While methylenetetrahydrofolate reductase (MTHFR) polymorphisms are known to regulate homocysteine (Hcy) metabolism, their contribution to the clinical heterogeneity, immune dysregulation, and treatment responses in axial spondyloarthritis (axSpA) remains critically undefined. To address this gap, this study aims to pioneeringly delineate the multifaceted role of MTHFR polymorphisms in axSpA, with a particular focus on elucidating their associations with distinct clinical, immunological, and metabolic phenotypes.

We conducted a cross-sectional analysis of 180 patients with active axSpA, complemented by a retrospective cohort analysis of treatment response in a subset of 50 patients. All patients underwent baseline assessment of radiographic damage (mSASSS), serum Hcy, liver enzymes, and inflammatory markers. CD4⁺ T cell subsets were profiled by flow cytometry, and serum neurotransmitter metabolites by UPLC-MS/MS. Patients were stratified by MTHFR C677T and A1298C variants, with a cumulative risk allele score (0, 1, 2). Dose-response relationships were evaluated using the Jonckheere-Terpstra test for continuous variables and the Cochran-Armitage test for trends, with ANCOVA adjustment for confounders. Patients from this cohort who completed 12 weeks of TNF inhibitor (TNFi) therapy were assessed for response using ΔASDAS-CRP.

C677T risk allele carriers had more severe spinal damage (median mSASSS: 20.50 vs. 13.50; p=0.009) and higher liver enzymes (GGT, p=0.002; ALP, p<0.001) than CC homozygotes. A clear gene-dose effect was observed: higher allele burden progressively correlated with increased Hcy, liver enzymes, and mSASSS ((all p for trend ≤0.05), persisting after covariate adjustment. The 677T allele was linked to a pro-inflammatory shift, evidenced by elevated Th17/Th2 and Th17/Treg ratios (all p<0.05). In the treatment cohort, TNFi response was stratified by genetic burden. Furthermore, response to TNFi therapy was stratified by genetic burden. While CRP levels decreased significantly following TNFi treatment in patients carrying 0 or 1 risk allele (p=0.001 for each), the reduction did not reach statistical significance in those with 2 risk alleles (p=0.188).

In axSpA, MTHFR risk alleles are dose-dependently associated with a phenotype of hyperhomocysteinemia, elevated liver enzymes, Th17 polarization, and severe spinal damage. A higher cumulative genetic burden is further linked to attenuated TNFi response. MTHFR genotyping may serve as a potential biomarker for risk stratification and personalized treatment planning.