Evaluating the Cardiovascular Impact of Genetically Proxied PCSK9 and HMGCR Inhibition in East Asian and European Populations: A Drug-Target Mendelian Randomization Study

Daniel B Rosoff, Andrew A Bell, Lucas A Mavromatis, Ali Hamandi1, Lauren Park, Jeesun Jung, Josephin Wagner, George Davey Smith, Falk W Lohoff*, et al

*Corresponding author for this work

Research output: Contribution to journalArticle (Academic Journal)peer-review

26 Downloads (Pure)


As new proprotein convertase subtilisin/kexin 9 inhibitors (PCSK9is) become approved for clinical use, it is important to evaluate their long-term impact on cardiovascular health in diverse populations representative of patients in clinical settings, including across ancestries and sexes, where both differences in cardiovascular disease (CVD) risk and drug responses have been suggested.1,2 However, as PCSK9i approval is recent, whether there are differences in long-term cardiovascular efficacy of PCSK9i and 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibition across ancestries and between men and women remains unknown. More broadly, there exists a pressing need to both improve race/ancestry data in genetics-based studies and increase diversity in randomized controlled trials across all disciplines.3

We perform drug-target Mendelian randomization (MR) analyses4 comparing the impact of inhibiting PCSK9 (proprotein convertase subtilisin/kexin 9) and HMGCR (3-hydroxy-3-methylglutaryl coenzyme A reductase) on 5 CVD outcomes publicly available in East Asian (EAS) and European cohorts (coronary artery disease [CAD], peripheral artery disease [PAD], atrial fibrillation, arrhythmia, and heart failure [HF]; Figure [A]). We also obtained corresponding sex-stratified CVD data to compare PCSK9 and HMGCR efficacy between men and women in the EAS and European populations. This study uses publicly available data, and ethical approval was obtained in the original studies. Study code and harmonized data are available from authors upon request. We extracted genetic variants within or near the PCSK9 and HMGCR loci (±100 kilobases) strongly associated (P<5×10−8, clumped at r2<0.2 [±250 kilobases]) with low-density lipoprotein cholesterol levels in genome-wide association study data derived from the Global Lipids Genetics Consortium EAS (N≤146 462) and European (N≤1 320 016) cohorts. EAS PCSK9 and HMGCR instruments had average F-statistics of 78.9 and 280.1, respectively. European PCSK9 and HMGCR instruments had average F-statistics of 81.7 and 210.3. We used the inverse-variance weighted incorporating correlation matrices between the instrument variants (matrices generated using the relevant 1000G ancestry reference panels) along with MR Egger and MR maximum likelihood as sensitivity analyses. Drug-target estimates were aligned with the physiological impact of PCSK9is and statins corresponding to an SD lowering in low-density lipoprotein cholesterol. For inverse-variance weighted estimates with P values surpassing Bonferroni correction (0.005 [5 end points×2 drugs per population]), we performed post hoc hypothesis testing evaluating differences in the estimates. Colocalization was also performed to assess evidence of a shared causal variant between low-density lipoprotein cholesterol and the CVDs in these loci (evidence of colocalization defined as PP.H4>0.5).

Original languageEnglish
Article numbere004224
Pages (from-to)59-62
Number of pages4
Issue number1
Early online date23 Jan 2024
Publication statusPublished - 1 Feb 2024

Structured keywords

  • Bristol Population Health Science Institute


Dive into the research topics of 'Evaluating the Cardiovascular Impact of Genetically Proxied PCSK9 and HMGCR Inhibition in East Asian and European Populations: A Drug-Target Mendelian Randomization Study'. Together they form a unique fingerprint.

Cite this