Methotrexate (MTX) has long been the anchor drug for rheumatoid arthritis (RA). However, clinical concerns persist that MTX may induce or exacerbate interstitial lung disease (ILD), often forcing clinicians to discontinue therapy in patients with respiratory symptoms. This study aimed to provide robust evidence for this clinical controversy by integrating two-sample Mendelian randomization (MR) and network pharmacology.

A two-sample MR design was employed using summary data from large-scale genome-wide association studies. First, genetic variants significantly associated with clinical MTX use from the UK Biobank database were utilized as instrumental variables (P < 5×10⁻⁸, F > 10). Outcome data included ILD and idiopathic pulmonary fibrosis (IPF). The primary causal estimation was performed using the inverse-variance weighted (IVW) method, supplemented by sensitivity analyses including MR-Egger and weighted median to assess pleiotropy. To address confounding, multivariable Mendelian randomization (MVMR) was conducted, adjusting for genetic predisposition to autoimmune diseases including RA, Sjögren’s syndrome, myositis, and smoking status. In addition, network pharmacology was used to analyze molecular overlap between MTX targets and pulmonary fibrosis-related genes, along with GO and KEGG functional enrichment analyses to explore potential drug-disease relationships at the molecular mechanism level.

Univariate MR demonstrated no consistent causal association between genetic predisposition to MTX use and increased risk of IPF or ILD in the European population (P > 0.05). Although weak signals were observed in individual datasets, these were not robust after validation by sensitivity analyses and replication in independent cohorts.MVMR showed that the direct effect of MTX on pulmonary fibrosis was completely attenuated and no longer statistically significant after adjusting for autoimmune diseases and smoking. In contrast, the analysis confirmed RA (OR=1.17, p=0.017), myositis (OR=1.12, p=0.018), and smoking history (OR=6.94, p=0.044) as significant independent risk factors for ILD. Network pharmacology identified 34 shared targets between MTX and pulmonary fibrosis (e.g., TP53, NFKB1, MAPK1). However, functional enrichment revealed that these targets were mainly involved in general cellular stress responses and viral infection pathways, rather than specific pro-fibrotic toxic signaling cascades.

This study provides strong genetic evidence that does not support the traditional assumption that MTX use causally increases the risk of ILD.