No genetic evidence for an association of apolipoprotein A-I with cardiovascular outcomes at different LDL cholesterol levels: drug-target Mendelian randomization analyses

Background and aims Exploratory post-hoc analyses of the AEGIS-II trial suggested that the efficacy of CSL112 (apolipoprotein A-I infusion) may be greater in patients with high baseline LDL-C. This genetic study aimed to assess whether LDL-C levels modify the effect of apoA-I on cardiovascular outcomes. Methods We conducted drug-target Mendelian randomization analyses using data from 339,210 UK Biobank participants. Genetic proxies for apoA-I elevation ( cis -acting variants within APOA1 gene) were used to mimic the effect of CSL112 therapy. The primary outcome was a composite of ischemic heart disease, myocardial infarction, and stroke; secondary outcomes were the individual components. To mitigate collider bias, we stratified by residual LDL-C, derived from a model adjusting LDL-C for the genetic instrument, age, sex, genotyping array and top 40 genetic principal components. Sensitivity analyses included alternative genetic instruments and formal tests for effect modification. Results Genetically proxied apoA-I elevation was associated with elevated circulating apoA-I and HDL-C, but there was no concomitant association with any cardiovascular outcome. Null findings were consistent across strata of residual LDL-C. Sensitivity analyses with alternative instrument sets and formal interaction tests uniformly confirmed the absence of any association or effect modification. Conclusions Genetically proxied lifelong elevation of apoA-I concentration was not associated with reduced cardiovascular risk, even among individuals with high residual LDL-C.