Abstract
Steroid-induced osteonecrosis of the femoral head (SONFH) is a major cause of disability among young and middle-aged adults. However, current diagnosis relies primarily on imaging findings and clinical manifestations, as stable and reliable molecular biomarkers for adjunctive diagnosis and risk stratification remain lacking, thereby hindering timely and effective intervention. Aberrant lactate metabolism is thought to contribute to the onset and progression of various inflammatory diseases by reshaping the inflammatory microenvironment and reprogramming immune responses. However, its role and regulatory mechanisms in SONFH remain understudied. In this study, we analyzed transcriptomic data from SONFH patients in the GEO database, integrating differential expression analysis with weighted gene co-expression network analysis (WGCNA) to identify SONFH-associated genes and co-expression modules. Cross-screening with lactate-related genes (LRGs) curated in the MSigDB database yielded a set of LRGs closely associated with SONFH. Unsupervised consensus clustering was then applied to stratify patients into molecular subtypes, and a machine-learning-based diagnostic model was constructed. In parallel, gene set variation analysis (GSVA) and CIBERSORT were used to characterize metabolic states and immune cell infiltration across subtypes, with a focus on LRGs implicated in metabolic reprogramming and immune dysregulation. Finally, bone marrow-derived mesenchymal stem cells (BMSCs) were collected from Sprague-Dawley rats and humans, along with peripheral blood from patients, and in vitro experiments confirmed significant downregulation of BPGM, FBXL4, and RHAG in SONFH, genes closely linked to bone metabolic imbalance and immune microenvironment remodeling. Collectively, these findings systematically elucidate the potential molecular regulatory role of LRGs in SONFH and provide a theoretical basis for its auxiliary diagnosis and the development of targeted therapeutic strategies.