Premium
Predictive clinical‐genetic model of long‐term non‐response to tumor necrosis factor‐alpha inhibitor therapy in spondyloarthritis
Author(s) -
Polo y La Borda Jessica,
Campos José,
Sanz Jesús,
Andréu José Luis,
Mulero Juan,
Sánchez Alejandra
Publication year - 2019
Publication title -
international journal of rheumatic diseases
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.795
H-Index - 41
eISSN - 1756-185X
pISSN - 1756-1841
DOI - 10.1111/1756-185x.13607
Subject(s) - basdai , medicine , basfi , ankylosing spondylitis , rheumatoid arthritis , hazard ratio , etanercept , infliximab , proportional hazards model , golimumab , univariate analysis , oncology , tumor necrosis factor alpha , multivariate analysis , psoriatic arthritis , confidence interval
Aim Tumor necrosis factor inhibitors (TNFi) are effective in controlling disease activity in spondyloarthritis (SpA). However, in a proportion of patients these treatments are ineffective or lead to adverse events. Recently, alternative therapies, such as interleukin (IL)‐17 or IL‐23 inhibitors, have emerged in the treatment of these pathologies. This study aimed to determine clinical and genetic predictors of non‐response to TNFi treatment in 118 spondyloarthritis patients diagnosed according to Assessment in SpondyloArthritis International Society (ASAS) criteria. Method From the literature, 41 single nucleotide polymorphisms (SNPs) were selected that had previously been associated with TNFi treatment response in spondyloarthropathies, rheumatoid arthritis and psoriasis. A clinical non‐response was defined as a decrease of <50% of initial Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) in axial involvement, or a reduction of less than 1.2 of initial Disease Activity Score of 28 joints—C‐reactive protein (DAS28‐CRP) in patients with only peripheral involvement. Univariate and multivariate hazard ratios (HR) were determined using Cox proportional hazard models to analyze the potential prognostic factors affecting non‐response to TNFi treatment. Results The clinical factors that significantly increased the non‐response rate were: global visual analog scale (VAS), CRP, BASDAI, Bath Ankylosing Spondylitis Functional Index (BASFI), and the number of TNFi used. Only rs11591741 SNP showed an association with non‐response. In the multivariate analysis, females had a non‐response rate 4.46 times higher than males; each one‐point increase in the BASFI index increased the non‐response rate by 75%, and being a genotype GG vs GC or CC carrier was associated with an almost 4 times greater non‐response rate. Conclusion We developed a clinical‐genetic model to identify SpA patients with a long‐term non‐response to TNFi therapy.