Genomic methods to reduce cardiovascular risk

Abstract

This provides a method to reduce cardiovascular risk in patients with a genetic profile associated with a response to proprotein converter subtilisin / kexin type 9 (PCSK9) inhibitor therapy. [Solution] A method for treating a patient at risk of major cardiovascular adverse events (MACE) includes: determining the patient's polygenetic risk score (CAD-PRS), which includes a weighted sum of multiple gene variants associated with coronary artery disease; identifying the patient as being at high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and administering a PCSK9 inhibitor to the patient if the patient has been identified as being at high risk of MACE.

Description

[Technical Field]

[0001] This disclosure relates to the field of therapeutic treatments for diseases and disorders associated with elevated lipid and lipoprotein levels. More specifically, this disclosure relates to methods for enhancing the efficacy of PCSK9 inhibitor therapy in patients at high cardiovascular risk by identifying patients who are likely to respond to PCSK9 inhibitors. [Background technology]

[0002] Despite modern treatments including rapid coronary revascularization, dual antiplatelet therapy, and intensive statin therapy, major cardiovascular adverse events (MACE) continue to occur frequently in patients with a history of MACE. Registry data show a cardiovascular mortality rate of as high as 13% over 5 years, with the vast majority occurring after the initial discharge. Patients with recent MACE are at very high risk of suffering from short-term recurrent MACE. Approximately 10% of patients with a history of MACE experience cardiovascular death, recurrent myocardial infarction, or stroke within one year.

[0003] PCSK9 is a serine protease involved in regulating the levels of low-density lipoprotein receptor (LDLR) protein. In vitro experiments have shown that adding PCSK9 to HepG2 cells reduces LDLR levels on the cell surface. Mouse experiments have shown that increasing PCSK9 protein levels decreases LDLR protein levels in the liver, while PCSK9 knockout mice have high levels of LDLR in the liver. Furthermore, various human PCSK9 mutations have been identified that result in elevated or decreased plasma LDL levels. PCSK9 has been shown to directly interact with LDLR protein, undergo endocytosis together with LDLR, and co-immunofluorescence with LDLR throughout the endosomal pathway. Degradation of LDLR by PCSK9 has not been observed, and the mechanism by which it reduces extracellular LDLR protein levels remains unclear.

[0004] Following the establishment of the association between PCSK9 and cholesterol metabolism, it was immediately discovered that selected mutations in the PCSK9 gene cause autosomal dominant hypercholesterolemia, suggesting that the mutations confer a gain of function by enhancing the normal activity of PCSK9. Conversely, loss-of-function PCSK9 mutations and inhibition of PCSK9 function have been shown to significantly reduce LDL levels and the frequency of MACE.

[0005] PCSK9 inhibition reduces the risk of MACE in both primary and secondary intervention settings, but not all patients respond equally well to PCSK9 inhibitor treatment. The etiology of cardiovascular disease is complex and may be influenced by genetics, environment, and various additional risk factors including dyslipidemia, age, gender, hypertension, diabetes, obesity, and smoking. Genome-wide association studies (GWAS) have identified gene variants widely associated with coronary artery disease, but genomic data needs to be utilized to identify patients who are particularly likely to benefit from PCSK9 inhibitor therapy for the purpose of preventing or reducing the likelihood of MACE. SUMMARY OF THE INVENTION

[0006] The present disclosure provides a method for treating patients at risk of MACE, comprising determining a patient's coronary artery disease polygenic risk score (CAD-PRS), which includes the weighted sum of multiple gene variants associated with coronary artery disease; identifying a patient as having a high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and administering a PCSK9 inhibitor to the patient if the patient is identified as having a high risk of MACE.

[0007] The disclosure also provides a method for lowering serum LDL levels in patients at high risk of MACE, comprising: determining a patient's CAD-PRS, which includes a weighted sum of multiple gene variants associated with coronary artery disease; identifying a patient as being at high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and, if the patient is identified as being at high risk of MACE, targeting the patient with a PCSK9 inhibitor in a dose effective in lowering the patient's serum LDL level.

[0008] The disclosure also provides a method for lowering serum LDL levels in patients at high risk of MACE, comprising: determining a patient's CAD-PRS, which includes a weighted sum of multiple gene variants associated with coronary artery disease; identifying a patient as being at high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and, if the patient is identified as being at high risk of MACE, targeting the patient with a PCSK9 inhibitor in a dose effective in lowering the patient's serum LDL level.

[0009] The disclosure also provides a method for screening candidate subjects for inclusion in a clinical trial for the treatment of a cardiovascular condition, the method comprising: determining the CAD-PRS of a candidate subject, wherein the CAD-PRS comprises a weighted sum of multiple gene variants associated with coronary artery disease; including the candidate subject in the clinical trial if the candidate subject has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; or excluding the candidate subject from the clinical trial if the candidate subject has a CAD-PRS lower than a threshold CAD-PRS determined from a reference population.

[0010] These and other purposes and features of the present disclosure will be better understood and recognized from the following detailed description of one embodiment thereof, which has been selected for illustrative purposes and is shown in the accompanying drawings. [Brief explanation of the drawing]

[0011] [Figure 1-1] A table is provided listing the demographic and baseline characteristics of patients in pharmacological genomics analysis, along with comparisons of high-risk and low-risk gene groups, and generalizability to the ODYSSEY OUTCOMES trial. [Figure 1-2] Same as above. [Figure 1-3] Same as above. [Figure 1-4] Same as above. [Figure 2] The incidence of MACE and secondary endpoints in the placebo group in the low-gene risk group (polygene risk score (PRS) ≤ 90th percentile) and the high-gene risk group (PRS > 90th percentile) is shown. The figures represent the overall incidence of MACE (composite of death from coronary heart disease, non-fatal myocardial infarction, fatal or non-fatal ischemic attack, or unstable angina requiring hospitalization) and primary secondary endpoints in patients with all ancestry, stratified by genetic risk. The numbers at the bottom of each panel represent the number of patients in each group, and the numbers within each bar represent the percentage of MACE in each group. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting and congestive heart failure. [Figure 3-1]The incidence of MACE in the placebo group in low-gene risk groups (PRS ≤ 90th percentile) and high-gene risk groups (PRS > 90th percentile), stratified by baseline risk factors, is shown. The figures represent the overall incidence of MACE in patients with all ancestry, stratified by genetic risk for LDL-C at baseline (<100 mg / dL or ≥100 mg / dL) (Panel A); Framingham relapse risk score (<median or ≥median) (Panel B); and Lp(a) at baseline (<50 mg / dL or ≥50 mg / dL) (Panel C). The numbers at the bottom of each panel represent the number of patients in each group, and the numbers within each bar represent the percentage of MACE in each group. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting; and congestive heart failure. [Figure 3-2] Same as above. [Figure 3-3] Same as above. [Figure 4-1] The cumulative incidence of MACE in the low-gene risk group (Panel A; PRS ≤ 90th percentile) and the high-gene risk group (Panel B; PRS > 90th percentile) is shown. The figures shown are the cumulative incidence of MACE in patients with all ancestry, stratified by genetic risk. Hazard ratios and p-values ​​were calculated from Cox proportional hazards models adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting and congestive heart failure. In addition to genetic risk stratification analysis, Cox models were performed including treatment group, genetic risk (high / low), treatment interaction by genetic risk, and the above covariates. The p-value for treatment interaction by genetic risk was 0.040. [Figure 4-2] Same as above. [Figure 5-1]A table is provided listing the primary and secondary endpoints across low-risk and high-risk genetic risk groups. [Figure 5-2] Same as above. [Figure 5-3] Same as above. [Figure 6-1] This panel shows the incidence of MACE stratified by baseline genetic risk and LDL cholesterol levels. The figures represent the overall incidence of events in patients across all ancestry, stratified by baseline genetic risk and / or LDL-C. Panel A is stratified by genetic risk (high genetic risk is PRS > 90th percentile, and low genetic risk is PRS ≤ 90th percentile). Panel B is stratified by baseline LDL-C (LDL-C ≥ 100 mg / dL and LDL-C < 100 mg / dL). Panel C is stratified by both baseline genetic risk and LDL-C. The numbers at the bottom of each panel represent the number of patients in each group, and the numbers within each bar represent the percentage of MACE in each group. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting; and congestive heart failure. [Figure 6-2] Same as above. [Figure 6-3] Same as above. [Figure 7-1]The cumulative incidence of MACE, stratified by baseline genetic risk and LDL cholesterol levels, is shown. The figures represent the cumulative incidence of all ancestral patients with high genetic risk (PRS > 90th percentile; panels A and B) and low genetic risk (PRS ≤ 90th percentile; panels C and D), further stratified by baseline LDL-C. Patients with LDL-C < 100 mg / dL and high genetic risk are shown in panel A, and patients with LDL-C ≥ 100 mg / dL and high genetic risk are shown in panel B. Similarly, patients with LDL-C < 100 mg / dL and low genetic risk are shown in panel C, and patients with LDL-C ≥ 100 mg / dL and low genetic risk are shown in panel D. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting; and congestive heart failure. The LDL-C interaction of treatment based on baseline genetic risk was p > 0.05. [Figure 7-2] Same as above. [Figure 7-3] Same as above. [Figure 7-4] Same as above. [Figure 8-1] A table is provided listing additional demographic and baseline characteristics of patients in pharmacological genomics analysis. [Figure 8-2] Same as above. [Figure 8-3] Same as above. [Figure 8-4] Same as above. [Figure 9-1]This report presents candidate SNPs (27–57), pruning and thresholding (P&T), and LDPred results on the UK Biobank (UKB) test dataset. Results are presented for a composite endpoint of myocardial infarction, angina, or ischemic attack. Panel A shows the area under the curve (AUC) and odds ratio per SD for each candidate SNP list or set of algorithmic adjustment parameters. Panel B shows the number of markers used to generate the genetic risk score for each adjustment parameter. [Figure 9-2] Same as above. [Figure 10-1] This panel presents candidate SNPs (27-57), pruning and thresholding (P&T), and LDPred results from the DiscovEHR test dataset. Results are presented for a composite endpoint of myocardial infarction, angina, or ischemic attack. Panel A shows the AUC and odds ratio per SD for each candidate SNP list or set of algorithmic adjustment parameters. Panel B shows the number of markers used to generate the genetic risk score for each adjustment parameter. [Figure 10-2] Same as above. [Figure 11-1] The results for LDPred (ρ = 0.001) in the UKB and DiscovEHR test datasets are shown. Results are presented for a composite endpoint of myocardial infarction, angina, or ischemic attack. Panel A shows the proportion of UKB participants with myocardial infarction, angina, or ischemic attack, divided into the 2.5% percentile of their genetic risk score. Panel B shows this proportion for DiscovEHR participants. [Figure 11-2] Same as above. [Figure 12] The table below lists the incidence of MACE for each ancestral group. [Figure 13-1]This plot shows treatment-stratified deciles of MACE, including endpoints such as death from coronary heart disease, non-fatal myocardial infarction, fatal or non-fatal ischemic attack, or unstable angina requiring hospitalization. Panel A shows the event proportions by genetic risk score decile in the alirocumab group, while Panel B shows the risk by decile in the placebo group. The mean PRS Z score for each decile is shown to the right of the decile. The gray dashed line represents the overall event proportion for each group. [Figure 13-2] Same as above. [Figure 14-1] Secondary endpoints—plots of treatment-stratified deciles for any cardiovascular event—are shown. These endpoints include cardiovascular death, non-fatal myocardial infarction, or unstable angina requiring hospitalization, coronary revascularization due to ischemia, or non-fatal ischemic attack. Panel A shows the event proportions by genetic risk score decile in the alirocumab group, while Panel B shows the risk by decile in the placebo group. The mean PRS Z score for each decile is shown to the right of the decile. The gray dashed line represents the overall event proportion for each group. [Figure 14-2] Same as above. [Figure 15-1] Secondary endpoints—plots of treatment-stratified deciles for any coronary heart disease event—are shown. These endpoints include death due to coronary heart disease, non-fatal myocardial infarction, unstable angina requiring hospitalization, and coronary revascularization due to ischemia. Panel A shows the event proportions by genetic risk score decile in the alirocumab group, while Panel B shows the risk by decile in the placebo group. The mean PRS Z score for each decile is shown to the right of the decile. The gray dashed line represents the overall event proportion for each group. [Figure 15-2] Same as above. [Figure 16-1]This plot shows the treatment-stratified deciles for secondary endpoints: death from any cause, non-fatal myocardial infarction, or ischemic attack. Panel A shows the event proportions by genetic risk score decile in the alirocumab group, while Panel B shows the risk by decile in the placebo group. The mean PRS Z score for each decile is shown to the right of the decile. The gray dashed line represents the overall event proportion for each group. [Figure 16-2] Same as above. [Figure 17-1] This panel shows plots of treatment-stratified deciles for secondary endpoints of major coronary heart disease events. These endpoints include death and non-fatal myocardial infarction due to coronary heart disease. Panel A shows the proportion of events by genetic risk score decile in the alirocumab group, while Panel B shows the risk by decile in the placebo group. [Figure 17-2] Same as above. [Figure 18-1] This plot shows the treatment-stratified deciles for secondary endpoints of coronary artery revascularization procedures due to ischemia. Panel A shows the event ratios by genetic risk score decile in the alirocumab group, while Panel B shows the risk by decile in the placebo group. The mean PRS Z score for each decile is shown to the right of the decile. The gray dashed line represents the overall event ratio for each group. [Figure 18-2] Same as above. [Figure 19]The incidence of MACE in the placebo group for low-gene risk (PRS ≤ 90th percentile) and high-gene risk (PRS > 90th percentile) groups, stratified by very high-risk (VHR) groups, is shown. The overall incidence of MACE in all ancestral patients, stratified by VHR category, is also shown. VHR categories follow the definitions described in doi:10.1161 / CIRCULATIONAHA.119.042551. VHR* (Multiple Prior Major ASCVD Events) includes patients with one or more prior ischemic events preceding a qualifying index ACS event, including ischemic attack, myocardial infarction, or peripheral artery disease. VHR* (Major Previous ASCVD Event + Multiple High-Risk Conditions) includes patients with one major ASCVD event (eligible index ACS event) and two or more high-risk conditions (diabetes mellitus, current smoking, age 65 or older, history of hypertension, baseline eGFR ≥ 15 to < 60 mL·1 min·1.73 m-2, congestive heart failure, revascularization prior to index ACS, or LDL-C ≥ 100 mg / dL using both statins and ezetimibe). VHR* is a combination of both categories, while non-VHR includes patients without these risk factors. The numbers at the bottom of each panel represent the number of patients in each group, and the numbers in each bar represent the proportion of MACE in each group. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry. Since the composite VHR* risk group includes multiple risk factors, covariate adjustments for additional risk factors are not included in this model. [Figure 20-1] UKB: Median Lp(a) nmol / L excluding / including the LPA gene region. The results show the median Lp(a) as a percentile, excluding and including the LPA gene region (+ / -1MB). Panel A displays the PRS for the entire genome excluding the LPA gene region (+ / -1MB), while Panel B displays the score for the entire genome. [Figure 20-2] Same as above. [Figure 21-1]ODYSSEY: Shows median Lp(a) mg / dL (Q1~Q3) with / without the LPA gene region. The results show the median Lp(a) as percentiles, with and without the LPA gene region (+ / -1MB) in the score. Panel A shows the PRS for the entire genome with the LPA gene region (+ / -1MB) excluded, and Panel B shows the score for the entire genome. [Figure 21-2] Same as above. [Figure 22-1] UKB: Shows composite endpoints for myocardial infarction, angina pectoris, or ischemic attack, excluding / including the LPA gene region. Results are shown for composite endpoints for myocardial infarction, angina pectoris, or ischemic attack, excluding and including the LPA gene region (+ / -1MB) in the score. Panel A shows the PRS for the entire genome excluding the LPA gene region (+ / -1MB), and Panel B shows the score for the entire genome. [Figure 22-2] Same as above. [Figure 23-1] This shows the incidence of MACE in the ODYSSEY placebo group with / without the LPA gene region. Results are shown for MACE (a composite endpoint including death from coronary heart disease, non-fatal myocardial infarction, fatal or non-fatal ischemic attack, or unstable angina requiring hospitalization) with and without the LPA gene region (+ / -1MB) in the score. Panel A shows the whole genome PRS with the LPA gene region (+ / -1MB) excluded, and Panel B shows the whole genome score. [Figure 23-2] Same as above. [Figure 23-3] Same as above. [Figure 23-4] Same as above. [Figure 24-1]This report shows the cumulative incidence of MACE in patients of European ancestry in the low-gene risk group (PRS ≤ 90th percentile; Panel A) and the high-gene risk group (PRS > 90th percentile; Panel B). The figures presented are the cumulative incidence of MACE (a composite of death from coronary heart disease, non-fatal myocardial infarction, fatal or non-fatal ischemic attack, or unstable angina requiring hospitalization) in patients of European ancestry, stratified by genetic risk. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, percutaneous coronary intervention, or coronary artery bypass grafting for index acute coronary syndrome, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting and congestive heart failure. In addition to stratified analysis of genetic risk, a Cox model was also performed including treatment group, high / low genetic risk, treatment interaction due to genetic risk, and the above covariates. The p-value for genetic risk due to treatment group interaction was 0.113. [Figure 24-2] Same as above. [Figure 25-1] The table below lists the median changes in lipids and related proteins from baseline at 4 months. [Figure 25-2] Same as above. [Figure 26-1]The cumulative incidence of MACE, stratified by baseline Lp(a) level and genetic risk, is shown. The cumulative incidence is shown for all ancestral patients further stratified by baseline Lp(a) into high genetic risk (PRS > 90th percentile; panels A and B) and low genetic risk (PRS ≤ 90th percentile; panels C and D). Patients with Lp(a) ≥ 50 mg / dL and high genetic risk are shown in panel A, and patients with Lp(a) ≥ 50 mg / dL and high genetic risk are shown in panel B. Similarly, patients with Lp(a) < 50 mg / dL and low genetic risk are shown in panel C, and patients with Lp(a) ≥ 50 mg / dL and low genetic risk are shown in panel D. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, percutaneous coronary intervention or coronary artery bypass grafting for indicative acute coronary syndrome; and the following medical features prior to indicative ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting and congestive heart failure. [Figure 26-2] Same as above. [Figure 26-3] Same as above. [Figure 26-4] Same as above. [Figure 27-1]This panel shows the incidence of MACE stratified by genetic risk and baseline Lp(a) levels. The figures represent the proportion of events in patients of all ancestry, stratified by genetic risk and / or baseline LDL-C. Panel A is stratified by genetic risk (high genetic risk is PRS > 90th percentile; low genetic risk is PRS ≤ 90th percentile). Panel B is stratified by baseline Lp(a) (Lp(a) ≥ 50 mg / dL and Lp(a) < 50 mg / dL). Panel C is stratified by both genetic risk and baseline Lp(a). The numbers at the bottom of each panel represent the number of patients in each group, and the numbers within each bar represent the proportion of MACE in each group. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting; and congestive heart failure. [Figure 27-2] Same as above. [Figure 27-3] Same as above. [Figure 28-1] This panel shows MACE stratified by genetic risk considering VHR categories and baseline Lp(a). Panel A is stratified by genetic risk, where high genetic risk is PRS > 90th percentile and low genetic risk is PRS ≤ 90th percentile. Panel B is stratified by baseline Lp(a) (Lp(a) ≥ 50 mg / dL and Lp(a) < 50 mg / dL). Panel C is stratified by both genetic risk and baseline Lp(a). The numbers at the bottom of each panel represent the number of patients in each group, and the numbers in each bar represent the percentage of MACE in each group. Hazard ratios and p-values ​​were calculated from a Cox proportional hazards model adjusted for ancestry, baseline LDL-C, Lp(a), age, sex, family history of juvenile coronary heart disease, and the following medical features prior to index ACS: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting; and congestive heart failure. [Figure 28-2]Same as above. [Figure 28-3] Same as above. [Figure 29] A table is shown listing the risks by genetic decile, summarized across the entire PRS generation algorithm. [Figure 30-1] Plots of treated-stratified deciles for MACE in the LDPred, 27-SNP, and 57-SNP models are shown. These endpoints include death from coronary heart disease, non-fatal myocardial infarction, fatal or non-fatal ischemic attack, or unstable angina requiring hospitalization. Panel A shows the results for LDPred (p=0.001), Panel B shows the results for the 27-SNP model, and Panel C shows the results for the 57-SNP model. The upper column shows the event rate by decile of the genetic risk score in the alirocumab group, and the lower column shows the risk by decile in the placebo group. [Figure 30-2] Same as above. [Figure 30-3] Same as above. [Figure 30-4] Same as above. [Figure 30-5] Same as above. [Figure 30-6] Same as above. [Modes for carrying out the invention]

[0012] Genetic factors can play a significant role in the risk of developing a disease and may influence how individuals respond to drug treatment. PRS combines information from numerous gene variants derived from disease-related trials to create a single composite quantitative assessment criterion for each individual that reflects each individual's genetically generated disease risk. Individuals with a higher number of risk alleles for a given disease will have a higher PRS than individuals with fewer alleles. Risk can be assessed at several thresholds, such as percentiles or standard deviations of population distribution. This disclosure generally relates to the unexpected finding that subject stratification using CAD-PRS is useful in identifying individuals who may benefit from PCSK9 inhibitor treatment, independently of conventional clinical criteria such as LDL cholesterol levels.

[0013] Various terms relating to aspects of this disclosure are used throughout this specification and the claims. Unless otherwise specified, such terms shall be given their ordinary meanings in the art. Other specifically defined terms shall be construed in a manner consistent with the definitions provided herein.

[0014] Unless otherwise explicitly stated, no method or embodiment described herein is intended to be construed as requiring the steps to be performed in a specific order. Therefore, unless a method claim explicitly states in the claims or description that it limits the steps to a specific order, no order should be inferred in any way. This also applies to any possible implicit criteria of interpretation, including logical matters relating to the arrangement of a process or workflow, general meanings arising from grammatical structure or punctuation, or the number or types of embodiments described herein.

[0015] When used herein, unless otherwise clearly indicated by the context, the singular forms "a," "an," and "the" refer to multiple objects. When used herein, the term “approximately” means that the cited figures are approximate and small variations will not significantly affect the practice of the disclosed embodiments. Where figures are used, unless otherwise indicated by the context, the term “approximately” means that the figures may vary by ±10% and remain within the scope of the disclosed embodiments.

[0016] As used herein, the terms “subject” and “patient” are interchangeable. A subject may include any animal, including mammals. Examples of mammals include, but are not limited to, livestock (e.g., horses, cattle, pigs), companion animals (e.g., dogs, cats), laboratory animals (e.g., mice, rats, rabbits), and non-human primates. In some embodiments, the subject is human.

[0017] As used herein, “Major Cardiovascular Adverse Event” or “MACE” means one or more of the following: death due to coronary heart disease (CHD death), coronary artery disease (CAD), non-fatal myocardial infarction (MI), unstable angina requiring hospitalization, fatal or non-fatal ischemic attack, ischemic coronary vascular regeneration, arrhythmia, cardiovascular death, valvular heart disease, cardiomyopathy, or congestive heart failure.

[0018] As used herein, “MACE-risk patient” or “risk patient” refers to a patient with hypercholesterolemia and / or elevated levels of at least one atherosclerotic lipoprotein. In some embodiments, a MACE-risk patient has hypercholesterolemia and / or elevated levels of at least one atherosclerotic lipoprotein. In some embodiments, a MACE-risk patient is a patient who has previously had MACE.

[0019] The term "coronary revascularization due to ischemia" refers to percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG). For the clinical studies disclosed herein, coronary revascularization performed solely for restenosis at a previous PCI site is excluded from this definition. In some embodiments, coronary revascularization due to ischemia must be triggered by one of the following: a) acute ischemia, b) new or progressive symptoms (angina or equivalent), or 3) new or progressive functional abnormalities (e.g., stress tests or imaging).

[0020] As used herein, “death from coronary heart disease,” “CHD death,” and “death due to coronary heart disease” are interchangeable and include deaths following acute myocardial infarction (MI), sudden death, heart failure, complications of coronary artery revascularization performed for symptoms, progression of coronary artery disease, or new myocardial ischemia where the cause of death is clearly related to the procedure, unobserved unexpected death, and other deaths that cannot be clearly attributed to non-vascular causes.

[0021] As used herein, the terms “cardiovascular event” or “CV event” refer to non-fatal coronary heart disease events, cardiovascular death, and non-fatal ischemic attacks. Exemplary CV events include, but are not limited to, myocardial infarction, attacks, unstable angina requiring hospitalization, heart failure requiring hospitalization, and ischemic revascularization.

[0022] As used herein, the terms “cardiovascular death,” “CV death,” and “cardiovascular mortality” are interchangeable and refer to death due to acute myocardial infarction, sudden cardiac death, death due to heart failure, death due to seizure, and death due to other cardiovascular causes. In some embodiments, CV death is CHD death. In other embodiments, CV death is selected from the group consisting of heart failure or cardiogenic shock, seizure, ischemic cardiovascular causes, or non-ischemic cardiovascular causes.

[0023] As used herein, the term “non-fatal cardiovascular event” refers to any CV event that does not result in death. In some embodiments, non-fatal CV events may occur consecutively with a time interval between the first (e.g., first) CV event and subsequent (e.g., second, third, or fourth) events.

[0024] As used herein, “non-cardiovascular death” and “non-CV death” are interchangeable and refer to any death that is not considered to be cardiovascular death. Examples of non-cardiovascular deaths include, but are not limited to, lung infections, lung malignancies, gastrointestinal / hepatobiliary / pancreatic infections, gastrointestinal / hepatobiliary / pancreatic malignancies, hemorrhage, seizures / non-hemorrhagic neurological processes, suicide, non-cardiovascular procedures or surgeries, accidents or trauma, kidney infections, kidney malignancies, other non-cardiovascular infections, and other non-cardiovascular malignancies.

[0025] As used herein, “non-fatal myocardial infarction” is defined and subdivided according to the ACC / AHA / ESC universal definition of myocardial infarction (see Thygesen et al., J.Amer.Coll.Cardiol., 2012, 60, 1581-98).

[0026] As used herein, “coronary artery bypass grafting (CABG)” refers to a procedure in which an autologous artery or vein is used as a graft to bypass a coronary artery that is partially or completely occluded by atherosclerotic plaque (see Alexander & Smith, New Eng. J. Med, 2016, 374, 1954-64).

[0027] When used herein, the terms “unstable angina requiring hospitalization” and “hospitalization for unstable angina” are interchangeable and refer to hospitalization or emergency department admission due to symptoms of myocardial ischemia requiring either an accelerating tempo and / or resting chest discomfort of ≥20 minutes within the past 48 hours, and further including: a) a novel or presumed novel ischemic ECG change defined by ST depression greater than 0.5 mm in two nearby leads; T-wave inversion greater than 1 mm in two nearby leads with a prominent R wave or R / S > 1; ST elevation in two or more nearby leads with >0.2 mV in V2 or V3 in males, >0.15 mV in V2 or V3 in females, or >0.1 mV or LBBB in other leads; and b) hospitalization or emergency department admission due to symptoms of myocardial ischemia requiring either coronary revascularization or clear, modern evidence of coronary occlusion due to at least one epicardial stenosis ≥70%. For clinical trials disclosed herein, coronary revascularization or stenosis solely at the site of previous PCI was excluded.

[0028] As used herein, “ischemic attack” means 1) a) pathological, radiographic, or other objective evidence of acute, focal ischemic injury of the brain, spinal cord, or retina in a defined vascular distribution; or b) an acute episode of focal brain, spinal cord, or retinal dysfunction caused by an infarction defined by at least one of the symptoms of acute ischemic injury of the brain, spinal cord, or retina lasting 24 hours or until death, excluding other etiologies; 2) a hemorrhagic infarction that is not due to intracerebral or subarachnoid hemorrhage; or 3) an attack that is not otherwise subdivided.

[0029] As used herein, “high-intensity statin therapy” and “high-dose atorvastatin / rosuvastatin” are interchangeable and refer to the administration of 40-80 mg of atorvastatin or 20-40 mg of rosuvastatin daily.

[0030] As used herein, “maximum tolerated statin therapy” or “maximum tolerated statin therapy” are interchangeable and refer to a therapeutic dosing regimen that includes the administration of a daily dose of statins, which is the highest dose of statins that can be administered to a particular patient without causing unacceptable adverse side effects in the patient. Maximum tolerable statin therapy includes, but is not limited to, high-intensity statin therapy.

[0031] As used herein, a patient is considered “statin intolerant” or “intolerant to statins” if they have a medical history of experiencing one or more adverse reactions that began or worsened during a daily statin treatment regimen and ceased when statin therapy was discontinued. In some embodiments, adverse reactions are musculoskeletal in nature, such as skeletal muscle pain, aches, weakness, or spasms (e.g., myalgia, myopathy, rhabdomyolysis, etc.). Such adverse reactions are often more severe after exercise or exertion. Statin-related adverse reactions also include hepatic, gastrointestinal, and psychiatric symptoms that correlate with statin administration. In some embodiments, a patient is considered “statin intolerant” or “intolerant to statins” if any of the following apply to the patient: (1) having a history of skeletal muscle-related symptoms associated with at least two different daily statin therapy regimens; (2) exhibiting one or more statin-related adverse reactions to one or more approved minimum daily doses of statins; (3) being unable to tolerate a cumulative once-weekly statin dose seven times the approved minimum tablet size; (4) being able to tolerate low-dose statin therapy but experiencing symptoms when the dose is increased (e.g., to achieve a target LDL-C level); or (5) statins are contraindicated for the patient.

[0032] As used herein, “not adequately controlled” with respect to hypercholesterolemia means that a patient’s serum low-density lipoprotein cholesterol (LDL-C) concentration, total cholesterol concentration, and / or triglyceride concentration have not decreased to a recognized medically acceptable level (taking into account the patient’s relative risk of coronary heart disease) after at least four weeks of a therapeutic dosing regimen including a stable daily dose of statins. For example, patients with hypercholesterolemia not adequately controlled by statins include one or more patients with serum LDL-C concentrations of approximately 70 mg / dL or higher, approximately 80 mg / dL or higher, approximately 90 mg / dL or higher, approximately 100 mg / dL or higher, approximately 110 mg / dL or higher, approximately 120 mg / dL or higher, approximately 130 mg / dL or higher, or approximately 140 mg / dL or higher (depending on the patient’s potential risk of heart disease) after the patient has received a stable daily statin dosing regimen for at least four weeks.

[0033] As used herein, the expression “not adequately controlled” with respect to atherosclerotic lipoproteins means that the patient’s serum low-density lipoprotein cholesterol (LDL-C) concentration, non-high-density lipoprotein cholesterol, and / or apolipoprotein B concentration has not decreased to a recognized medically acceptable level after at least four weeks of a therapeutic dosing regimen including a stable daily dose of statins (taking into account the patient’s relative risk of coronary heart disease). For example, patients with elevated levels of atherosclerotic lipoproteins that are not adequately controlled by statins include patients (single) or patients (multiple) with serum LDL-C concentrations of approximately 70 mg / dL or higher, approximately 80 mg / dL or higher, approximately 90 mg / dL or higher, approximately 100 mg / dL or higher, approximately 110 mg / dL or higher, approximately 120 mg / dL or higher, approximately 130 mg / dL or higher, or approximately 140 mg / dL or higher (depending on the patient's potential risk of heart disease); non-high-density lipoprotein cholesterol concentrations of approximately 100 mg / dL or higher; or apolipoprotein B concentrations of approximately 80 mg / dL or higher.

[0034] This disclosure generally relates to methods and compositions for treating patients at high risk of MACE. In some embodiments, patients at high risk of MACE who are treatable by the methods of this disclosure have hypercholesterolemia (e.g., serum LDL-C concentration of 70 mg / dL or higher, or serum lipoprotein(a) (LPA or LP(a)) level of at least about 30 mg / dL). In some embodiments, patients at high risk of MACE who are treatable by the methods of this disclosure have been or are currently being treated with high doses of statins.

[0035] This disclosure also generally relates to methods and compositions for treating patients at high risk of MACE who have elevated levels of atherosclerotic lipoproteins. In some embodiments, patients at high risk of MACE who are treatable by the methods of this disclosure have high levels of atherosclerotic lipoproteins (e.g., serum LDL-C concentration of 70 mg / dL or higher, or serum lipoprotein(a) (LPA or Lp(a) level) of at least about 30 mg / dL). In some embodiments, patients at high risk of MACE who are treatable by the methods of this disclosure have been or are currently being treated with high doses of statins.

[0036] This disclosure generally relates to methods and compositions for reducing serum LDL and lipoprotein(a) levels in patients at high risk of MACE. In some embodiments, patients at high risk of MACE who are treatable by the methods of this disclosure have hypercholesterolemia (e.g., serum LDL-C concentration of 70 mg / dL or higher, or serum lipoprotein(a) (LPA or LP(a)) level of at least about 50 mg / dL). In some embodiments, patients at high risk of MACE who are treatable by the methods of this disclosure have been or are currently being treated with high doses of statins.

[0037] This disclosure also includes methods for treating patients at high risk of MACE with hypercholesterolemia and high levels of atherosclerotic lipoprotein that are not adequately controlled by the maximum tolerated dose of statin therapy. In some embodiments, the maximum tolerated dose of statin therapy includes daily administration of statins such as cerivastatin, pitavastatin, fluvastatin, lovastatin, and pravastatin.

[0038] Without being limited to any particular theory, the CAD-PRS calculated according to the method presented herein is thought to enable the identification of MACE-risk patients most likely to respond to PCSK9 inhibitor therapy. Furthermore, surprisingly and unexpectedly, the CAD-PRS also predicts the patient response to PCSK9 inhibitor therapy in patients without elevated levels of lipoprotein(a) (LPA or LP(a)) or LDL-C.

[0039] In some embodiments, patients at high risk of MACE, treatable by the methods of the present disclosure, have had MACE within the past 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, or 18 months. Patients at high cardiovascular risk, treatable by the methods of the present disclosure, include patients who have been hospitalized for MACE.

[0040] In some embodiments, patients at high risk of MACE may be selected based on CAD-PRS, where CAD-PRS includes a weighted sum of multiple gene variants associated with coronary artery disease and is calculated using at least about 2, at least about 3, at least about 4, at least about 5, at least about 10, at least about 20, at least about 30, at least about 40, at least about 50, at least about 60, at least about 70, at least about 80, at least about 100, at least about 120, at least about 150, at least about 200, at least about 250, at least about 300, at least about 400, at least about 500, or at least about 1,000 gene variants, and if a patient has a CAD-PRS above a threshold score, a PCSK9 inhibitor is administered targeted at an effective dose to lower serum LDL and lipoprotein(a) levels.

[0041] Risk assessment using a large number of gene variants offers the advantage of improved predictive power. In some embodiments, one or more gene variants are single nucleotide polymorphisms (SNPs). In some embodiments, one or more gene variants are insertions. In some embodiments, one or more gene variants are deletions. In some embodiments, one or more gene variants are structural variants. In some embodiments, one or more gene variants are copy number variants.

[0042] In some embodiments, this disclosure may, for example, include a large number of alleles derived from one or more gene variant databases, e.g., Nikpay et al., Nat. Genet., 2015, 47, 1121-1130 ("the Database") and gene variant databases available on the World Wide Web at "cardiogramplusc4d.org / media / cardiogramplusc4d-consortium / data-downloads / cad.additive.Oct2015.pub.zip", e.g., at least about 500,000 gene variants, at least about 1,000,000 gene variants, at least about 2,000,000 gene variants, The risk assessment should include at least approximately 3,000,000 gene variants, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants.

[0043] In some embodiments, the risk assessment may include evaluating all gene variants listed in the database. In some embodiments, this disclosure provides a method for determining CAD-PRS in a subject, the method comprising at least about 2 gene variants, at least about 5 gene variants, at least about 10 gene variants, at least about 15 gene variants, at least about 20 gene variants, at least about 30 gene variants, at least about 40 gene variants, at least about 50 gene variants, at least about 60 gene variants, at least about 70 gene variants, at least about 100 gene variants, at least about 200 gene variants, at least about 500 gene variants, at least about 1000 gene variants, at least about 2000 gene variants, at least about 5000 gene variants, at least about 10,000 gene variants, This involves identifying whether a genetic variant, at least approximately 20,000 gene variants, at least approximately 50,000 gene variants, at least approximately 75,000 gene variants, at least approximately 100,000 gene variants, at least approximately 500,000 gene variants, at least approximately 1,000,000 gene variants, at least approximately 2,000,000 gene variants, at least approximately 3,000,000 gene variants, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants are present in a biological sample from the subject; in this case, the presence of a risk allele increases CAD-PRS, and the presence of an alternative allele decreases CAD-PRS.

[0044] In some embodiments, the disclosure provides a method for determining the risk of MACE in a subject, which includes identifying whether database-derived gene variants are present in a biological sample from the subject, and calculating a CAD-PRS for the subject based on the identified gene variants, wherein the CAD-PRS is calculated by summing the weighted risk scores associated with each identified gene variant. The number of identified gene variants may be at least about 2, at least about 5, at least about 10, at least about 15, at least about 20, or at least about 30. At least approximately 40 gene variants, at least approximately 50 gene variants, at least approximately 95 gene variants, at least approximately 100 gene variants, at least approximately 200 gene variants, at least approximately 500 gene variants, at least approximately 1000 gene variants, at least approximately 2000 gene variants, at least approximately 5000 gene variants, at least approximately 10,000 gene variants, at least approximately 20,000 gene variants, at least approximately 50,000 gene variants, at least approximately 75,000 gene variants, at least approximately 100,000 gene variants, at least approximately 500,000 gene mutations, and a small number of others. It can be at least approximately 1,000,000 gene mutations, at least approximately 2,000,000 gene mutations, at least approximately 3,000,000 gene mutations, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants.

[0045] In some embodiments, the Disclosure provides a method for determining the risk of MACE in a subject, which includes identifying whether a database-derived gene variant is present in a biological sample from the subject, in which the identification may include at least about 50 gene variants, at least about 95 gene variants, at least about 100 gene variants, at least about 200 gene variants, at least about 500 gene variants, at least about 1000 gene variants, at least about 2000 gene variants, at least about 5000 gene variants, at least about 10,000 gene variants, at least about 20,000 gene variant variants, at least about 50,000 gene variants, at least about 75,000 gene variants, and at least This includes measuring the presence of approximately 100,000 gene variants, at least approximately 500,000 gene variants, at least approximately 1,000,000 gene variants, at least approximately 2,000,000 gene variants, at least approximately 3,000,000 gene variants, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants.

[0046] In some embodiments, the disclosure includes at least about 50 gene variants, at least about 95 gene variants, at least about 100 gene variants, at least about 200 gene variants, at least about 500 gene variants, at least about 1000 gene variants, at least about 2000 gene variants, at least about 5000 gene variants, at least about 10,000 gene variants, at least about 20,000 gene variant variants, at least about 50,000 gene variants, at least about 75,000 gene variants, at least about 100,000 gene variants, at least about 500,000 gene variants, at least about 1,000,000 gene variants, at least about 2,000,000 gene variants, and at least The present invention provides a method for determining the risk of MACE in a subject, comprising: selecting at least approximately 3,000,000 gene variants, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants from a database; identifying whether the gene variants are present in a biological sample from the subject; and calculating the PRS based on the presence of the gene variants.

[0047] In some embodiments, the disclosure provides a method for determining the risk of MACE in a subject, which includes identifying whether a database-derived gene variant is present in a biological sample from the subject, calculating a CAD-PRS for the subject based on the identified gene variant, and assigning the subject to a risk group based on the CAD-PRS. The CAD-PRS may be divided into quintiles, for example, upper quintile, middle quintile, and lower quintile, where the upper quintile of the polygene score corresponds to the highest gene risk group and the lower quintile of the polygene score corresponds to the lowest gene risk group. The number of identified gene variants is as follows: at least approximately 50 gene variants, at least approximately 95 gene variants, at least approximately 100 gene variants, at least approximately 200 gene variants, at least approximately 500 gene variants, at least approximately 1000 gene variants, at least approximately 2000 gene variants, at least approximately 5000 gene variants, at least approximately 10,000 gene variants, at least approximately 20,000 gene variant variants, at least approximately 50,000 gene variants, at least approximately 75,000 gene variants, at least approximately 100,000 gene variants, at least approximately 500,000 gene variants, and a small number of others. It can be at least approximately 1,000,000 gene variants, at least approximately 2,000,000 gene variants, at least approximately 3,000,000 gene variants, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants.

[0048] In some embodiments, this disclosure includes at least about 50 gene variants, at least about 95 gene variants, at least about 100 gene variants, at least about 200 gene variants, at least about 500 gene variants, at least about 1000 gene variants, at least about 2000 gene variants, at least about 5000 gene variants, at least about 10,000 gene variants, at least about 20,000 gene variant variants, at least about 50,000 gene variants, at least about 75,000 gene variants, at least about 100,000 gene variants, at least about 500,000 gene variants, at least about 1,000,000 gene variants, at least about 2,000,000 gene variants, and at least about 3,000,000 gene variants derived from a database. The present invention provides a method for selecting subjects or candidates at risk of developing MACE, comprising: identifying whether at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants are present in biological samples from each subject or candidate; calculating a polygenic risk score (CAD-PRS) for each subject or candidate based on the identified gene variants; and selecting subjects or candidates for a desired risk group.

[0049] For all MACE risk assessments, incorporating a large number of gene variants offers the advantage of improved predictive power. This disclosure further provides the risk assessment outlined above, for example, incorporating at least 500,000, at least 1,000,000, at least 2,000,000, at least 3,000,000, at least 4,000,000, at least 5,000,000, or at least 6,000,000 gene variants from a database, or at least 6,500,000 gene variants, or at least 7,000,000 gene variants, or at least 8,000,000 gene variants, or at least 9,000,000, or at least 10,000,000 gene variants.

[0050] In some embodiments, this disclosure includes at least about 50 gene variants, at least about 95 gene variants, at least about 100 gene variants, at least about 200 gene variants, at least about 500 gene variants, at least about 1000 gene variants, at least about 2000 gene variants, at least about 5000 gene variants, at least about 10,000 gene variants, at least about 20,000 gene variant variants, at least about 50,000 gene variants, at least about 75,000 gene variants, at least about 100,000 gene variants, at least about 500,000 gene variants, at least about 1,000,000 gene variants, at least about 2,000,000 gene variants, and at least about 3,000,000 gene variants derived from a database. The present invention provides a method for selecting a population of subjects or candidates at high risk of MACE, comprising: identifying whether gene variants, at least approximately 4,000,000 gene variants, at least approximately 5,000,000 gene variants, or at least approximately 6,000,000 gene variants, or at least approximately 6,500,000 gene variants, or at least approximately 7,000,000 gene variants, or at least approximately 8,000,000 gene variants, or at least approximately 9,000,000 gene variants, or at least approximately 10,000,000 gene variants, are present in biological samples from each subject or candidate; calculating the CAD-PRS for each subject or candidate based on the identified gene variants; and selecting subjects or candidates for a desired risk group.

[0051] In some embodiments, the number of identified gene variants is at least 20. In some embodiments, the number of identified gene variants is at least 30. In some embodiments, the number of identified gene variants is at least 40. In some embodiments, the number of identified gene variants is at least 50. In some embodiments, the number of identified gene variants is at least 70. In some embodiments, the number of identified gene variants is at least 100. In some embodiments, the number of identified gene variants is at least 500. In some embodiments, the number of identified gene variants is at least 1,000. In some embodiments, the number of identified gene variants is at least 2,000. In some embodiments, the number of identified gene variants is at least 5,000. In some embodiments, the number of identified gene variants is at least 10,000. In some embodiments, the number of identified gene variants is at least 20,000. In some embodiments, the number of identified gene variants is at least 50,000. In some embodiments, the number of identified gene variants is at least 75,000. In some embodiments, the number of identified gene variants is at least 100,000. In some embodiments, the number of identified gene variants is at least 500,000. In some embodiments, the number of identified gene variants is at least 1,000,000. In some embodiments, the number of identified gene variants is at least 2,000,000. In some embodiments, the number of identified gene variants is at least 3,000,000. In some embodiments, the number of identified gene variants is at least 4,000,000. In some embodiments, the number of identified gene variants is at least 5,000,000.In some embodiments, the number of identified gene variants is at least 6,000,000 gene variants. In some embodiments, the number of identified gene variants is at least 6,500,000 gene variants. In some embodiments, the number of identified gene variants is at least 7,000,000 gene variants. In some embodiments, the number of identified gene variants is at least 8,000,000 gene variants. In some embodiments, the number of identified gene variants is at least 9,000,000 gene variants. In some embodiments, the number of identified gene variants is at least 10,000,000 gene variants.

[0052] In some embodiments of the present disclosure, the risk assessment includes the highest weighted CAD-PRS scores including, but not limited to, the top 50%, 55%, 60%, 70%, 80%, 90%, or 95% of the CAD-PRS scores derived from the patient population.

[0053] In some embodiments, the identified gene variants include the most riskiest gene variants in the database or gene variants with top 10%, top 20%, top 30%, top 40%, or top 50% weighted risk scores.

[0054] In some embodiments, the identified gene variants include gene variants associated with MACE in the top 10%, top 20%, top 30%, top 40%, or top 50% of the p-value range of the database. In some embodiments, each of the identified gene variants is about 10 in the database -1 about 10 -2 about 10 -3 about 10- 4 about 10 -5 about 10 -6 about 10 -7 about 10 -8 about 19 -9 about 10 -10 about 10 -11 about 10 -12 about 10 -13 about 10-14 , or about 10 -15 The following gene variants associated with MACE have p-values. In some embodiments, the identified gene variants are listed in the database as 5 × 10⁶. -8 Includes gene variants associated with MACE with p-values ​​less than <

[0055] In some embodiments, identified gene variants include gene variants associated with MACE in high-risk patients compared to the rest of the reference population where the odds ratio (OR) is approximately ≥1.0, ≥1.5, ≥1.75, ≥2.0, or ≥2.25 for the top 50% of the distribution; or approximately ≥1.5, ≥1.75, ≥2.0, ≥2.25, ≥2.5, or ≥2.75 compared to the rest of the reference population. In some embodiments, the odds ratio (OR) may be in the range of approximately 1.0 to approximately 1.5, approximately 1.5 to approximately 2.0, approximately 2.0 to approximately 2.5, approximately 2.5 to approximately 3.0, approximately 3.0 to approximately 3.5, approximately 3.5 to approximately 4.0, approximately 4.0 to approximately 4.5, approximately 4.5 to approximately 5.0, approximately 5.0 to approximately 5.5, approximately 5.5 to approximately 6.0, approximately 6.0 to approximately 6.5, or approximately 6.5 to approximately 7.0. In some embodiments, high-risk patients include patients in the reference population whose CAD-PRS score is in the upper decile, quintile, or tertile.

[0056] In some embodiments, the identified gene variants include those with the best gene variant performance in the reference population. In some embodiments, the gene variant performance is calculated with respect to the risk of coronary artery disease based on statistical significance, strength of association, and / or probability distribution.

[0057] In some embodiments, gene variant scores are calculated using a PRS calculation method such as the LDPred method (or its variations and / or versions), which is a Bayesian approach that calculates the posterior mean effect of all variants based on prior (effect size in previous GWAS) and subsequent reduction based on linkage disequilibrium. LDPred produces the PRS using genome-wide variations with weights derived from a set of GWAS summary statistics. See Vilhjalmsson et al., Am.J.Hum.Genet., 2015, 97, 576-92. In some embodiments, alternative approaches to calculating gene variant scores may be used, including SBayesR (Lloyd-Jones, LR, World Wide Web at “biorxiv.org / content / biorxiv / early / 2019 / 01 / 17 / 522961.full.pdf”), Pruning and Thresholding (P&T) (Purcell, Nature, 2009, 460, 748-752), and COJO (Yang et al., Nat. Genet., 2012, 44, 369-375). SBayesR is a Bayesian approach similar to LDPred, but allows for greater flexibility with posterior mean effects. Pruning and thresholding involve a minimum p-value threshold (the p-value associated with the variant from the original data file) and r between variants. 2 A threshold (LD criterion) needs to be specified. P&T identifies the mutant with the smallest p-value in each region, and then, under that mutant, the specified r 2 Larger than r 2 PRS "collates" all other variants within a region that has a value. In PRS, the index variant represents all variants in the collection (only the index variant is included in PRS, and all other variants are excluded). COJO, or Conditional and Co-Association Analysis, is conceptually similar to P&T, but after conditioning with the index variant, it incorporates additional variants in a given LD block into the score if they show an independent contribution to disease risk.

[0058] In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.0001 to approximately 0.5. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.5. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.1. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.05. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.01. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.005. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.001. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.0005. In some embodiments, the performance of the gene variant is calculated using the LDPred method, where the ρ value is approximately 0.0001.

[0059] In some embodiments, the method further includes a first step of obtaining a biological sample from the subject. When used herein, “biological sample” may contain whole cells and / or living cells and / or cellular debris. A biological sample may contain (or be derived from) “body fluids.” This disclosure encompasses embodiments in which body fluids are selected from amniotic fluid, aqueous body fluids, vitreous fluid, bile, serum, milk, cerebrospinal fluid, earwax, chyle, atherosclerotic fluid, endolymph, perilymphatic fluid, exudate, feces, vaginal fluid, gastric acid, gastric juice, lymph, mucus (including nasal discharge and sputum), pericardial fluid, ascites, pleural fluid, pus, eye discharge, saliva, sebum (skin oil), semen, saliva, synovial fluid, sweat, tears, urine, vaginal secretions, vomit, and mixtures of one or more thereof. Biological samples include cell cultures, body fluids, and cell cultures derived from body fluids. Body fluids may be obtained from mammalian organisms, for example, by puncture or by other collection or sampling procedures.

[0060] In some embodiments, the method is used to select a population of subjects or candidates for a clinical trial, for example, a clinical trial to determine whether a particular treatment or treatment plan is effective against MACE or recurrent MACE. In some embodiments, the selected candidates or subjects are divided into subgroups based on gene variants identified for each subject or candidate, and the method is used to determine whether a particular treatment or treatment plan is effective for a particular gene variant or a particular group of gene variants. In other words, the method can be used to determine the susceptibility of a population of subjects to a particular treatment or treatment plan, in which case the population of subjects is selected based on gene variants identified in the subjects.

[0061] In some embodiments, the method is used to select a population of subjects or candidates for a clinical trial, for example, a clinical trial to determine whether a particular treatment or treatment plan is effective against MACE or recurrent MACE. In some embodiments, the desired risk group is a population that includes high-risk subjects or candidates. In some embodiments, the selected population of subjects or candidates are responders, i.e., subjects or candidates respond to the treatment or treatment plan.

[0062] In some embodiments, subjects are selected based solely on CAD-PRS. For example, if a patient or candidate subject has a CAD-PRS above a predetermined threshold, the patient is selected to initiate treatment, or the candidate subject is included in a clinical trial. In some embodiments, the threshold for initiating treatment or clinical trial inclusion is determined in relative terms. For example, in some embodiments, the threshold CAD-PRS score is in the top 50% of the reference population. In some embodiments, the threshold CAD-PRS score is in the top 40% of the reference population. For example, in some embodiments, the threshold CAD-PRS score is in the top 30% of the reference population. For example, in some embodiments, the threshold CAD-PRS score is in the top 25% of the reference population. For example, in some embodiments, the threshold CAD-PRS score is in the top 20% of the reference population. For example, in some embodiments, the threshold CAD-PRS score is in the top 15% of the reference population. For example, in some embodiments, the threshold CAD-PRS score is in the top 10% (decile) of the reference population. For example, in some embodiments, the threshold CAD-PRS score is in the top 5% of the reference population.

[0063] In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 100 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 200 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 500 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 1,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 3,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 5,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 7,500 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 10,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 12,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 15,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 20,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 30,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 50,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 70,000 patients. In some embodiments, the reference population for determining relative CAD-PRS scores is at least about 100,000 patients.

[0064] In some embodiments, the reference population is enriched with respect to the members of the ancestral group. In some embodiments, the ancestral group is self-reported. In some embodiments, the ancestral group is derived from principal component analysis of the ancestors. In some embodiments, the ancestral group is European. In some embodiments, the ancestral group is African. In some embodiments, the ancestral group is mixed American. In some embodiments, the ancestral group is East Asian. In some embodiments, the ancestral group is South Asian. In some embodiments, the ancestral group is any two or more mixtures of the European, African, mixed American, East Asian, and South Asian populations.

[0065] In some embodiments, the method further includes determining a composite risk score comprising CAD-PRS and low-density lipoprotein (LDL) levels in a biological sample obtained from a patient. For example, if a patient or candidate subject has both CAD-PRS and LDL levels above a predetermined threshold in a biological sample obtained from the patient or subject, the patient is selected to initiate treatment, or the candidate subject is included in a clinical trial. In some embodiments, the biological sample includes serum. In some embodiments, the threshold serum LDL level is at least about 100 mg / dL. In some embodiments, the threshold serum LDL level is at least about 120 mg / dL. In some embodiments, the threshold serum LDL level is at least about 140 mg / dL. In some embodiments, the threshold serum LDL level is at least about 160 mg / dL. In some embodiments, the threshold serum LDL level is at least about 180 mg / dL. In some embodiments, the threshold serum LDL level is at least about 200 mg / dL.

[0066] In some embodiments, the method further includes determining a composite risk score comprising CAD-PRS and low-density lipoprotein(a) (LPA or LP(a)) levels in a biological sample obtained from a patient. For example, if a patient or candidate subject has both CAD-PRS and LPA levels above a predetermined threshold in a biological sample obtained from the patient or subject, the patient is selected to initiate treatment, or the candidate subject is included in a clinical trial. In some embodiments, the biological sample includes serum. In some embodiments, the threshold serum LPA level is at least about 30 mg / dL. In some embodiments, the threshold serum LPA level is at least about 40 mg / dL. In some embodiments, the threshold serum LPA level is at least about 50 mg / dL. In some embodiments, the threshold serum LPA level is at least about 120 mg / dL. In some embodiments, the threshold serum LPA level is at least about 60 mg / dL. In some embodiments, the threshold serum LPA level is at least about 70 mg / dL. In some embodiments, the threshold serum LPA level is at least about 80 mg / dL. In some embodiments, the threshold serum LPA level is at least about 100 mg / dL. In some embodiments, the threshold serum LPA level is at least about 120 mg / dL. In some embodiments, the threshold serum LPA level is at least about 140 mg / dL.

[0067] In some embodiments, the method further includes determining a composite risk score comprising CAD-PRS, LPA levels, and LDL-C levels in a biological sample obtained from the patient. For example, if a patient or candidate has CAD-PRS and both LDL and LPA levels above a predetermined threshold in a biological sample obtained from the patient or candidate, the patient is selected to initiate treatment, or the candidate is included in a clinical trial.

[0068] In some embodiments, the method further includes determining a composite risk score including CAD-PRS and LPA levels in a biological sample obtained from a patient. In some embodiments, the method further includes determining a composite risk score including CAD-PRS, LDL levels and LPA levels in a biological sample obtained from a patient. In some embodiments, the method further includes determining a composite risk score including CAD-PRS and Framingham (FHS) recurrence risk score (see D'Agostino et al., Am. Heart J., 2000, 139, 272-281) in a biological sample obtained from a patient. In some embodiments, the method further includes determining a composite risk score including CAD-PRS and very high risk (VHR) group (Roe et al., Circulation, 2019, 140, 1578-1589) in a biological sample obtained from a patient. Therefore, in some embodiments, the composite risk score may include CAD-PRS and one or more of the following in a biological sample obtained from the patient: LPA level, LDL level, Framingham (FHS) recurrence risk score, and VHR group. In some embodiments, the biological sample includes blood.

[0069] In some embodiments, the method further includes initiating treatment on a subject. The treatment may include statins, ezetimibe, beta-blockers, angiotensin-converting enzyme inhibitors, aspirin, anticoagulants, antiplatelet agents, angiotensin II receptor blockers, angiotensin receptor neprilysin inhibitors, calcium channel blockers, cholesterol-lowering agents, vasodilators, antidiuretics, renin-angiotensin system agents, lipid modifiers, anti-inflammatory agents, nitrates, antiarrhythmics, steroidal or nonsteroidal anti-inflammatory drugs, DNA methyltransferase inhibitors and / or histone deacetylase inhibitors. The DNA methyltransferase inhibitor may be any DNA methyltransferase known in the art, e.g., 5-aza-2'-deoxycytidine or 5-azacitidine. The histone deacetylase inhibitor may be any histone deacetylase inhibitor known in the art, e.g., barinostat, romidepsin, panobinostat, bellinostat or entinostat. The statin may be any statin known in the art, such as atorvastatin, fluvastatin, lovastatin, pravastatin, rosuvastatin, and simvastatin. The lipid modifier may be any lipid modifying compound known in the art, such as a PCSK9 inhibitor, an antisense oligonucleotide targeting apolipoprotein C-III, and an antisense oligonucleotide that reduces lipoprotein (a).

[0070] Initiating treatment may involve devising a treatment plan based on risk groups, which corresponds to a CAD-PRS calculated for the patient. In some embodiments, the CAD-PRS predicts the treatment effect or the patient's response to the treatment plan. In some embodiments, a composite risk score (CAD-PRS combined with LDL levels, LPA levels, or both) predicts the treatment effect of the patient's response to the treatment plan. Therefore, treatment can be determined or adjusted according to the CAD-PRS.

[0071] In some embodiments, initiating treatment involves modifying the dosage or dosing plan of treatment already being received by a patient at risk of MACE or with hypercholesterolemia (e.g., treatment with statins that do not adequately control hypercholesterolemia) based on a CAD-PRS calculated for the patient. In some embodiments, initiating treatment involves replacing one therapeutic agent with another based on a CAD-PRS calculated for a patient at risk of MACE or with hypercholesterolemia, for example, if the patient is intolerant to statins. In some embodiments, initiating treatment involves initiating a dosing plan for a therapeutic agent in addition to the therapeutic agent the patient is already receiving, for example, initiating a PCSK9 inhibitor dosing plan in a patient at risk of MACE or with hypercholesterolemia who is receiving statin treatment such as high-intensity statin therapy or maximally tolerable statin therapy. In some embodiments, initiating treatment involves initiating the implementation of a therapeutic dosing plan in a previously untreated patient at risk of MACE or with hypercholesterolemia.

[0072] In some embodiments, the additional therapeutic agent is a PCSK9 inhibitor. In some embodiments, the CAD-PRS predicts the therapeutic effect or the patient's response to PCSK9 inhibitor therapy. In some embodiments, a composite risk score (CAD-PRS combined with LDL levels, LPA levels, or both) predicts the patient's response to PCSK9 inhibitor therapy. Therefore, PCSK9 inhibitor therapy can be determined or adjusted according to the CAD-PRS calculated for the patient.

[0073] When used herein, the term “proprotein convertase subtilisin / kexin type 9” or “PCSK9” refers to Sequence ID No. 1: Amino acid sequence of Sequence ID No. 2: , or human PCSK9 having those fragments that are biologically active.

[0074] As used herein, the term “inhibitor” means that a given compound can inhibit the activity of a particular protein or other substance in a cell to at least a specific degree. This can be achieved by a direct interaction between the compound and a given protein or substance ("direct inhibition"), or by an interaction between the compound and other proteins or substances inside or outside the cell that leads to at least partial inhibition of the activity of the protein or substance ("indirect inhibition"). Inhibition of protein activity can also be achieved by suppressing the expression of a target protein. Techniques for inhibiting protein expression include, but are not limited to, antisense inhibition, siRNA-mediated inhibition, miRNA-mediated inhibition, ribozyme-mediated inhibition, DNA-directed RNA interference (DdRNAi), RNA-directed DNA methylation, activator-like effector nuclease (TALEN)-mediated inhibition, zinc finger nuclease-mediated inhibition, aptamer-mediated inhibition, and CRISPR-mediated inhibition.

[0075] As used herein, “antisense inhibition” means a decrease in the level of the target nucleic acid in the presence of an oligonucleotide complementary to the target nucleic acid, compared to the level of the target nucleic acid in the absence of the oligonucleotide.

[0076] In some embodiments, PCSK9 inhibitors are small molecules. Numerous small molecule inhibitors of PCSK9 are described, for example, in U.S. Patent No. 10,131,637. In some embodiments, the PCSK9 inhibitor is an siRNA. An example siRNA is inclisilan (see Ray et al., Circulation, 2018, 138, 1304-1316).

[0077] In some embodiments, the PCSK9 inhibitor is an anti-PCSK9 antibody or its antigen-binding moiety. As used herein, the term “antibody” is intended to refer to an immunoglobulin molecule comprising four polypeptide chains, two heavy (H) chains and two light (L) chains linked together by disulfide bonds, as well as its polymer (e.g., IgM). Each heavy chain has a heavy chain variable region (HCVR or V) as used herein. H It includes the heavy chain steady region (which is abbreviated as C). H 1, C H 2, and C H It contains three domains. Each light chain has a light chain variable region (LCVR or V in this specification). L It includes the (abbreviated) and light chain steady region. The light chain steady region is one domain (C L Includes 1). V H and V L The region can be further subdivided into a highly variable region called the Complementarity Determination Region (CDR), which is sandwiched between even more conserved regions called the Framework Region (FR). H and V L It consists of three CDRs and four FRs arranged from the amino terminus to the carboxyl terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. In various embodiments, the FRs of the anti-PCSK9 antibody (or its antigen-binding moiety) may be identical to the human germline sequence, or they may be naturally or artificially modified. The amino acid consensus sequence may be defined based on a parallel analysis of two or more CDRs.

[0078] As used herein, the term “antibody” also includes the antigen-binding fragment of a complete antibody molecule. As used herein, terms such as “antigen-binding portion” of an antibody and “antigen-binding fragment” of an antibody include any naturally occurring, enzymatically obtained, synthetic, or genetically engineered polypeptide or glycoprotein that specifically binds to an antigen to form a complex. The antigen-binding fragment of an antibody may be derived, for example, from a complete antibody molecule using any preferred standard method such as protein digestion, or recombinant genetic engineering methods including the manipulation and expression of DNA encoding the variable domain and optionally constant domain of the antibody. Such DNA is readily available, for example, from commercial sources, DNA libraries (including, for example, phage antibody libraries), or can be synthesized. The DNA may be sequenced and manipulated, for example, using chemical or molecular biological methods, to arrange one or more variable domains and / or constant domains in a preferred configuration, or to modify, add, or delete amino acids, to introduce codons and create cysteine ​​residues.

[0079] Examples of anti-PCSK9 antibodies include, but are not limited to, evolocumab, alirocumab, and vococizumab. Additional anti-PCSK9 antibodies are described, for example, in U.S. Patents 10,259,885, 10,023,654, 9,266,961, 9,561,155, 9,550,837, 9,540,449, 9,029,515, 8,951,523, 8,859,741, 8,530,414, 8,829,165, 8,802,827, 8,710,192, 8,344,114, and 8,188,233. Additional anti-PCSK9 antibodies include evolocumab, alirocumab, or vococizumab V H , V L , and / or antibodies containing CDRs are included.

[0080] In the context of the method, additional therapeutically active ingredients, such as any of the above-mentioned agents or their derivatives, may be administered immediately before, concurrently with, or immediately after the administration of the PCSK9 inhibitor (for the purposes of this disclosure, such a dosing regimen is considered to be the administration of the PCSK9 inhibitor "in combination" with the additional therapeutically active ingredients). The methods of the present invention include pharmaceutical compositions and methods of use thereof in which a PCSK9 inhibitor is formulated concurrently with one or more additional therapeutically active ingredients, as described elsewhere herein.

[0081] All patent documents, websites, other publications, accession numbers, etc., cited above or below are incorporated by reference in whole for any purpose to the same extent as if each individual item were specifically and individually instructed to be incorporated by reference in that manner. Where different versions of an array are associated with different accession numbers, it means the version associated with the accession number of the effective filing date of this application. The effective filing date means, where applicable, the earlier of the actual filing date or the filing date of the priority claim application to which the accession number is referred. Similarly, where different versions of a publication, website or similar are published over time, it means the most recent published version as of the effective filing date of this application, unless otherwise specified. Any feature, process, element, embodiment, or aspect of this disclosure may be used in combination with any other feature, process, element, embodiment, or aspect unless otherwise specifically specified. While this disclosure has been described in some detail as examples and embodiments for the purposes of clarity and understanding, it will be clear that certain changes and modifications may be implemented within the scope of the attached claims.

[0082] The following examples are provided to illustrate the embodiments in more detail. They are intended to illustrate, and not limit, the embodiments claimed. The following examples provide to those skilled in the art a disclosure and description of how the compounds, compositions, articles, devices and / or methods described herein are prepared and evaluated, and are intended to be merely illustrative and not to limit any of the claims. With respect to numbers (e.g., quantities, temperatures, etc.), efforts have been made to ensure accuracy, but some degree of error and deviation should be taken into consideration. Unless otherwise indicated, parts are parts by weight, temperatures are °C or ambient temperature, and pressures are atmospheric pressure or near atmospheric pressure. [Examples]

[0083] Example 1: ODYSSEY OUTCOMES Clinical Trial The ODYSSEY OUTCOMES trial was a randomized, double-blind comparison of alirocumab or placebo in 18,924 patients with recent hospitalization (1–12 months prior) for ACS (myocardial infarction or unstable angina). Eligible patients had LDL-C cholesterol levels of ≥70 mg / dL, apolipoprotein B levels of ≥80 mg / dL, or non-HDL cholesterol levels of ≥100 mg / dL despite high-intensity or maximally tolerable statin treatment. Patients were assigned in a 1:1 ratio to receive either alirocumab or the corresponding placebo every two weeks. The primary endpoint, MACE, was a composite of death from coronary heart disease, non-fatal myocardial infarction, ischemic attack, or unstable angina requiring hospitalization. The median follow-up was 2.8 years. MACE occurred in 1052 patients (11.1%) in the placebo group and 903 patients (9.5%) in the alirocumab group (hazard ratio (HR), 0.85; 95% confidence interval (CI), 0.78–0.93; p-value < 0.001).

[0084] Genetic data generation DNA samples were obtained from 12,118 trial participants who provided written informed consent to participate in the pharmacogenomics trial. Genotyping was performed using the Illumina Global Screening Array (GSA), v1.0. Additional genetic data were allocated using the Minimac3 program. The imputation reference population was obtained from 1000 genomes phase 3 version 5 data. Of the 12,118 samples, 11,953 (98%) met the genetic data quality control procedures.

[0085] The gene variants and summary statistics used in the development of the PRS were obtained from a genome-wide meta-analysis of 60,801 cases of coronary artery disease and 123,504 controls. These variants (up to n=6,579,025) and their corresponding disease-related effect sizes (odds ratios) were used in the development of the PRS genome-wide using a pruning and threshold (P&T) approach and the LDPred algorithm. For comparison with previous publications of CAD-PRS in statin response, 27- and 57-variant models were also evaluated. The PRS was calculated for each patient by taking the product of the number of risk alleles in the patient and the weight of each variant (log odds ratio or LDPred-adjusted log odds ratio) for each variant, and summing them over all variants. These scores were examined and validated using two large, independent databases: DiscovEHR (n=84,243) and UK Biobank (n=446,208). Patients in the ODYSSEY OUTCOMES trial were assigned to one of five ancestral groups (Africa, Mixed Americas, East Asia, Europe, or South Asia). Ancestral population classification was based on the similarity between each patient's genotype and publicly available genetic data from the International HapMap Project. Population structure was assessed using principal component analysis with plink software. Subsequent risk score calculations were stratified by ancestry. Within each ancestral group, PRS was standardized to mean zero and standard deviation 1 to integrate the datasets and enable comparisons between ancestry. High genetic risk was defined as patients within the upper decile (>90 PRS percentile) of the PRS distribution. Those below the upper decile were defined as having low genetic risk (≤90 PRS percentile). This threshold was selected in a post-hoc analysis evaluating high genetic risk thresholds ranging from 50% to 90% in 10% increments. PRS was also evaluated as a continuous assessment criterion.

[0086] Processing of genetic data Genotyping method. Microarray genotypes for genome-wide association analysis (GWAS) were generated using the Illumina Global Screening Array (GSA), v1.0 (GSA-24v1-0_A1). This array contains approximately 660,000 markers, with an average marker spacing of 4.2 kb.

[0087] Illumina microarray genotyping data QC. Individual samples with a call rate of less than 90%, and those with a call rate of less than 90% or a Hardy-Weinberg equilibrium p-value of 1 × 10⁻⁶. -6 Genetic variants below a certain threshold were excluded from the analysis. In paired samples with IBD ≥ 0.25, samples with low call rates were excluded. Samples were also excluded if a sex discrepancy was detected between the sex estimated from the X chromosome and the sex reported in the clinical database.

[0088] Principal Component Analysis (PCA). Population structure was assessed using PCA within plink version 1.9. Two sets of analysis were performed: 1) ancestral group assignment and 2) generation of ancestral-specific PCs. Ancestral population assignment was based on the similarity between each patient's genotype and publicly available genetic data from the International HapMap Project. PCA was performed on an integrated dataset of ODYSSEY CVOT and HapMap samples. The probability of each sample belonging to one of five HapMap hyperpopulations / ancestral groups (African (AFR), Mixed American (AMR), East Asian (EAS), European (EUR), or South Asian (SAS)) was calculated and used to classify the samples. PCA was performed within the entire PGx population and ancestral groups to generate ancestral-specific PCs. The top 4-12 PCs (depending on ancestry) were used as covariates in the analysis.

[0089] Imputation. Genotype imputation was performed using Minimac3. The reference population for imputation was obtained from 1000 genomes from Phase 3 Version 5. Post-QC variants were limited to those with an INFO score greater than 0.3. Similar thresholds were applied to deletions and HWE. For assigned variants, allele doses were used to calculate the PRS.

[0090] Generation of a multi-gene risk score Datasets. The primary data source for polygenic risk scores is a GWAS of CAD risk containing 9.4 million variants derived from a meta-analysis of 60,801 CAD cases and 123,504 controls. A set of PRS algorithm adjustment parameters were assessed based on performance in two datasets: UK Biobank (UKB) and DiscovEHR. Composite cardiovascular endpoints of myocardial infarction, unstable angina, and ischemic attack (defined by ICD-10 codes I21*, I22*, I23*, I24.1, I25.2, I2.0, I63.0) were used, along with self-report code 20002* (UKB), to define the status of cases and controls.

[0091] Selection of PRS algorithms. Three approaches for generating polygene risk scores were examined: candidate SNP models from previous studies on the merits of PRS in statins including 27 and 57 variants, pruning and thresholding (P&T), and LDPred. P&T identifies variants with the lowest p-values ​​in each region, and then, under those variants, a specified r 2 Larger than r 2All other variants within a region with a value are "collected". In PRS, the indicator variant represents all variants in the collection (only the indicator variant is included in the PRS, and all other variants are excluded). LDPred is a Bayesian approach to PRS development that calculates the posterior mean effect (adjusted effect size) for all variants based on prior and LD information from a reference panel. To facilitate discovery, LDPred jointly models the size and variance of the effect of each marker, and the effect sizes generated from LDPred differ from P&T in that it incorporates the LD structure when reducing the effect size. Adjustment or reduction of variant weights is based not only on the magnitude of the variant's association with the disease, but also on the linkage disequilibrium (LD) between variants. For both the P&T and LDPred approaches, data from 1000 Genomes Phase 3 Version 5 were used as the LD reference panel.

[0092] PRS calculation. A set of variants and their respective weights were generated using either the LDPred or P&T approach. In the P&T case, the weights were the log odds ratios from the meta-analysis raw data. In the LDPred case, the variant weights were the adjusted log odds ratios (posterior mean). After weight generation, the process of score calculation and standardization is identical. For a set of i=1...M variants in j=1...N patients, the PRS for patient j is:

[0093]

number

[0094] It is calculated by, in the formula, B i is the log odds ratio for mutant i, and x ij is the number of risk alleles possessed by patient j in mutant i (or the allele dose of mutant i in the case of a assumed mutant). The score was standardized to ~N(0.1) by subtracting the mean PRS and dividing by the PRS standard deviation within each ancestral group.

[0095] Examination and validation of the PRS algorithm. For each set of adjusted parameters in LDPred or P&T, the PRS was calculated, and logistic regression was performed with the composite endpoint as the dependent variable and PRS, age, sex, genotyping array (UK Biobank only), and ancestral covariates as independent variables. For each model, the odds ratio (OR) and area under the curve (AUC) per PRS standard deviation (SD) were reported. 28 P&T models (5 × 10) were analyzed. -1 ~5×10 -8 The index p-values ​​in the range of 0.2, 0.4, 0.6, and 0.8 r 2 Value) and LDPred model of 8 (ρ=3×10 -1 , 1 x 10 -1 , 3 x 10 -2 , 1 x 10 -2 , 3 x 10 -3 , 1 x 10 -3 , 3 x 10 -4 and 3 × 10 -4 We investigated the following. In both the UKB dataset and the DiscovEHR dataset, LDPred with ρ=0.001 showed the best performance and was used in the primary analysis; the results are shown in Figures 9-11.

[0096] Selection of thresholds for defining high risk. Previous publications on PRS-CAD risk have used different thresholds to define high risk by PRS, with most thresholds ranging from the upper tertile to the upper quintile. Genetically high risk was defined as patients within the upper decile of the distribution of polygeneic risk scores. This threshold was selected in a post-hoc analysis evaluating high genetic risk thresholds ranging from top >10% to top >50% in 10% increments. In the placebo group, the risk of the top 10% events by PRS was consistently higher than the overall event rate, and this effect was consistent across ancestral groups (Figure 12). While no risk trend was observed in the placebo group (specifically percentiles 70–90) for the primary endpoint (Figure 13), these deciles had a higher-than-mean risk across many secondary endpoints, including coronary heart disease (Figure 15), the endpoint that most closely matches the CAD criteria used in the original dataset. However, although trends were observed in several secondary endpoints, only the top 10% of deciles showed a consistent difference in treatment efficacy.

[0097] Further analysis was performed, which showed that the difference in risk for the primary endpoint in the upper decile differed from all other deciles. Thirty-six genetic risk scores were calculated using ODYSSEY OUTCOMES across the PRS method and threshold range described above. Using each score, a Cox model was run within each treatment group to assess the risk for the primary endpoint in each decile compared to all other deciles. Next, within each decile, the proportion of events was calculated for each group and for the hazard ratio of the treatment difference (risk in alirocumab compared to placebo). These results are summarized in Figure 29, presenting the results for the first (Q1) and third (Q3) quartiles, as well as the median HR (or proportion of events) for each decile. Columns 2 and 3 show the genetic risk results (risk in that decile compared to all others), and columns 4-6 focus on the differences in treatment-related risk within deciles.

[0098] In these summarized estimates, only the upper decile showed an increased risk in the placebo group and a treatment effect exceeding the overall trial estimate. The hazard ratio associated with genetic risk in the placebo group increased from 1.03 to 1.24 from the 9th decile to the upper decile. Conversely, the difference was small in the alirocumab group (1.07 vs. 1.09). Although the results for a certain proportion of events were similar, the proportion of the upper decile in the placebo group was more than 2% higher than any other decile, whereas the risk in the upper decile in the alirocumab group was indistinguishable from that of the other deciles. Due to the difference in risk between the placebo and alirocumab groups at the upper decile, the median hazard ratio for the treatment difference was estimated to be 0.70, compared to 0.80–0.89 for all other deciles.

[0099] It should be noted that at the upper decile, the greatest difference in treatment was observed in the genetic score derived from LDPred across all 36 gene risk score generation algorithms (rho=0.001). This method was a priori selected from testing on two independent datasets because it demonstrated the best ability to distinguish CAD cases from controls. Across all gene risk score algorithms with scores ranging from fewer than 100 markers to more than 6 million genetic markers, the effect at the upper decile for the placebo group generally differed from that at other deciles. The median treatment effect observed at the upper genetic decile (HR=0.70) was greater than that observed in patients with elevated baseline LDL-C across the entire trial (HR=0.76). These PRS findings are consistent with previous analyses of gene risk scores in statin therapy (n=10,456), where the benefits of statin use did not follow a clear linear trend. While some variability in the magnitude of treatment effectiveness is acceptable, the complete summary of the results suggests that patients in the high-gene risk group (top 10%) may enjoy even greater treatment benefits.

[0100] statistical analysis Baseline disease and medical history characteristics were analyzed to assess the distribution of cardiovascular risk factors by genetic risk status and high (> percentile threshold) and low (≤ percentile threshold) levels. Continuous baseline characteristics were compared using t-tests, and two-component or categorical characteristics were tested using chi-square tests or Fisher's exact test. Baseline lipids were regressiond with age, sex, and ancestry as covariates; residuals from the models were transformed using rank-inverse normal transformation (RINT) before comparing genetic risk groups. Lipid changes at 4 months were similarly analyzed using RINT residuals from linear regression models adjusted for baseline, age, sex, and ancestry as covariates.

[0101] In this analysis, MACE and all secondary endpoints followed the definitions and treatment intention analysis approach of the ODYSSEY OUTCOMES trial. The primary analysis was the time to first occurrence of a component of the composite primary endpoint. The relationship between PRS and MACE or other efficacy endpoints was evaluated using two different analytical approaches. First, the risk of MACE and secondary endpoints in patients receiving placebo was assessed using a Cox proportional hazards model. PRS was modeled as both continuous and bicomponent (above / below threshold) covariates. Second, treatment effects were evaluated using a Cox model stratified by genetically defined high-risk and low-risk groups. An unstratified Cox model with a treatment interaction term by genetic risk was also performed to determine whether the benefit of alirocumab treatment differed between genetically defined risk groups. An inverse variance-weighted meta-analysis was also used to combine PRS risk estimates for placebo and alirocumab. Unless otherwise specified, all inferential analyses were performed after adjusting for baseline and clinical covariates, including ancestry, age, sex, baseline LDL-C, Lp(a), family history of juvenile coronary heart disease, and the following pre-ACS medical features that strongly predicted the trial outcomes and were disproportionate between genetically defined risk groups: myocardial infarction; percutaneous coronary intervention; coronary artery bypass grafting; and congestive heart failure. Risk factor stratification analyses were also performed by Lp(a) (≥50 mg / dL vs <50 mg / dL), LDL-C (≥100 mg / dL vs <100 mg / dL), Framingham (FHS) recurrence risk score, and very high risk (VHR) group. The risk algorithms for FHS and VHR are described herein. All of these analyses, except for the VHR analysis, included the above covariates (other than stratification coefficients, where applicable). Since this was an exploratory analysis, a p-value < 0.05 from the covariate-adjusted Cox model was considered significant.

[0102] Framingham recurrent coronary heart disease risk score. The score is based on regression coefficients of risk prediction for up to 4 years, based on age, logarithmic ratio of total cholesterol to HDL cholesterol, diabetic status, systolic blood pressure (women), and smoking status (women). The score was calculated for all ODYSSEY patients, analyzed as a continuous measure, and stratified by median score (≥median vs <median). A second risk factor analysis classified patients into very high risk (VHR) categories (as described in Roe et al., Circulation, 2019, 140, 1578-1589). VHR was classified by two sets of criteria. The first criterion (multiple previous ASCSD events) identified patients with at least one previous ischemic event, including ischemic attack, myocardial infarction, or peripheral artery disease. The second set of criteria (major previous ASCVD event + multiple high-risk conditions) includes one major ASCVD event (eligible criterion ACS event) and a history of diabetes, current smoking, being 65 years of age or older, hypertension, and ≥15 to <60 mL·min. 1 1.73m -2 We identified patients with at least two high-risk conditions, including baseline eGFR, congestive heart failure, revascularization prior to index ACS, or LDL-C ≥ 100 mg / dL using both statins and ezetimibe.

[0103] Example 2: Results Identifying patients at high risk of cardiovascular events using a polygeneic risk score. PRS for CAD was initially examined and its association with CAD prevalence was validated in two large, independent databases totaling over 530,000 individuals (DiscovEHR, n=84,243; UK Biobank, n=446,208). From these analyses, the LDPred algorithm (with an adjusted parameter ρ=0.001) was selected as the optimal PRS generation method, consistent with previous CAD trials (Figures 9-11). As a continuous score, PRS was significantly associated with MACE in DiscovEHR (OR per standard deviation (SD) of PRS = 1.4, p<0.001) and UK Biobank (OR per SD of PRS = 1.5, p<0.001) (Figures 9-11). Distribution analysis was performed to compare CAD risk across deciles. The MACE risk observed at the highest decile compared to the lowest decile was 1.9OR and 2.3OR in the DiscovEHR and UK Biobank studies, respectively (p<0.001 for each study).

[0104] Baseline characteristics of the study population and genetic risk groups. Baseline characteristics of the pharmacogenomics study population were evaluated relative to the entire study population (Figure 1). Since the pharmacogenomics group is a subset of the entire study population, p-values ​​were not calculated for these comparisons. In the genetic study, the proportion of Asian patients was smaller than that of the overall study, mainly due to differences in pharmacogenomics participation rates across study enrollment regions (Figure 8). Despite this difference, baseline medical features and lipid profiles were generally very similar across the overall study and the patients in the genetic analysis.

[0105] Demographic and baseline characteristics of high- and low-genetic subgroups were also compared to determine if there was an imbalance between the groups. At baseline, patients with high genetic risk (PRS > 90%) had numerous significant differences compared to patients with low genetic risk (PRS ≤ 90%). Individuals with high genetic risk were also younger (up to 1.8 years), had a higher incidence and likelihood of having a history of myocardial infarction, percutaneous coronary intervention, coronary artery bypass grafting, congestive heart failure (up to a benchmark event), and a family history of juvenile coronary heart disease. Patients with high genetic risk had moderately elevated baseline concentrations (approximately 2–5 mg / dL) of total cholesterol, LDL-C, non-HDL cholesterol, and apolipoprotein B. In particular, patients with high genetic risk had a median substantially elevated baseline Lp(a) level (49.4 mg / dL) compared to patients with low genetic risk (19.9 mg / dL; beta = 0.48 standard deviation units, p < 0.001). This finding is based on data from 351,224 Europeans in the UK, including those with genetic data and LP(a) levels. The findings were replicated in the Biobank, and CAD-PRS > 90% was also associated with even higher Lp(a) levels (beta = 0.39 standard deviation units, p < 0.001). While the association between baseline genetic risk score and Lp(a) was statistically very significant in this study, it should be noted that the variance ratio of serum Lp(a) levels explained by PRS was a modest 3.1%. Additional patient characteristics are shown in Figure 8.

[0106] Assessment of MACE risk by genetic risk group. Subsequently, we investigated whether PRS could identify patients at high risk of cardiovascular events in the patient population after the ACS ODYSSEY trial. PRS deciles were evaluated for the incidence of MACE and each secondary endpoint (Figures 13-18). In the placebo group, the risk of any of the endpoints was consistently higher than the overall event rate and was consistent across the ancestral group (Figure 12). In the placebo group, patients with high genetic risk (defined as the upper PRS decile) had approximately 50% higher incidence of MACE (17.0% vs. 11.4%, HR=1.59, p<0.001) and 40% higher incidence of secondary coronary heart disease events (20.4% vs. 14.6%, HR=1.55, p<0.001) compared to patients with low genetic risk (PRS<90%) (Figure 2). All analyses were adjusted for previously specified covariates. It should be noted that even lower PRS thresholds, >80th and >70th percentiles, showed statistically significant differences in MACE between high-risk and low-risk patients in the placebo group (p=0.004 and p=0.013, respectively). In a meta-analysis of the placebo and alirocumab treatment groups, the combination of continuous PRS was p=0.027; the p-values ​​for the placebo and alirocumab groups were 0.079 and 0.202, respectively.

[0107] Comparison of genetic risk for cardiovascular disease with conventional risk factors. In addition to adjusting for the baseline clinical characteristics and risk factors described above, the impact of these risk factors (LDL-C, Lp(a), and other conventional risk factors) on PRS in placebo-treated patients was further evaluated by performing risk stratification analysis. Stratification by LDL-C (divided at 100 mg / dL) showed that PRS was independent of baseline LDL-C levels (Figure 3A). Patients with high baseline LDL-C (≥100 mg / dL) and high PRS had the highest incidence of MACE at 22.7%, 95% CI (17.0–28.4), while patients with low baseline LDL-C and low genetic risk had the lowest incidence at 9.9%, 95% CI (9.0–10.8). Using both high baseline LDL-C and high PRS identifies patients at even higher risk of MACE than using either risk factor alone (Figure 3A).

[0108] These analyses were extended to a broader range of conventional risk factors (age, systolic blood pressure, smoking status, lipid levels, and type 2 diabetes) established in the Framingham Heart Trial (FHS) for recurrent coronary heart disease. Continuous PRS was associated with MACE even after adjusting for baseline FHS risk score p=0.003 (adjusting for age, sex, ancestry, and FHS score). The bifurcated PRS also showed a consistent effect across the entire FHS stratified by median score, demonstrating the independent additional value of these assessment criteria (Figure 3B). PRS was also assessed by the very high risk (VHR) category and showed a consistent effect across risk groups. As shown in Figure 19, high genetic risk was associated with increased MACE risk even in the absence of the VHR criterion (non-VHR), HR=1.84 (p=0.007). Next, the effect of Lp(a) on the association between PRS and MACE risk was investigated. The risk of Lp(a) at 50 mg / dL was divided into two groups, and a combined subgroup analysis with PRS was performed to re-emphasize the additive value of Lp(a) and PRS (Figure 3C).

[0109] Due to the strong association between baseline Lp(a) and PRS, the relationship between PRS and Lp(a) levels was further investigated. Variants within and around the LPA gene (+ / - 1 MB) were removed from PRS. This modified PRS was evaluated for its impact on Lp(a) levels and MACE risk in ODYSSEY and the UK Biobank. Removal of these variants (+ / - 1 MB) weakened the association of PRS with Lp(a) levels, and the ratio of Lp(a) variability explained by the modified PRS was nearly zero in both studies (Figures 20-21).

[0110] In the UKB, the modified PRS remained strongly associated within MACE in the UK Biobank (OR 1.5 per SD, p<0.001). In MACE risk in ODYSSEY, removing these LPA regions split the placebo group's risk in the upper decile into the top two deciles (Figures 22-23). ​​Among placebo-treated patients in the upper decile with an event, 35% (36 out of 104) moved away from the upper decile, with the majority moving to the next highest decile (suggesting that the LPA region is important but not the sole contributor to risk). Of these “moved” patients, approximately 28% (10 out of 36) had baseline serum Lp(a) < 50 mg / dL. Further evaluation of this region showed that the PRS in the LPA region had only a moderate correlation with baseline serum LP(a) levels, and r 2 It was found to be 27.7%. The results from the UK Biobank were similar, and the r between PRS in the LPA region and serum Lp(a) levels (nmol / L) 2The prevalence was 29.0%. In summary, serum Lp(a) is not a simple surrogate for MACE risk from the LPA genomic region, which is why PRS remains strongly associated with risk even after adjusting for or stratifying by baseline Lp(a). The LPA genomic region had an even stronger influence in the 27SNP score, as two of the 27 variants in the 27SNP score (rs10455872 and rs3798220) originated from the LPA region. Of the patients who received placebo in the upper decile of the 27 variant risk score, 96 had a MACE event. After excluding these two variants from the score, 56 of the 96 patients (58%) moved out of the upper decile. These and other variants originating from the LPA locus may play a significant role across a wide range of PRS scores, from small scores like the 27SNP score to larger, genome-wide PRS scores.

[0111] Evaluation of the impact on genetic risk and major cardiovascular events. Next, we investigated whether patients with high genetic risk would receive greater benefit from alirocumab treatment. Patients with high genetic risk who received alirocumab showed a significant reduction in both absolute and relative risk of MACE compared to patients with low genetic risk. In the high-gene risk group, the 3-year Kaplan-Meier cumulative incidence of MACE was 11.4% in the alirocumab group and 17.4% in the placebo group, representing a 6.0% absolute risk reduction. In the low-risk genetic group, the rates were 10.0% and 11.5%, respectively (Figure 4), representing a 1.5% absolute risk reduction. To prevent the occurrence of one primary endpoint, 17 patients (95% CI, 11–96) with high genetic risk or 64 patients (95% CI, 34–546) with low genetic risk would need to be treated for 3 years. Patients with a high genetic risk also showed a greater relative reduction in MACE with alirocumab compared to patients with a low genetic risk (HR 0.87; 95% CI: 0.78~0.98; p=0.022) (HR 0.63; 95% CI: 0.46~0.86; p=0.004). This difference was statistically significant (PRS due to treatment interaction; p=0.04) (Figure 4).

[0112] These analyses also demonstrated that patients with high genetic risk showed a greater reduction with alirocumab than those with low genetic risk in the predefined primary secondary endpoints that were significantly reduced with alirocumab across the entire trial (any cardiovascular event, any coronary heart disease event, major coronary heart disease event, and a composite endpoint of death from any cause, non-fatal myocardial infarction, or non-fatal ischemic attack) (Figure 5). Analysis of causes of death was limited by a small number of events (44 events in total) in the high genetic risk group. The overall number of patients in the high genetic risk group was smaller in those treated with alirocumab (n=20 / 584; 3.4%) compared to those treated with placebo (n=24 / 613; 3.9%), while the number of events was too small for inferential analysis.

[0113] Patients with European ancestry comprised 78% of the analyzed population, making this subgroup the largest ancestral group in the overall analysis. Therefore, subgroup analysis was performed on patients with European ancestry. The results for patients with European ancestry at high genetic risk (MACE HR 0.64; 95% CI: 0.45~0.92; p=0.016) were consistent with the overall analysis including patients of all ancestry (Figures 12 and 24).

[0114] Independent additional values ​​of PRS and pre-treatment LDL-C levels were used to predict the benefit from alirocumab. The relationship between treatment and the risk of baseline LDL-C (bifurcated at 100 mg / dL), PRS, and MACE was also investigated. In the group with high genetic risk and high baseline LDL-C, the 3-year Kaplan-Meier cumulative incidence of MACE was 22.7% in the placebo group and significantly reduced with alirocumab treatment (13.4%) (Figure 6C), corresponding to a 9.2% risk reduction (95% CI: 1.8% to 6.6%). In the group with low genetic risk and low baseline LDL-C, the rates were 9.9% and 9.2%, respectively (Figure 6C), corresponding to a 0.7% risk reduction, 95% CI (-0.6% to 2.1%). The hazard ratio for MACE (alirocumab: placebo) was lowest numerically in patients with high genetic risk and high LDL-C (≥100 mg / dL) (HR 0.55; 95% CI: 0.33~0.89; p=0.015), and highest numerically in patients with low LDL-C (<100 mg / dL) and low PRS (HR 0.94; 95% CI: 0.81~1.09; p=0.424, Figure 7). It should be noted that this difference was not statistically significant (p>0.05) when evaluating a complete Cox regression model.

[0115] The impact of alirocumab treatment on genetic risk due to VHR status. The treatment effect based on VHR status was evaluated using the aforementioned VHR criteria. In the non-VHR high-gene risk group, the absolute risk reduction associated with alirocumab was 7.3%, HR=0.26 (95% CI: 0.10-0.63), p=0.003. In the VHR* high-gene risk group, the absolute risk reduction was 5.6%, HR=0.73 (95% CI: 0.52-1.03), p=0.076. Although there were only 736 patients in the VHR* high-gene risk group, the p-value showed a relative risk reduction of 27%, which was statistically significant (Figure 28). These results suggest that patients in the high-gene risk group benefit from alirocumab treatment regardless of VHR classification.

[0116] Genetic risk and the effect of alirocumab treatment on lipid reduction. Next, the degree of lipid reduction after alirocumab treatment was examined in both patients with high and low genetic risk. The reduction in LDL-C by alirocumab was similar in both PRS groups: the median reduction at 4 months was 57.0 mg / dL in patients with high genetic risk and 58.7 mg / dL in patients with low genetic risk (Figure 25).

[0117] Because there is a strong association between baseline Lp(a) levels and genetic risk (Figure 1), the influence of genetic risk on changes in Lp(a) due to alirocumab treatment was also investigated. Patients in the high-gene risk group had a median Lp(a) reduction of 8.2 mg / dL (16.6% decrease from baseline median Lp(a) at 4 months of the trial), which was compared to a median reduction of 5.1 mg / dL (25.6% decrease from baseline median Lp(a) in the low-gene risk group) (Figure 25). In stratified analyses of high- and low-gene risk patients (Lp(a) divided at 50 mg / dL), events were significantly reduced in the high-gene risk subgroup compared to the low-gene risk subgroup in both groups (Figures 26-27). Figure 28 shows MACE stratified by genetic risk considering VHR categories and baseline Lp(a). Panel A was stratified by genetic risk, where high genetic risk was PRS > 90th percentile and low genetic risk was PRS ≤ 90th percentile. Panel B was stratified by baseline Lp(a) (Lp(a) ≥ 50 mg / dL and Lp(a) < 50 mg / dL). Panel C was stratified by both genetic risk and baseline Lp(a). These results suggest that the significant reduction in MACE observed in patients with high genetic risk is not fully explained by baseline Lp(a) or the change in Lp(a) due to alirocumab treatment.

[0118] Comparison of LDPred and 27 and 57 variant PRS models. The efficacy of alirocumab treatment was also evaluated for the selected LDPred model (ρ=0.001) and the 27 and 57 candidate PRS models. These analyses were performed using a Cox proportional hazards model and adjusted for the above covariates. The decile-based results for each of the three models are shown in Figure 30. In the 27 SNP model, the HR and p-values ​​for the high-risk and low-risk groups were HR=0.68, p=0.008, and HR=0.85, p=0.016, respectively. In the 57 SNP model, the high-risk group had an HR of 0.65 and p=0.010, while the low-risk group had an HR of 0.86 and p=0.010. These results are similar to the LDPred findings (Figure 4), with an HR of 0.63 (p=0.004) and 0.87 (p=0.022), and contrast with the UK Biobank and DiscovEHR results shown in Figures 9 and 10. In the UK Biobank, the OR for MACE at the 90th percentile was 2.33, in contrast to the OR of 1.65 for 27 SNPs. The differences between the ODYSSEY findings and these larger EHR datasets may be due to either differences in study size or population when converting primary CAD risk to evaluate the therapeutic effect of PCSK9 inhibition on recurrent events in high-risk populations. The similarity of results for high-gene risk groups in ODYSSEY is not due to strong correlations across genetic risk scores. 2 These are r 2 =31.3% and r 2 =31.4%; between 27 mutant PRS and 57 mutant PRS, r 2 The figure was 40.1%. In any case, the consistency of findings across models supports the benefit of alirocumab treatment in genetically high-risk populations.

[0119] These findings support the significant contribution of genetic risk scores to precision medicine, particularly by providing an independent, supplemental tool (which can be combined with conventional risk assessment criteria) to enhance risk assessment and predict potential benefits from treatment. Such tools can better manage limited medical resources and target patients at highest risk or those most likely to respond to treatments with limited access. In particular, this trial demonstrates that the PRS for CAD, previously developed to assess the prevalence of cardiovascular disease in large populations, is also highly useful in predicting relapse risk in high-gene risk patients in a post-ACS setting. Furthermore, this trial shows that the PRS's ability is independent and additive with many more conventional risk factors, such as LDL-C levels and FHS risk scores. Importantly, it also shows that the PRS can be combined with LDL-C levels and other risk assessment criteria to predict patients most likely to benefit from treatment (in this case, from the ability of the PCSK9 antibody alirocumab to prevent relapsed MACE).

[0120] This pharmacogenomics analysis of approximately 12,000 post-ACS patients from the ODYSSEY OUTCOMES trial showed that patients with high PRS were substantially at higher risk of recurrent MACE, despite intensive statin therapy and even after adjusting for demographic and clinical features that capture known and established risk factors for atherosclerotic cardiovascular disease. When all patients were included in the trial, the overall incidence of MACE in the placebo group was 11.1%, while the subgroup of patients with high genetic risk in the upper decile of PRS had an incidence of 17.0%. The subgroup with high baseline LDL-C (≥100 mg / dL) had an incidence of 16.1%, and the subgroup of patients with high genetic risk and high baseline LDL-C had the highest incidence of 22.7%. These data suggest that both LDL-C and PRS are independent and important risk factors for identifying post-ACS patients at highest risk of MACE despite intensive statin therapy. These analyses also demonstrated that PRS is an independent risk factor compared to the Framingham cardiac trial risk score for recurrent coronary heart disease, which is a composite score of established risk factors.

[0121] PRS also identified a group of high-gene risk patients who would have received an even greater benefit from treatment. These high-gene risk patients received a greater benefit from alirocumab treatment in terms of both absolute and relative reductions in MACE (as well as secondary endpoints) (an absolute reduction of 6.0% in MACE risk compared to 1.5% in the low-gene risk group, and a relative reduction of 37% in MACE risk compared to 13% in the low-gene risk group). Furthermore, this study demonstrates that combining PRS with conventional lipid biomarkers may not only identify individuals at highest risk of MACE, but also enable the greatest risk reduction to be achieved from treatment. Patients with high PRS and high LDL-C were not only at the highest risk (22.7%) of relapsed MACE despite intensive statin treatment, but also experienced the greatest absolute and relative risk reduction (9.2% absolute reduction in risk, 45% relative reduction in risk) with the addition of alirocumab to statin treatment; patients with high PRS and low LDL-C, or low PRS and high LDL-C, had both moderate risk and moderate benefit. Patients with low PRS and low LDL-C had the lowest risk and received only slight benefit from alirocumab treatment. These findings have clear implications in that access to such treatments should be targeted to those at the highest risk and most likely to benefit. This trial also provides evidence for the addition of alirocumab, a PCSK9 inhibitor, from a different class of lipid-lowering therapies when added to patients already receiving intensive statin treatment. These results suggest that the improvement in clinical outcomes in patients with high genetic risk is not mediated by a significant reduction in LDL-C after treatment or by high baseline levels of LDL-C, for either statins or alirocumab.

[0122] Lp(a) is recognized as a major risk factor for coronary artery disease, and Lp(a) levels are primarily genetically determined. PCSK9 inhibition is currently one of the few therapeutic approaches to lower Lp(a). In the ODYSSEY trial, alirocumab treatment reduced Lp(a) levels by 23.4% overall. At baseline, this trial observed a strong association between high genetic risk and baseline Lp(a) levels (Figure 1). Several pieces of evidence demonstrate that the greater benefit regarding MACE observed in patients with high genetic risk is not attributable to baseline Lp(a) levels or the degree of Lp(a) reduction by alirocumab alone: ​​1) even though PRS is significantly associated with elevated serum Lp(a) levels, the proportion of serum Lp(a) level variability explained by PRS is only 3.1%; 2) PRS and high genetic risk remained associated with a higher incidence of MACE and a greater reduction in MACE by alirocumab, even after adjusting for baseline Lp(a) and Lp(a) reduction; and 3) stratified analyses of patients with low and high baseline Lp(a) levels demonstrated that high genetic risk is not fully explained by Lp(a) levels, as it was equally associated with higher incidence and a greater reduction in events in the Lp(a) subgroup (Figures 26-27). In summary, these results suggest that while Lp(a) may be a strong contributing factor, it is not the sole driving mechanism for the PRS outcomes in this study.

[0123] The PRS for this trial was developed using GWAS data derived from individuals of European ancestry. This analysis in the ODYSSEY OUTCOMES trial applied this PRS to patients from all ancestral groups combined (Figure 1). Additional subgroup analyses were also performed on patients of European ancestry (Figure 24), and the results were consistent with a larger analysis including patients of all ancestry. Although the sample size was small in some ancestral groups, it was also observed that the reduction in MACE after alirocumab treatment was consistent in direction in patients with high genetic risk across all ancestral groups examined (Figure 12). As GWAS data becomes available in more diverse populations, the polygenic risk score may similarly improve over time for non-European populations as well.

[0124] In addition to those described herein, various modifications to the subject matter described herein will be apparent to those skilled in the art from the foregoing description. Such modifications are also intended to fall within the scope of the appended claims. Each reference cited in this application (including, but not limited to, journal articles, U.S. and non-U.S. patents, patent application publications, international patent application publications, gene bank acceptance numbers, etc.) is incorporated herein by reference in its entirety.

[0125] Exemplary Embodiments 1. A method for treating a patient at risk of major cardiovascular adverse events (MACE), comprising: determining the patient's polygenetic risk score (CAD-PRS), wherein the CAD-PRS comprises a weighted sum of multiple gene variants associated with coronary artery disease; identifying the patient as being at high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and, if the patient is identified as being at high risk of MACE, administering a proprotein converter subtilisin / kexin type 9 (PCSK9) inhibitor to the patient. 2. The method according to Embodiment 1, wherein the CAD-PRS threshold score is in the top 30% of the reference population. 3. The method according to Embodiment 1, wherein the CAD-PRS threshold score is the upper quintile within the reference population. 4. The method according to Embodiment 1, wherein the CAD-PRS threshold score is the upper decile in the reference population. 5. The method according to any one of Embodiments 2 to 4, wherein the reference population includes at least 100 patients. 6. The method according to any one of Embodiments 2 to 4, wherein the reference population includes at least 1,000 patients. 7. The method according to any one of Embodiments 2 to 4, wherein the reference population includes at least 5,000 patients. 8. The method according to any one of Embodiments 2 to 4, wherein the reference population includes at least 10,000 patients. 9. The method according to any one of embodiments 2 to 4, wherein the reference population is well-developed with respect to members of the ancestral group. 10. The method according to Embodiment 9, wherein the reference population is enriched with respect to members of an ancestral group selected from a group consisting of European ancestral groups, African ancestral groups, mixed American ancestral groups, East Asian ancestral groups, or South Asian ancestral groups. 11. The method according to Embodiment 9 or Embodiment 10, wherein the ancestor group is self-reported. 12. The method according to Embodiment 9 or Embodiment 10, wherein the ancestral group is derived from the main component of the ancestor. 13. The method according to Embodiment 1, wherein the gene variant is a single nucleotide polymorphism (SNP), insertion, deletion, structural variant, or copy number variant. 14. The method according to Embodiment 1, wherein the plurality of gene variants are determined by calculating the performance of gene variants in the reference population and selecting the gene variant with the most performance. 15. The method according to Embodiment 14, wherein the performance of gene variants is calculated with respect to the risk of coronary artery disease based on statistical significance, strength of association, and / or probability distribution. 16. The method according to Embodiment 15, wherein the CAD-PRS is calculated using the LDPred method. 17. The method according to Embodiment 16, wherein the proportion (ρ) of the causative marker is set to 0.001, and the plurality of gene variants include at least 6,500,000 gene variants. 18. The method according to Embodiment 15, wherein the CAD-PRS is calculated using a pruning and thresholding method. 19. The p-value threshold is 5 × 10 -8 And r 2 The method according to embodiment 18, wherein the value is 0.2. 20. The p-value threshold is 5 × 10 -2 And r 2 The method according to embodiment 18, wherein the value is 0.8. 21. The method according to Embodiment 14, wherein the plurality of gene variants comprises at least 20 gene variants. 22. The method according to Embodiment 14, wherein the plurality of gene variants comprises at least 1,000 gene variants. 23. The method according to Embodiment 14, wherein the plurality of gene variants comprises at least 10,000 gene variants. 24. The method according to Embodiment 14, wherein the plurality of gene variants comprises at least 100,000 gene variants. 25. The method according to Embodiment 14, wherein the plurality of gene variants comprises at least 1,000,000 gene variants. 26. The method according to Embodiment 14, wherein the plurality of gene variants comprises at least 6,500,000 gene variants. 27. The method according to any one of Embodiments 1 to 21, wherein the PRS is determined from a biological sample obtained from the patient, and the biological sample includes blood, semen, saliva, urine, feces, hair, teeth, bone, tissue, or cells. 28. The method according to Embodiment 27, wherein the biological sample includes blood. 29. The method according to any one of Embodiments 1 to 28, further comprising determining the patient's serum low-density lipoprotein (LDL) level and identifying the patient as being at high risk of MACE if the patient further has a serum LDL level of at least about 100 mg / dL. 30. The method according to any one of Embodiments 1 to 28, further comprising determining the level of the patient's serum lipoprotein(a) (LPA or LP(a)) and identifying the patient as being at high risk of MACE if the patient further has a serum LPA level of at least about 30 mg / dL. 31. The method according to any one of Embodiments 1 to 28, further comprising determining the level of the patient's serum lipoprotein(a) (LPA or LP(a)) and identifying the patient as being at high risk of MACE if the patient further has a serum LPA level of at least about 50 mg / dL. 32. The method according to any one of Embodiments 1 to 28, further comprising determining the patient's serum LDL level and LPA level, and identifying the patient as being at high risk of MACE if the patient further has a serum LDL level of at least about 100 mg / dL and a serum LPA level of at least about 30 mg / dL. 33. The method according to any one of Embodiments 1 to 28, further comprising determining the patient's serum LDL level and LPA level, and identifying the patient as being at high risk of MACE if the patient further has a serum LDL level of at least about 100 mg / dL and a serum LPA level of at least about 50 mg / dL. 34. The method according to any one of embodiments 1 to 33, wherein the patient had previously had MACE. 35. The method according to any one of Embodiments 1 to 34, wherein the patient has been administered high doses of statins in the past or is currently being administered high doses. 36. The method according to any one of Embodiments 1 to 35, wherein the PCSK9 inhibitor is alirocumab. 37. The method according to any one of Embodiments 1 to 35, wherein the PCSK9 inhibitor is evolocumab or vococizumab. 38. The method according to any one of Embodiments 1 to 37, wherein MACE includes coronary artery disease (CAD), myocardial infarction (MI), unstable angina, ischemic attack, coronary artery regeneration due to ischemia, arrhythmia, cardiovascular death, valvular heart disease, cardiomyopathy, or congestive heart failure. 39. A method for lowering serum LDL levels in a patient at high risk of major cardiovascular adverse events (MACE), comprising: determining the patient's polygenic risk score (CAD-PRS), wherein the CAD-PRS includes a weighted sum of multiple gene variants associated with coronary artery disease; identifying the patient as being at high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and, if the patient is identified as being at high risk of MACE, administering to the subject a proprotein converter subtilisin / kexin type 9 (PCSK9) inhibitor in an amount effective in lowering the patient's serum LDL levels. 40. The method according to embodiment 39, wherein the CAD-PRS threshold score is in the top 30% of the reference population. 41. The method according to Embodiment 39, wherein the CAD-PRS threshold score is the upper quintile in the reference population. 42. The method according to Embodiment 39, wherein the CAD-PRS threshold score is the upper decile in the reference population. 43. The method according to any one of embodiments 39 to 42, wherein the reference population includes at least 1,000 patients. 44. The method according to any one of embodiments 39 to 42, wherein the reference population includes at least 5,000 patients. 45. The method according to any one of embodiments 39 to 42, wherein the reference population includes at least 10,000 patients. 46. ​​The method according to any one of embodiments 39 to 42, wherein the reference population is enriched with respect to members of the ancestral group. 47. The method according to Embodiment 46, wherein the reference population is enriched with respect to members of an ancestral group selected from a group consisting of European ancestral groups, African ancestral groups, mixed American ancestral groups, East Asian ancestral groups, or South Asian ancestral groups. 48. The method of Embodiment 46 or Embodiment 47, wherein the ancestral group is self-reported. 49. The method according to Embodiment 46 or Embodiment 47, wherein the ancestral group is derived from the main component of the ancestor. 50. The method according to Embodiment 39, wherein the gene variant is a single nucleotide polymorphism (SNP), insertion, deletion, structural variant, or copy number variant. 51. The method according to Embodiment 39, wherein the plurality of gene variants are determined by calculating the performance of gene variants in the reference population and selecting the gene variant with the most performance. 52. The method according to Embodiment 51, wherein the performance of gene variants is calculated with respect to the risk of coronary artery disease based on statistical significance, strength of association, and / or probability distribution. 53. The method according to Embodiment 52, wherein the CAD-PRS is calculated using the LDPred method. 54. The method according to Embodiment 53, wherein the proportion (ρ) of the causative marker is set to 0.001, and the plurality of gene variants include at least 6,500,000 gene variants. 55. The method according to Embodiment 52, wherein the CAD-PRS is calculated using a pruning and thresholding method. 56. The p-value threshold is 5 × 10 -8 And r 2 The method according to embodiment 55, wherein the value is 0.2. 57. The p-value threshold is 5 × 10 -2 And r 2 The method according to embodiment 55, wherein the value is 0.8. 58. The method according to Embodiment 51, wherein the plurality of gene variants comprises at least 20 gene variants. 59. The method according to Embodiment 51, wherein the plurality of gene variants comprises at least 1,000 gene variants. 60. The method according to Embodiment 51, wherein the plurality of gene variants comprises at least 10,000 gene variants. 61. The method according to Embodiment 51, wherein the plurality of gene variants comprises at least 100,000 gene variants. 62. The method according to Embodiment 51, wherein the plurality of gene variants comprises at least 1,000,000 gene variants. 63. The method according to Embodiment 51, wherein the plurality of gene variants comprises at least 6,500,000 gene variants. 64. The method according to any one of Embodiments 39 to 57, wherein the PRS is determined from a biological sample obtained from the patient, and the biological sample includes blood, semen, saliva, urine, feces, hair, teeth, bone, tissue, or cells. 65. The method according to embodiment 64, wherein the biological sample includes blood. 66. The method according to any one of embodiments 39 to 65, further comprising determining the patient's serum low-density lipoprotein (LDL) level and identifying the patient as being at high risk of MACE if the patient further has a serum LDL level of at least about 100 mg / dL. 67. The method according to any one of embodiments 39 to 65, further comprising determining the level of the patient's serum lipoprotein(a) (LPA or LP(a)) and identifying the patient as being at high risk of MACE if the patient further has a serum LPA level of at least about 30 mg / dL. 68. The method according to any one of embodiments 39 to 65, further comprising determining the level of the patient's serum lipoprotein(a) (LPA or LP(a)) and identifying the patient as being at high risk of MACE if the patient further has a serum LPA level of at least about 50 mg / dL. 69. The method according to any one of embodiments 39 to 65, further comprising determining the patient's serum LDL level and LPA level, and identifying the patient as being at high risk of MACE if the patient further has a serum LDL level of at least about 100 mg / dL and a serum LPA level of at least about 30 mg / dL. 70. The method according to any one of embodiments 39 to 65, further comprising determining the patient's serum LDL level and LPA level, and identifying the patient as being at high risk of MACE if the patient further has a serum LDL level of at least about 100 mg / dL and a serum LPA level of at least about 50 mg / dL. 71. The method according to any one of embodiments 39 to 70, wherein the patient had previously had MACE. 72. The method according to any one of embodiments 39 to 71, wherein the patient has been administered high doses of statins in the past or is currently being administered high doses of statins. 73. The method according to any one of embodiments 39 to 72, wherein the PCSK9 inhibitor is alirocumab. 74. The method according to any one of embodiments 39 to 72, wherein the PCSK9 inhibitor is evolocumab. 75. The method according to any one of embodiments 39 to 74, wherein MACE includes coronary artery disease (CAD), myocardial infarction (MI), unstable angina, ischemic attack, ischemic coronary artery regeneration, arrhythmia, cardiovascular death, valvular heart disease, cardiomyopathy, or congestive heart failure. 76. A method for reducing lipoprotein(a) (LPA or LP(a)) levels in patients at high risk of major cardiovascular adverse events (MACE), comprising: determining the patient's polygenic risk score (CAD-PRS), wherein the CAD-PRS includes a weighted sum of multiple gene variants associated with coronary artery disease; identifying the patient as being at high risk of MACE if the patient has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; and administering to the subject a proprotein converterse subtilisin / kexin type 9 (PCSK9) inhibitor in an amount effective in reducing the patient's lipoprotein(a) levels if the patient has been identified as being at high risk of MACE. 77. The method according to embodiment 76, wherein the CAD-PRS threshold score is in the top 30% of the reference population. 78. The method according to embodiment 76, wherein the CAD-PRS threshold score is the upper quintile in the reference population. 79. The method according to embodiment 76, wherein the CAD-PRS threshold score is the upper decile in the reference population. 80. The method according to any one of embodiments 77 to 79, wherein the reference population includes at least 1,000 patients. 81. The method according to any one of embodiments 77 to 79, wherein the reference population includes at least 5,000 patients. 82. The method according to any one of embodiments 77 to 79, wherein the reference population includes at least 10,000 patients. 83. The method according to any one of embodiments 77 to 79, wherein the reference population is enriched with respect to members of the ancestral group. 84. The method according to Embodiment 83, wherein the reference population is enriched with respect to members of an ancestral group selected from a group consisting of European ancestral groups, African ancestral groups, mixed American ancestral groups, East Asian ancestral groups, or South Asian ancestral groups. 85. The method according to Embodiment 83 or Embodiment 84, wherein the ancestral group is self-reported. 86. The method according to Embodiment 82 or Embodiment 83, wherein the ancestor group is derived from the main component of the ancestor. 87. The method according to Embodiment 76, wherein the gene variant is a single nucleotide polymorphism (SNP), insertion, deletion, structural variant, or copy number variant. 88. The method according to Embodiment 76, wherein the plurality of gene variants are determined by calculating the performance of gene variants in the reference population and selecting the gene variant with the most performance. 89. The method according to Embodiment 88, wherein the performance of gene variants is calculated with respect to the risk of coronary artery disease based on statistical significance, strength of association, and / or probability distribution. 90. The method according to embodiment 89, wherein the CAD-PRS is calculated using the LDPred method. 91. The method according to Embodiment 90, wherein the proportion (ρ) of the causative marker is set to 0.001, and the plurality of gene variants include at least 6,500,000 gene variants. 92. The method according to Embodiment 89, wherein the CAD-PRS is calculated using a pruning and thresholding method. 93. The p-value threshold is 5 × 10 -8 And r 2 The method according to embodiment 92, wherein the value is 0.2. 94. The p-value threshold is 5 × 10-2 And r 2 The method according to embodiment 92, wherein the value is 0.8. 95. The method according to Embodiment 88, wherein the plurality of gene variants comprises at least 70 gene variants. 96. The method according to Embodiment 88, wherein the plurality of gene variants comprises at least 1,000 gene variants. 97. The method according to Embodiment 88, wherein the plurality of gene variants comprises at least 10,000 gene variants. 98. The method according to Embodiment 88, wherein the plurality of gene variants comprises at least 100,000 gene variants. 99. The method according to Embodiment 88, wherein the plurality of gene variants comprises at least 1,000,000 gene variants. 100. The method according to Embodiment 88, wherein the plurality of gene variants comprises at least 6,500,000 gene variants. 101. The method according to any one of Embodiments 76 to 94, wherein the PRS is determined from a biological sample obtained from the patient, and the biological sample includes blood, semen, saliva, urine, feces, hair, teeth, bone, tissue, or cells. 102. The method according to Embodiment 101, wherein the biological sample includes blood. 103. The method according to any one of embodiments 76 to 102, wherein the patient had previously had MACE. 104. The method according to any one of embodiments 76 to 103, wherein the patient has been administered high doses of statins in the past or is currently being administered high doses of statins. 105. The method according to any one of embodiments 76 to 104, wherein the PCSK9 inhibitor is alirocumab. 106. The method according to any one of embodiments 76 to 104, wherein the PCSK9 inhibitor is evolocumab. 107. The method according to any one of embodiments 76 to 106, wherein MACE includes coronary artery disease (CAD), myocardial infarction (MI), unstable angina, ischemic attack, ischemic coronary artery regeneration, arrhythmia, cardiovascular death, valvular heart disease, cardiomyopathy, or congestive heart failure. 108. The method according to any one of embodiments 1 to 107, further comprising determining a composite risk score including the PRS and LPA levels in the patient. 109. The method according to any one of embodiments 1 to 107, further comprising determining a composite risk score including the PRS and the LDL level in the patient. 110. The method according to any one of embodiments 1 to 107, further comprising determining a composite risk score including the PRS, the LPA level and the LDL level in the patient. 111. A method for screening candidates for inclusion in a clinical trial for the treatment of a cardiovascular condition, the method comprising: determining the CAD-PRS of a candidate, wherein the CAD-PRS comprises a weighted sum of multiple gene variants associated with coronary artery disease; including the candidate in the clinical trial if the candidate has a CAD-PRS greater than a threshold CAD-PRS determined from a reference population; or excluding the candidate from the clinical trial if the candidate has a CAD-PRS less than a threshold CAD-PRS determined from a reference population. 112. The method according to embodiment 111, wherein the CAD-PRS threshold score is in the top 30% of the reference population. 113. The method according to Embodiment 111, wherein the CAD-PRS threshold score is the upper quintile in the reference population. 114. The method according to Embodiment 111, wherein the CAD-PRS threshold score is the upper decile in the reference population. 115. The method according to any one of embodiments 112 to 114, wherein the reference population includes at least 1,000 patients. 116. The method according to any one of embodiments 112 to 114, wherein the reference population includes at least 5,000 patients. 117. The method according to any one of embodiments 112 to 114, wherein the reference population includes at least 10,000 patients. 118. The method according to any one of embodiments 112 to 114, wherein the reference population is enriched with respect to members of the ancestral group. 119. The method according to Embodiment 118, wherein the reference population is enriched with respect to members of an ancestral group selected from the group consisting of European ancestral groups, African ancestral groups, mixed American ancestral groups, East Asian ancestral groups, or South Asian ancestral groups. 120. The method according to Embodiment 118 or Embodiment 119, wherein the ancestral group is self-reported. 121. The method according to Embodiment 118 or Embodiment 119, wherein the ancestral group is derived from the principal component of the ancestor. 122. The method according to Embodiment 111, wherein the gene variant is a single nucleotide polymorphism (SNP), insertion, deletion, structural variant, or copy number variant. 123. The method according to Embodiment 111, wherein the plurality of gene variants are determined by calculating the performance of gene variants in the reference population and selecting the gene variant with the most performance. 124. The method according to Embodiment 123, wherein the performance of gene variants is calculated with respect to the risk of coronary artery disease based on statistical significance, strength of association, and / or probability distribution. 125. The method according to Embodiment 124, wherein the CAD-PRS is calculated using the LDPred method. 126. The method according to Embodiment 125, wherein the proportion (ρ) of the causative marker is set to 0.001, and the plurality of gene variants include at least 6,500,000 gene variants. 127. The method according to Embodiment 124, wherein the CAD-PRS is calculated using a pruning and thresholding method. 128. The p-value threshold is 5 × 10 -8 And r 2 The method according to embodiment 127, wherein the value is 0.2. 129. The p-value threshold is 5 × 10 -2 And r 2 The method according to embodiment 127, wherein the value is 0.8. 130. The method according to Embodiment 123, wherein the plurality of gene variants comprises at least 70 gene variants. 131. The method according to Embodiment 123, wherein the plurality of gene variants comprises at least 1,000 gene variants. 132. The method according to Embodiment 123, wherein the plurality of gene variants comprises at least 10,000 gene variants. 133. The method according to Embodiment 123, wherein the plurality of gene variants comprises at least 100,000 gene variants. 134. The method according to Embodiment 123, wherein the plurality of gene variants comprises at least 1,000,000 gene variants. 135. The method according to Embodiment 123, wherein the plurality of gene variants comprises at least 6,500,000 gene variants. 136. The method according to any one of Embodiments 111 to 129, wherein the PRS is determined from a biological sample obtained from the patient, and the biological sample includes blood, semen, saliva, urine, feces, hair, teeth, bone, tissue, or cells. 137. The method according to embodiment 136, wherein the biological sample includes blood. 138. The method according to any one of embodiments 111 to 137, wherein the patient had previously had MACE. 139. The method according to any one of embodiments 111 to 138, wherein the patient has been administered high doses of statins in the past or is currently being administered high doses. 140. The method according to any one of Embodiments 111 to 139, wherein MACE includes coronary artery disease (CAD), myocardial infarction (MI), unstable angina, ischemic attack, coronary artery regeneration due to ischemia, arrhythmia, cardiovascular death, valvular heart disease, cardiomyopathy, or congestive heart failure. 141. The method according to any one of embodiments 111 to 140, further comprising determining a composite risk score including the PRS and LPA levels in the patient. 142. The method according to any one of embodiments 111 to 140, further comprising determining a composite risk score including the PRS and the LDL level in the patient. 143. The method according to any one of embodiments 111 to 140, further comprising determining a composite risk score including the PRS, the LPA level and the LDL level in the patient. [Table 1-1] [Table 1-2] [Table 1-3] [Table 1-4] [Table 1-5]

Claims

[Claim 1] A method for treating patients at risk of major cardiovascular adverse events (MACEs), To determine the CAD-PRS of the patient, which includes a weighted sum of multiple gene variants associated with coronary artery disease; If a patient has a CAD-PRS higher than the threshold CAD-PRS determined from the reference population, the patient is identified as being at high risk for MACE; The method comprising administering a proprotein converter subtilisin / kexin type 9 (PCSK9) inhibitor to the patient if the patient is identified as being at high risk of MACE.

Images