Association between serum uric acid concentration and radiographic axial spondylarthritis: a cross-sectional study of 202 patients.

To the Editor: The risk factors for the development of radiographic axial spondyloarthritis (SpA) are still uncertain. Previous studies found that a high concentration of serum uric acid (SUA) leads to deposition of monosodium urate (MSU) in different joints of other rheumatic arthritis diseases except gout. Furthermore, it has been reported that gout could mimic SpA. Thus, our study explored whether the development of radiographic axial SpA is related to high SUA concentration. The study was approved by the Ethics Committee of Shenzhen Second People's Hospital (No. 20200224001). Written consent was not required for this retrospective study as individual privacy and commercial interests were not involved, according to the medical ethical law of China. We utilized data from a cross-sectional study in Chinese Shenzhen Second People's Hospital from January 1, 2016, and December 31, 2018, which included 202 SpA patients. We searched keyword of SpA by inputting the Chinese character for SpA in the discharge diagnosis. Demographics and disease characteristics were collected. Disease activity was detected by the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI). The Ankylosing Spondylitis Disease Activity Score (ASDAS) was calculated with the C-reactive protein (CRP). Disease severity was measured by the Bath Ankylosing Spondylitis Functional Index (BASFI). Past and current medications, including non-steroidal anti-inflammatory drug (NSAID), corticosteroids, conventional synthetic disease-modifying antirheumatic drugs (DMARDs), and tumor necrosis factor inhibitors (TNFis), were also collected. Extra-articular manifestations including uveitis, psoriasis, and IBD (Crohn's/ulcerative colitis) were examined. Furthermore, additional information about human leucocyte antigen B27 (HLA-B27) and sacroiliac joint imaging (including radiography, computed tomography [CT]) was extracted from the clinical laboratory system and imaging system. The unit of measurement for uric acid was μmol/L, which was relatively small for observing changes. Thus, we transformed the unit of measurement for uric acid from μmol/L into 10 μmol/L, after collecting the raw data on uric acid levels. We examined the sacroiliac joint imaging for all patients. Radiographic axial SpA is defined as grade II bilaterally or grade III to IV unilaterally, according to the Modification of the New York Criteria for radiography. It can also be defined as an erosion score and/or joint space score of two or higher in any of the 24 regions of both joints for CT. Non-radiographic axial SpA is defined as sacroiliac joint image without structural sacroiliitis described above.[1] Sacroiliac images were evaluated by a rheumatologist and a radiologist separately, independently, and retrospectively. Discordant examinations of CT were discussed between the rheumatologist and the radiologist and would be reported to the senior radiologist to establish a final consensus of structural sacroiliitis. After data collection, we evaluated the association between radiographic images of SpA and SUA concentration. We compared the SUA concentration between the radiographic axial SpA group and non-radiographic axial SpA group, with logistic regression models. Stratified and interaction analyses were also performed to further confirm the consistency of the relationship. Multiple imputation was used to address missing data. Data were analyzed using the statistical packages R (The R Foundation; http://www.r-project.org; version 3.4.3) and EmpowerStats (www.empowerstats.com; X&Y Solutions Inc., Boston, USA). Two hundred and two individuals were included in this study. We divided our subjects into the non-radiographic axial SpA group and radiographic axial SpA group. The percentages of males in the two groups were 57.4% and 79.4% (P = 0.001), respectively. Compared with the non-radiographic axial SpA group, the radiographic axial SpA group was associated with a higher level of disease duration (P < 0.001), SUA (P < 0.001), axial involvement (0.019), BASDAI (P = 0.020), ASDAS-CRP (P = 0.017), and the usage of conventional synthetic DMARDs (P = 0.002). No significant difference was found between the radiographic group and the non-radiographic group with respect to age, gout, peripheral involvement, enthesitis, dactylitis, uveitis, psoriasis, inflammatory bowel disease (IBD), erythrocyte sedimentation rate (ESR), CRP, creatinine, BASFI, and NSAID usage. Univariate analyses and multivariate analyses were performed. In the univariate logistic regression analyses for crude association between clinical variables and axial radiographic SpA, we found that male gender, disease duration (months), axial involvement (any), conventional synthetic DMARDs, BASDAI, ASDAS-CRP, HLA-B27, and SUA were associated with radiographic axial SpA (P < 0.05). In the multivariate regression analyses for associations between elevated SUA concentration (per 10 μmol/L) and radiographic axial SpA, SUA was found to be significantly associated with radiographic axial SpA (odds ratio [OR] = 1.06; 95% confidence interval [CI], 1.03–1.10; P < 0.001). To further analyze this association, we adjusted age, gender and other confounders. We selected gender, IBD and creatinine as confounders based on their associations with a change in effect estimate of more than 10%.[2] Elevated SUA was associated with higher odds of radiographic axial SpA independently after adjusting for confounders. In the fully adjusted model, the odds of having a diagnosis of radiographic axial SpA increased by 7% (adjusted OR = 1.07, 95% CI: [1.03, 1.12], P = 0.0013) per 10 μmol/L SUA increase. We then divided SUA (per 10 μmol/L) into three categories as low tertile (14.30–33.00), middle tertile (33.19–41.66), and high tertile (41.70–76.94). Using SUA as categorical variables, multivariate logistic regression analyses were conducted again to further confirm the association between SUA and radiographic axial SpA. Stratified analyses were performed to further confirm the association of the SUA and radiographic axial SpA. When SUA increased, the odds of diagnosing radiographic axial SpA increased consistently and steadily for almost all the strata. The association was also not statistically significant between radiographic axial SpA and difference of all strata (P-value for interaction >0.05). The interaction analyses revealed that the association between the SUA and radiographic axial SpA did not interact with other factors. Sensitivity analysis was performed for the association of SUA and radiographic axial SpA. After missing data (including seven for disease duration, three for HLA-B27, 16 for ESR, four for CRP, 80 for BASDAI, 84 for ASDAS-CRP, 80 for BASFI) underwent multiple imputation, logistic analysis was performed again on the imputed data. The result showed that the odds of developing radiographic axial SpA increased by 7% (adjusted OR = 1.07, 95% CI: [1.02, 1.12], P = 0.004) per 10 μmol/L SUA increase, which was consistent with that of the preimputation data, a 13% increase (adjusted OR = 1.13, 95% CI: [1.05, 1.22], P = 0.0014) per 10 μmol/L SUA increase. Furthermore, a smooth curve-fitting model was applied to present the OR of radiographic axial SpA which generally increased with rising level of SUA [Figure 1A]. A roughly positive non-linear correlation was observed between SUA and radiographic axial SpA. The smooth curve exhibited a two-stage change with a breakpoint at 26.09 (per 10 μmol/ L). For illustrative purposes, we presented two patients’ CT sacroiliac images for comparison [Figure 1B]. Patient A (CT image A) was a 30-year-old male, HLA-B27 positive, with disease duration of 120 months, and SUA of 489.2 μmol/L. Patient B (CT image B) was a 32-year-old male, HLA-B27 positive, with disease duration of 120 months, and SUA of 361.4 μmol/L. Patient A presented radiographic axial SpA, while patient B presented non-radiographic axial SpA.Figure 1: (A) The smooth curves for association between uric acid and radiographic axial SpA. (B) Two separate patients’ CT sacroiliac images comparison: patient A (CT image A) presented radiographic axial SpA, with sclerosis, erosions, and joint space of ankylosis; patient B (CT image B) presented non-radiographic axial SpA, with only sclerosis. CT: Computed tomography; SpA: Spondyloarthritis.Few population-based studies to date examined the relation between SUA and radiographic axial SpA. Our study reported that elevated SUA concentration was associated with higher odds of developing radiographic axial SpA among patients with SpA. Based on our results, we hypothesized that the potential biologic mechanism was that the high level of SUA led to deposits of MSU which destroyed the structure of sacroiliac joints. Prior studies have confirmed that high SUA increased the burden of MSU deposits.[3] Arthritic study reported that MSU deposits could be detected in different joints of other rheumatic arthritis diseases except gout. It has been ascertained by several axial arthritis surgeries that MSU deposits were found at the axial joints (including sacroiliac joints) when pathological examination was performed.[4] Furthermore, sacroiliitis also facilitated MSU deposition. In the early stages of axSpA, erosive cartilage defects were found in sacroiliac joints.[5] In damaged cartilage, xanthine oxidase (XO) expression, an expressing form of xanthine dehydrogenase, was increased in all cartilage areas. XO contributed to chondrocyte mineralization and pathological calcification.[6] XO also induced the formation of uric acid. When the elevated uric acid was over saturation concentration of approximately 6.8 mg/dl (405 μmol/L), MSU crystals were formed.[7] MSU crystals induced chondrocyte death and cartilage damage.[8] We hypothesize that a possible mechanism for injury would initially involve low levels of cartilage damage, leading to an increase in XO expression. The increased XO expression would then lead to chondrocyte mineralization followed by hyperuricemia and an elevated uric acid concentration, thereby leading to MSU. MSU would then further damage the cartilage of sacroiliac joints. This could also explain the phenomenon that the predictive risk of radiographic SpA increases rapidly based on our study's curve models from the beginning of the curves. The progression of sacroiliitis in axial SpA has been a difficult problem to solve. Despite the usage of TNFi, radiographic progression continued gradually. Tumor necrosis factor pathway was likely not only the major way for the progression. According to the previous report about MSU deposition at sacroiliac joints and our discovery about the relationship between SUA and radiographic axial SpA, we found that SUA was another possible factor for sacroiliitis progression. Our study had several strengths. All the data were measured according to a standard protocol. We considered and adjusted for many potentially confounding factors. Sensitive analyses were also performed to confirm the consistence of outcomes between multiple imputed data and pre-imputation data. This was a retrospective study which extracted data from hospital information system by searching keywords. This searching procedure included all SpA in the information system to avoid selection bias and allowed us to obtain a stronger result. There were also limitations for the study. The cross-sectional study could not define casual effect of SUA and the development of radiographic axial SpA. Furthermore, the sample size of this study was not large enough for stratified analyses. Thus, we need cohort studies with larger samples to further demonstrate the relation between SUA and radiographic axial SpA. In summary, this Chinese population-based study of SpA finds a significant positive relation between SUA and radiographic axial SpA. Further elucidation of how SUA interacts with radiographic axial SpA may provide us with its specific biologic mechanism and a new therapy for radiographic axial SpA in the future. Funding This work was supported by grants from the Science and Technology Program for Basic Research in Shenzhen (Nos. JCYJ20190809095811254, JCYJ20200109140412476). Conflicts of interest None.