Comparing the effects of CETP in East Asian and European ancestries: a Mendelian randomization study

Abstract

CETP inhibitors are a class of lipid-lowering drugs in development for treatment of coronary heart disease (CHD). Genetic studies in East Asian ancestry have interpreted the lack of CETP signal with low-density lipoprotein cholesterol (LDL-C) and lack of drug target Mendelian randomization (MR) effect on CHD as evidence that CETP inhibitors might not be effective in East Asian participants. Capitalizing on recent increases in sample size of East Asian genetic studies, we conducted a drug target MR analysis, scaled to a standard deviation increase in high-density lipoprotein cholesterol. Despite finding evidence for possible neutral effects of lower CETP levels on LDL-C, systolic blood pressure and pulse pressure in East Asians (interaction p-values < 1.6 × 10^-3), effects on cardiovascular outcomes were similarly protective in both ancestry groups. In conclusion, on-target inhibition of CETP is anticipated to decrease cardiovascular disease in individuals of both European and East Asian ancestries.

Fig. 1. Regional association plots for the CETP locus across LDL-C and HDL-C in European (top) and East Asian (bottom) populations.

The data were sourced from the global lipids genetics consortium (GLGC), with ancestry specific linkage disequilibrium data obtained from the UK biobank. The y-axis shows the -log₁₀(p-values) of the association between each SNP and lipid outcomes. The x-axis shows the chromosomal position (GRCh37). The purple circle shows the European-lead variant rs183130 (16:g.56991363C>T, GRCh37) at the CETP locus identified in the GLGC meta-analysis. The colour coding indicates the linkage disequilibrium with the lead fine-mapped European variant based on the UK Biobank European and East Asian reference population. The blue line shows the alignment of the CETP signals between lipids, as an indicator of colocalization, reported as posterior probability of both lipids sharing the same causal variant at the CETP locus (PP.H4). Coloc was used to estimate the posterior probabilities for all four hypotheses (H1: variants exclusively associated with the trait in European ancestries, H2: variants exclusively associated with the trait in East Asian ancestries, H3: independent variants associating with the trait in both ancestries, H4: the same variant associated with the trait in both ancestries) are: 0.981 (PP.H1), 0.000 (PP.H2), 0.017 (PP.H3), 0.002 (PP.H4) for LDL-C, and 0.000 (PP.H1), 0.000 (PP.H2), 0.027 (PP.H3), 0.974 (PP.H4) for HDL-C. With PP.H0 (no association at all) equal to the remainder after subtracting the PPs from 1. The source data underpinning this figure are available through figshare: 10.5522/04/24559537.v1.

Fig. 2. Mendelian randomisation effect estimates of lower CETP weighted by HDL-C on biomarkers in East Asian and European populations.

The rows represent the plasma biomarker outcomes, with ancestry specific effects presented in the first two columns, and their interaction effect presented in the final column. Cells are annotated by the point estimate per standard deviation increase in HDL-C (in the indicated trait units), when these were significant at an alpha of 0.05, or by a dot otherwise. Multiplicity corrected interaction effects (p-value < 1.6 × 10⁻³) were additionally annotated by a star symbol. The MR results are based on a Rucker selected IVW or Egger estimators, which were evaluated against a two-side alternative hypothesis. Cells are coloured by the direction of effect (dir) times the −log₁₀(p-value), which was truncated to ±8 for display purposes. Apo-A1: apolipoprotein A, Apo-B: apolipoprotein B, Lp[a] lipoprotein a, SBP: systolic blood pressure, DBP: diastolic blood pressure, PP: pulse pressure, CRP: C-reactive protein. Please see Supplementary Data 1–2 for the data underpinning this figure. Source data are provided as a Source Data file.

Fig. 3. Mendelian randomisation effect estimates of lower CETP weighted by HDL-C on cardiovascular outcomes in European and East Asian populations.

Effect estimates are presented as dots with 95% confidence intervals (95%CI) indicated as line segments. Effect estimates are odds ratios (OR) per standard deviation increase in HDL-C. The OR and 95% CI are shown on the right, together with MR estimate p-value. The significance of the interaction between the ancestry specific MR estimates is shown as “Interaction p-value”. The multiplicity corrected interaction test alpha was 1.6 × 10⁻³. The MR results are based on a Rucker selected IVW or Egger estimators, which were evaluated against a two-side alternative hypothesis. No. of cases/Total per ancestry: Coronary heart disease (European: 60,801/184,305 and East Asian: 32,512/146,214), angina (European: 30,025/440,906 and East Asian: 14,007/145,158), peripheral artery disease (European: 7114/457,964 and East Asian: 4112/173,601), any stroke (European: 110,182/1,614,080 and East Asian: 23,345/245,585) and ischaemic stroke (European: 86,668/1,590,566 and East Asian: 22,664/152,022), intracerebral haemorrhage (European: 1935/471,578 and East Asian: 1456/152,022), subarachnoid haemorrhage (European: 5140/77,092 and East Asian: 1203/152,022), heart failure (European: 47,309/977,323 and East Asian: 12,665/245,263), ventricular arrhythmia (European: 1018/327,198 and East Asian: 1673/155,540). Please see Supplementary Data 1–2 for the data underpinning this figure. Source data are provided as a Source Data file.

Fig. 4. Mendelian randomisation effect estimates of lower CETP weighted by HDL-C on non-cardiovascular outcomes in European and East Asian populations.

Effect estimates are presented as dots with 95% confidence intervals (95%CI) indicated as line segments. Effect estimates are odds ratios (OR) per standard deviation increase in HDL-C. The OR and 95% CI are shown on the right, together with MR estimate p-value. The significance of the interaction between the ancestry specific MR estimates is shown as “Interaction p-value”. The multiplicity corrected interaction test alpha was 1.6 × 10⁻³. The MR results are based on Rucker selected IVW or Egger estimators, which were evaluated against a two-side alternative hypothesis. No. of cases/Total per ancestry: Type 2 diabetes (European: 80,154/1,339,889 and East Asian: 45,383/132,032), chronic kidney disease (European: 64,164/180,698 and East Asian: 2117/174,345), pneumonia (European: 16,887/463,412 and East Asian 7423/171,303), COPD: chronic obstructive pulmonary disease (European: 58,559/995,917 and East Asian: 19,044/310,689), asthma (European: 38,369/411,131 and East Asian: 13,015/162,933), glaucoma (European: 15,655/179,925 and East Asian: 8448/168,903), lung cancer (European: 29,266/82,716 and East Asian 4444/178,726), prostate cancer (European: 46,939 /137,966 and East Asian: 5672/ 90,332) and breast cancer (European: 133,384/247,173 and East Asian 6325/79,550. Please see Supplementary Data 1–2 for the data underpinning this figure. Source data are provided as a Source Data file.

Fig. 5. Drug target Mendelian randomisation pathways.

Nodes are presented in bold face, with G representing a genetic variant, $\mathit{P}$ a protein drug target, $\mathit{X}$ a biomarker, $\mathit{D}$ the outcome, and $\mathit{U}$ (potentially unmeasured) common causes of both $\mathit{P}$, $\mathit{X}$, $\mathit{D}$. Labelled paths represent the effect magnitudes between nodes.