关键词: 骨坏死, 组织蛋白酶类, 孟德尔随机化, 多变量, 因果关系, 单核苷酸多态性, 工具变量, 全基因组关联研究, 逆方差加权法, Meta分析

Abstract: BACKGROUND: Osteonecrosis is an orthopedic disease that severely limits joint function, with complex pathogenesis involving multiple risk factors. Cathepsins, as a class of enzymes that play a key role in bone metabolism, are closely related to the proliferation, differentiation of bone cells, and remodeling of the bone matrix. However, previous studies have mostly focused on descriptive analyses, lacking direct evidence of causal relationships. 
OBJECTIVE: To clarify the potential causal relationship between cathepsins and osteonecrosis and to explore their possible mechanisms by analyzing large-scale sample data from the FinnGen database.
METHODS: We obtained osteonecrosis-related data from the FinnGen database, including R9 (a total of 359 399 samples: 1 385 cases and 358 014 controls) and R10 versions (a total of 392 580 samples: 1 543 cases and 391 037 controls). Single nucleotide polymorphisms associated with nine cathepsins (cathepsin B, E, F, G, H, O, S, L2, and Z) were acquired from a previous study (3 301 individuals). Univariate Mendelian randomization, reverse univariate Mendelian randomization, and multivariate Mendelian randomization analyses were conducted using the inverse variance weighted method, MR-Egger method, weighted median method, simple mode method, and weighted mode method. Initially, Mendelian randomization analysis was performed using osteonecrosis data from R9. Additionally, sensitivity analyses were conducted using Cochran’s Q test, MR-Egger intercept, MR-PRESSO global test, and leave-one-out analysis to check for horizontal pleiotropy and heterogeneity. Subsequently, a validation analysis study was carried out on the R10 dataset, and a meta-analysis was conducted to combine the two datasets to explore the joint effect.  
RESULTS AND CONCLUSION: Univariate Mendelian randomization analysis results showed that higher levels of cathepsin B were significantly associated with a reduced risk of osteonecrosis (inverse variance weighted: odds ratio (OR)=0.865, 95% confidence interval (CI): 0.762-0.982, P=0.025), and no reverse causal relationship was found between the nine cathepsins and osteonecrosis (P > 0.05). These associations were validated by meta-analysis. Multivariate analysis, using the nine cathepsins as covariates, revealed a reverse causal relationship between the levels of cathepsin B and the risk of osteonecrosis (inverse variance weighted: OR=0.8710, 95% CI: 0.761-0.997, P=0.045), consistent with the results before adjustment. Sensitivity analyses based on heterogeneity and horizontal pleiotropy suggested that the results were relatively robust. This study suggests that there is a causal relationship between high levels of cathepsin B and the reduced risk of osteonecrosis, and it may serve as a biomarker for osteonecrosis, providing new directions and insights for the diagnosis and treatment of osteonecrosis. Although this study is based on data analysis of European populations, these findings have important implications for Chinese biomedical research, especially in understanding disease mechanisms, developing biomarkers, and formulating treatment strategies. They also encourage similar studies conducted on Chinese populations to explore the impact of racial and genetic background differences on the occurrence of osteonecrosis.

中国组织工程研究杂志出版内容重点：组织构建；骨细胞；软骨细胞；细胞培养；成纤维细胞；血管内皮细胞；骨质疏松；组织工程

Key words: osteonecrosis, cathepsins, Mendelian randomization, multivariate, causal relationship, single nucleotide polymorphisms, instrumental variables, genome-wide association studies, inverse variance weighted method, Meta-analysis