FTO rs7195994 Is associated with TNF inhibitor response in lean rheumatoid arthritis patients: A BMI-stratified pharmacogenetic analysis

Tumor necrosis factor inhibitors (TNFi) have transformed the management of rheumatoid arthritis (RA); however, up to 40% of patients fail to achieve an adequate clinical response. Current clinical predictors remain insufficient, highlighting the need for pharmacogenetic biomarkers to guide biologic therapy selection. In this study, we investigated genetic variants associated with TNFi treatment response, with a particular focus on body mass index (BMI)-dependent effects, using a large real-world cohort. A total of 519 patients with RA were identified from the Taiwan Precision Medicine Initiative (TPMI), an electronic health record-linked biobank. Eligible patients had received TNFi therapy for at least 6 months and had available genotyping data. Ninety-seven candidate single nucleotide polymorphisms (SNPs) previously reported to be associated with TNFi response were identified through a systematic literature search. Five variants located in immune-metabolic genes (FTO, ZNF618, RANK, CD84, and LOC105375523) were further analyzed using univariable and multivariable logistic regression models. Subgroup analyses stratified by BMI were performed to explore potential effect modification. Three variants-FTO rs7195994, ZNF618 rs16911006, and LOC105375523 rs834811-were significantly associated with TNFi response in initial analyses. In multivariable models, only FTO rs7195994 remained an independent predictor of non-response (odds ratio OR 0.44, 95% confidence interval CI 0.22-0.87; p = 0.019). Among patients with BMI < 27 kg/m², carriers of the rs7195994 risk allele had a 49% lower likelihood of achieving treatment response (OR 0.51, 95% CI 0.28-0.92; p = 0.0249), whereas no significant association was observed in patients with higher BMI. These findings identify FTO rs7195994 as a novel, BMI-modulated pharmacogenetic marker of TNFi non-response in RA. Incorporating BMI-stratified genetic profiling into clinical decision-making may facilitate early identification of patients unlikely to benefit from TNFi therapy, thereby supporting precision treatment strategies. Further validation in multiethnic populations and functional studies is warranted.