95% CI 0.76-1.23) or with all the measured potential confounders included (OR 0.93, 95% CI 0.72-1.19).
"Fracture is an outcome of concern in axSpA given its association with morbidity and mortality," Dubreuil and colleagues explained, with epidemiological data confirming that bone breaks are more common in these patientsopens in a new tab or window compared with the population at large. "Local bone remodeling" in SpA is believed to raise risk for vertebral fractures, they added, while systemic inflammation likely drives weakening in other bones.
Current guidelines recommend NSAIDs as first-line treatment for axSpA, to be followed by TNF inhibitors if disease activity remains unacceptably high. Conventional DMARDs and steroids are discouraged unless there is non-axial involvement. (Advanced therapies with non-TNF targets may be used as third-line therapy; these were not addressed in the current study.)
Yet while anti-TNF drugs are well known to improve the major signs and symptoms of axSpA, their effects on fracture risk remain unstudied, apart from a few small analyses of bone density.
To address this gap, Dubreuil and colleagues identified 13,519 patients with axSpA from 2006 (when TNF inhibitors began to be used in the disease) to 2021 in the MarketScan database, among whom 1,229 cases of spine or hip fracture were recorded. For each of these cases, up to 10 axSpA patients without fractures were then selected as controls -- not matched by demographics or other parameters, as Dubreuil's group believed that would introduce confounding. The "index date" for fracture cases was the date of fracture; for the corresponding controls, it was a random date from the same year. All patients in the study had to have at least 1 year of continuous insurance prior to the index date.
Since patients and controls weren't matched, the two groups differed considerably. Mean age for fracture cases was about 53 versus 47 for controls; some 38% of fracture cases were women as opposed to 45% of controls. Not surprisingly, osteoporosis was more common in fracture cases, as were other factors associated with fracture risk such as tobacco and steroid use.
At least numerically, patients who had suffered previous fractures saw greater reductions in risk for new ones when taking TNF inhibitors, compared with those without such history (OR 0.59, 95% CI 0.23-1.51, vs OR 0.83, 95% CI 0.67-1.03, with NSAID users as the comparators).
A surprising number of patients among both fracture cases and controls -- nearly half -- were not using any of the three medication classes examined in the study. Men in the "no medication" group actually showed numerically greater fracture risk than NSAID users (OR 1.15, 95% CI 0.93-1.42, in the fully adjusted model); women, however, had virtually the same risk with "no medication" versus NSAIDs (OR 1.01).
"Our study demonstrates a protective effect of TNF [inhibitors] on fracture risk in axSpA compared with use of NSAIDs or [conventional] DMARDs," Dubreuil and colleagues wrote. "Future research investigating the impact of the timing of TNF [inhibitor] initiation and alternative axSpA treatment modalities, such as [interleukin-17 and JAK inhibitors], on fracture risk in axSpA will further enhance our understanding of how fracture risk may be mitigated in this population."

Limitations to the study included its reliance on administrative data, as well as a lack of information on relative disease activity, medication doses or subtypes, and patients' body mass index values. Also, Dubreuil's group noted that patients not using TNF inhibitors could have had contraindications for them that might also raise fracture risk. Finally, the relatively small number of fracture cases in the overall total of axSpA patients limited the analyses that could be conducted with meaningful results.