High Disease Activity Inuences the Presence of Vertebral Fractures in Rheumatoid Arthritis

Background: To investigate the prevalence and risk factors for vertebral fractures in patients with rheumatoid arthritis (RA) during an era of tight management. Methods: We retrospectively reviewed 426 RA patients who had visited our outpatient RA clinic between July 2017 and June 2020. Among them, we included 107 patients (19 males and 88 females) who had undergone lateral X-ray of the thoracolumbar spine and dual-energy X-ray absorption spectroscopy for the assessment of osteoporosis. We assessed the disease activity score for 28 joints (DAS28), the history of medication for RA and osteoporosis, the number and location of vertebral fractures, and the bone mineral density (BMD). Two board-certied specialists determined osteoporotic vertebral fractures on a lateral X-ray of the thoracolumbar spine. Results: The mean age, average disease duration, and average DAS28 of the analyzed patients were 67.9 years, 14.9 years, and 2.8, respectively. Vertebral fractures were found in 33 patients (30.8%). In this population, 84.8% of patients with vertebral fractures and 59.5% of those without vertebral fractures were treated for osteoporosis with active vitamin D3, bisphosphonate, and/or denosumab. RA patients with vertebral fractures had signicantly higher DAS28 values, a higher rate of patients with a history of glucocorticoid use, and lower BMD in comparison to those than without vertebral fractures (p = 0.009, p = 0.004, and p = 0.01, respectively). Logistic regression analysis showed DAS28 (p = 0.038) and BMD (p = 0.004) were independent factors associated with the presence of vertebral fractures. The ordered logistic regression analysis also showed DAS28 (p = 0.043) and BMD (p = 0.024) were independent factors that explained the number of vertebral fractures. Conclusions:


Background
Patients with rheumatoid arthritis (RA) are more likely to have osteoporosis than the general population and are considered to be at increased risk of fracture (1)(2)(3)(4)(5). The effects of in ammatory cytokines and osteoclast activation, which is associated with the disease itself and the effects of steroids used for treatment have been reported as risk factors for osteoporosis (1,2,5,6). However, in recent years, the arrival of biologic agents has dramatically changed the treatment strategy for RA, and has reduced the risk of fracture in RA patients (5,7).
Vertebral fractures (VFs) are associated with a reduced functional status (8- 15), and the presence of existing VFs are risk factors for a new VF and other fragile fractures (16,17). Therefore, it is important to assess the risk of VF, especially in RA patients with a high risk of fractures. Despite the era of tight control for RA, the risk of VF in RA patients remains inconclusive. Therefore, it is clinically meaningful to investigate the factors that in uence VFs. The aim of the present study was to investigate the prevalence and risk factors for VFs in patients with RA.

Patient
We retrospectively reviewed 426 RA patients who were treated at an outpatient RA clinic in Gunma University Hospital between July 2017 and June 2020. All patients met the American College of Rheumatology RA classi cation criteria in 1987 (18). Among them, we included 107 patients who had undergone lateral X-ray of the thoracolumbar spine and dual-energy X-ray absorption spectroscopy (DEXA) for the assessment of osteoporosis. We did not include patients with a history of thoracolumbar surgery ( Fig. 1). At the time of the survey, the patients had no subjective spinal symptoms. The research protocol was approved by the Institutional Review Board of Gunma University Hospital, and written informed consent for participation was obtained from all patients.

Clinical parameters
Clinical data were collected from medical records including age, sex, body mass index (BMI), RA disease duration, anti-cyclic citrullinated peptide antibody, rheumatoid factor, and treatment history. Disease activity-related parameters, C-reactive protein, erythrocyte sedimentation rate, tender joint count, swollen joint count, and the visual analog scale of patient's global assessment were also collected. The disease activity score for 28 joints (DAS28) was calculated. The Health Assessment Questionnaire-Disability Index was also assessed to evaluate dysfunction in RA patients. In addition, the use of medication for RA and osteoporosis for ≥ 3 months during the follow-up period was recorded.

Measurement of bone mineral density
The BMD of the femoral neck was measured by DEXA using the Discovery A System (Hologic, Inc., Waltham, MA, USA). T-scores were calculated based on normative data for Asian women.

Evaluation of VFs
The evaluation of VFs from T4 to L5 was performed using lateral X-ray of the thoracolumbar spine using a semi-quantitative method (19). Vertebral deformities were classi ed into grades 0 (normal), 1 (mild), 2 (moderate), and 3 (severe) according to the decrease in vertebra height (grade 1, 20-25% decrease; grade 2, 25-40% decrease; and grade 3, ≥ 40% decrease. Grade 1 and above was classi ed as VF in this study. The number of VFs was categorized as 0, 1, 2, or ≥ 3, according to a previous report (17).

Statistical analysis
Statistical analyses for univariate analysis and logistic regression analysis were performed using SPSS Statistics software program (version 25, IBM Corp., Armonk, NY, USA). The Mann-Whitney U test was used to compare continuous data, and the chi-squared test or Fisher's exact test was used for categorical variables. A logistic regression analysis was performed to identify risk factors for pre-existing VF in RA patients and to calculate odds ratio (OR) with 95% con dence interval (CI) for VF. An ordered logistic regression analysis was also performed to investigate risk factors for the categorized number of prevalent VFs in RA patients using the STATA/SE software program (version 15, StataCorp, College Station, TX, USA). P values of < 0.05 were considered statistically signi cant.

Results
The demographics and clinical characteristics of the RA patients in the study population are shown in Table 1. Of 107 RA patients, 19 were male and 88 were female. The mean age was 67.9 years. The average disease duration was 14.9 years and the average DAS28 was 2.9. At the time of the survey, 74 patients (69%) were receiving medication for osteoporosis. A total of 95 VFs were identi ed in 33 patients (30.8%). The VFs showed a bimodal distribution with peaks at the middle thoracic spine and thoracolumbar junction (Fig. 2). Of 33 patients with VF, 28 (85%) had been treated for osteoporosis.
Among those receiving anti-osteoporosis agents, 41% (30/74) had VFs. The prevalence of VFs was similar among patients treated with different anti-osteoporosis agents (active vitamin D3, 41%; bisphosphonate, 42%; denosumab, 46%). Table 2, in patients with VFs, the DAS28 was signi cantly higher (p = 0.009) and the BMD was signi cantly lower (p = 0.001) comparison to patients without VFs. Additionally, the prevalence of a history of corticosteroid use was signi cantly higher in patients with VFs than in patients without VFs (p = 0.004).

As shown in
A logistic regression analysis was performed to identify risk factors for existing VFs in RA patients. RA patient with or without VF was de ned as a dependent variable. When age, sex, BMI, history of corticosteroid use, history of biologic use, DAS28 and BMD were used as independent variables, DAS28 (OR: 1.814, p = 0.038, 95% CI: 1.032-3.187) and BMD (OR: 0.297, p = 0.004, 95% CI: 0.131-0.676) were de ned as independent risk factors for existing VFs in RA ( Table 3). The higher the DAS28, the higher the prevalence of VFs (Fig. 3A).
Next, the risk factors for the number of VFs were investigated by an ordered logistic regression analysis.
The categorized number of prevalent VFs was de ned as the dependent variable, while age, sex, BMI, history of corticosteroid use, history of biologic use, DAS28, and BMD were used as independent variables. As a result, DAS28 (p = 0.043) and BMD (p = 0.024) were identi ed as independent risk factors for the number of VFs (Table 4). Figure 3B shows the number of VFs in each disease activity group.

Discussion
Our study had three main ndings. First, VFs were detected as a morbidity in 30.8% of patients. Second, VFs were frequently observed, even when patients received the recommended anti-osteoporotic agents.
Finally, a high DAS28 and low BMD were independent risk factors associated with the presence and number of existing VFs.
The prevalence of VF in patients with RA has been reported to be 20.2-45.5%, although the background of those reports differed from this study (20)(21)(22)(23). In our study, VFs were frequently observed, even with the therapeutic agents recommended in the treatment guidelines: activated vitamin D3, bisphosphonate, and denosumab (24). The distribution of the sites of VF was similar to previous reports (6,20,22).
In this study, the disease activity of RA was determined as a risk factor for existing VFs in RA patients. Previous reports have shown the relationship between VFs and high disease activity (10,25). In the management of VFs in RA patients, it might be important not only to treat osteoporosis but also to manage disease activity.
The use of corticosteroids was also reported to affect VFs (6,8,9,11,20,22,26). However, in our study, the use of corticosteroids was not identi ed as an independent predictor of VF. One of the reasons might be the low dosage of corticosteroids used in this study. Since methotrexate and biologics were administered reduce RA activity, the dosage of corticosteroid was 3.9 mg/day. Anti-rheumatic biological drugs have been reported to prevent bone loss in RA patients (5,7,27), and tumor necrosis factor inhibitors were associated with reduced incidence of VF (28). On the other hand, other reports demonstrated that biologic agents did not contribute to increased bone density or reduce the risk of fracture (29,30). In this study, we found no association between the use of biologics and VF.
Besides, since our study was a cross-sectional in nature, we could not determine whether VF occurred before or after the use of biological agents. Longitudinal studies should be performed in performed in the future.
The present study was associated with some limitations. First, this was a retrospective study conducted at one hospital. The age and duration of RA were similar to the previous reports, and the effects of these factors may have been limited (6, 12,20). Second, the study population was relatively small. However, we conducted multivariate analyses and the results for the prevalence of VFs were comparable to past studies with larger cohorts (10,25). Therefore, the results of this study considered to be acceptable. Third, since this was a cross-sectional study, changes in disease activity and the therapeutic period were not taken into account. However, we analyzed the history of past drug use. Longitudinal research is needed to investigate the effects of disease activity and drug use on VFs more accurately.

Conclusions
VFs were frequently observed, even with the recommended anti-osteoporotic agents in RA patients. The presence and number of prevalent VFs in RA patients were associated with high disease activity as well as low BMD.
Abbreviations BMD: bone mineral density; BMI: body mass index, CI: con dence interval, DAS28: disease activity score for 28 joints; DEXA: dual-energy X-ray absorption spectroscopy; OR: odds ratio; RA: rheumatoid arthritis; VF: vertebral fracture Declarations Funding There has been no signi cant nancial support for this work that could have in uenced its outcome. The authors acknowledge the nancial support provided by a UCB Japan Grant (UCBJ Grant) for the English proofreading and article processing charge of the journal.

Availability of data and materials
All data generated or analyzed during this study are included in this published article.
Authors' contributions HS and KO made substantial contributions to the design of the study, acquisition, analysis, interpretation of data and writing of the manuscript. YI participated in the design of the study, acquiring data, helped to draft the manuscript and contributed to revising the manuscript critically. AH, ET, TM, SI and KI participated in acquiring data and critically contributing to revising the manuscript. TS, TK and YY contributed to revising the manuscript critically. HC critically contributed to revising the manuscript enhancing its intellectual content and approving the nal content of the manuscript. All authors read and approved the nal manuscript.
Ethics approval and consent to participate Ethical approval was obtained from the institutional review board of the Gunma University Hospital. Patients provided written informed consent before taking part in this study.     Figure 1 Flow diagram of the present study. ACR, American Rheumatism Association; RA, rheumatoid arthritis.