Abstract: BACKGROUND: Osteonecrosis is a common and refractory disease in clinical practice, which brings great distress to the patients' lives and also imposes a heavy burden on the medical system. Some observational studies have shown that lipid-lowering drugs may have a certain promoting effect on the prognosis of osteonecrosis. However, the causal relationship between the specific targets of lipid-lowering drugs and osteonecrosis remains unknown.
OBJECTIVE: To explore the causal relationships between low-density lipoprotein, lipid-lowering drugs [3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR) inhibitors, PCSK9 inhibitors, and NPC1L1 inhibitors] and osteonecrosis through Mendelian randomization and Meta-analysis.
METHODS: (1) The study first downloaded the HMGCR gene (Chromosome 5: 74632154-74657929), PCSK9 gene (Chromosome 1: 55505221-55530525), and NPC1L1 gene (Chromosome 7: 44552134-44580914) from the Global Lipids Genetics Consortium (GLGC) database as instrumental variables for lipid-lowering drug exposure. Osteonecrosis-related data were obtained from the FinnGen database, including Version R9 (with a total sample size of 359 399: 1 385 cases and 358 014 controls), Version R10 (with a total sample size of 392 580: 1 543 cases and 391 037 controls), and Version R11 (with a total sample size of 431 369: 1 543 cases and 429 826 controls). Meanwhile, low-density lipoprotein was selected as the biomarker for this study. (2) The two-sample Mendelian randomization analysis method was used to evaluate the causal relationship between lipid-lowering drug exposure and osteonecrosis. And LDlink was used to remove confounding factors that might affect the pathogenesis of osteonecrosis. (3) Four methods, namely MR-Egger, weighted median, inverse-variance weighting (IVW), and weighted mode, were employed to detect the causal relationship between lipid-lowering drug exposure and osteonecrosis. The IVW method was used as the primary analysis method. (4) To comprehensively and comprehensively evaluate the association between lipid-lowering drugs and osteonecrosis, a meta-analysis was further conducted on the odds ratio (OR) values obtained from the IVW method. (5) Finally, to ensure the robustness and reliability of the research results, this study used the Q test to evaluate heterogeneity, MR-Egger regression to evaluate horizontal pleiotropy, and Bayesian colocalization analysis for positive results. (6) The study also selected coronary heart disease as a positive control. By conducting Mendelian randomization analysis with it as the outcome factor, the reliability of the instrumental variables was determined.
RESULTS AND CONCLUSION: (1) The results of the Mendelian randomization analysis showed that the relationship between HMGCR inhibitors and osteonecrosis exhibited diverse characteristics across different versions of osteonecrosis data. In the R9 version data, HMGCR inhibitors were significantly effective in reducing the risk of osteonecrosis (IVW: OR=0.24, 95% confidence interval [CI]: 0.08-0.71; P=0.009). Similar conclusions were drawn from the R10 version data (IVW: OR=0.29, 95% CI: 0.10-0.86; P=0.025). However, in the R11 version data, there was no obvious causal relationship between them (P=0.09 > 0.05). (2) Neither NPC1L1 inhibitors, PCSK9 inhibitors, nor low-density lipoprotein showed an obvious causal relationship with osteonecrosis in the R9, R10, and R11 version data of osteonecrosis. The meta-analysis of the IVW results for HMGCR inhibitors further confirmed that they could significantly reduce the risk of osteonecrosis (OR=0.3, 95% CI: 0.16-0.56). (3) The sensitivity analysis did not find statistical evidence of heterogeneity or genetic confounding bias. The Bayesian colocalization analysis showed that the posterior probability of H4/(H4+H3) exceeded 90%, indicating that the causal relationship between HMGCR inhibitors and osteonecrosis was not accidental. (4) This study, through Mendelian randomization and meta-analysis, confirmed that there was a significant causal association between HMGCR inhibitors and the reduction of the risk of osteonecrosis. This suggests that existing HMGCR inhibitors are expected to become key targets for the prevention and treatment of osteonecrosis, bringing new treatment options for many patients with osteonecrosis. However, further in-depth clinical studies are still needed for verification.

Key words: osteonecrosis, lipid-lowering drugs, Mendelian randomization, meta-analysis, HMGCR inhibitors, PCSK9 inhibitors, NPC1L1 inhibitors