Real-World and Subgroup Data Synthesis to Guide PCSK9 Inhibitor Use in Cardiovascular Disease

Real-World and Subgroup Data Synthesis to Guide PCSK9 Inhibitor Use in Cardiovascular Disease | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Real-World and Subgroup Data Synthesis to Guide PCSK9 Inhibitor Use in Cardiovascular Disease Lingyao Sun, Dachuang Zhou, Hongfei Hu, Weiwei Ding, Yimei Zhong, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7274417/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Randomized controlled trials (RCTs) are the gold standard for assessing the efficacy of medical interventions; however, their findings often do not fully translate into real-world populations due to differences in patient characteristics. In this study, we aimed to estimate real-world populations treatment effects for PCSK9 inhibitors in patients with atherosclerotic cardiovascular disease (ASCVD) by integrating real-world data with subgroup-level data from two major RCTs, FOURIER and ODYSSEY LONG TERM. This approach seeks to improve precision medicine by aligning trial results with real-world data. Methods We combined individual-level data from six large population-based studies (N = 872,550) with subgroup data from the FOURIER and ODYSSEY LONG TERM trials. A meta-interpolation method was applied to estimate treatment effects for two PCSK9 inhibitors, evolocumab and alirocumab, using regression modeling to adjust for covariates such as age, sex, race, and cardiovascular history. Hazard ratios (HRs) for major adverse cardiovascular events were estimated, and sensitivity analyses were conducted to test the robustness of the findings. Results In a U.S.-representative population, evolocumab demonstrated consistent efficacy across subgroups with an HR of 0.73 (95% confidence interval [CI], 0.65–0.82). In contrast, alirocumab showed variable efficacy, with a significant difference in treatment effects between men (HR, 0.71) and women (HR, 0.89; P < 0.001). Sensitivity analyses confirmed the robustness of these findings across different covariate balance thresholds and data assumptions. Conclusions Integrating real-world data with RCT subgroup data via meta-interpolation offers a reliable method to estimate treatment effects tailored to real-world populations. Our results suggest that evolocumab provides consistent benefit across various patient subgroups, while alirocumab’s efficacy may vary based on factors such as sex and comorbidities. This approach can guide more personalized treatment decisions in ASCVD, advancing the field of precision cardiovascular pharmacotherapy. Trial Registration: Not applicable. PCSK9 inhibitors real-world evidence meta-interpolation cardiovascular disease precision medicine Figures Figure 1 Figure 2 Introduction Randomized clinical trials (RCTs) remain the cornerstone for evaluating drug efficacy due to their robust internal validity 1 . However, limitations persist in their application to real-world clinical decision-making 2 , 3 . Among these challenges are the lack of head-to-head trials comparing competing interventions and the frequent unavailability of individual patient data (IPD) 4 . These gaps are particularly pronounced in the field of cardiovascular disease, where evaluating long-term outcomes like major adverse cardiovascular events (MACE) often requires large-scale and prolonged follow-up studies 5 , 6 . The cost, ethical challenges, and feasibility constraints of conducting such trials often leave decision-makers reliant on indirect comparisons to guide treatment decisions 2 , 3 , 7 , 8 . To address these gaps, researchers and health technology assessment agencies have increasingly turned to advanced evidence synthesis techniques, including network meta-analysis (NMA), matching-adjusted indirect comparisons (MAIC), and simulated treatment comparisons (STC) 9 – 15 . While these methods provide valuable insights, their effectiveness is contingent on specific conditions. NMA requires a well-connected network of trials with consistent populations 16 ; MAIC and STC rely on access to IPD and assume that treatment effect modifiers are shared across studies 17 , 18 . When these assumptions are violated or data are unavailable—as is often the case in real-world settings—the utility of these approaches diminishes. In this context, we introduced and evaluated a novel method termed meta-interpolation 19 . This approach aimed to extend conventional meta-analytic strategies by integrating real-world data (RWD) and subgroup summary data from randomized trials to estimate individualized treatment effects without requiring IPD. Meta-interpolation leveraged the distributional structure of patient covariates observed in large-scale population datasets to impute missing subgroup characteristics in trial data. Using these estimates, we applied regression modeling to predict treatment effects tailored to specific patient populations. In this study, we applied the meta-interpolation framework to assess and compare the effectiveness of two PCSK9 inhibitors—evolocumab and alirocumab—in patients with atherosclerotic cardiovascular disease (ASCVD) and elevated low-density lipoprotein cholesterol (LDL-C). By integrating RWD from over 870,000 individuals with trial subgroup summaries from the FOURIER and ODYSSEY LONG TERM studies, we demonstrated how meta-interpolation could generate clinically relevant, covariate-adjusted hazard ratios in the absence of IPD 20 , 21 . These results may help inform more precise treatment decisions for ASCVD patients, promoting evidence synthesis and advancing precision therapeutics. Methods Study Design and Data Sources This comparative-effectiveness modeling study integrates subgroup-level trial data with real-world population datasets to estimate covariate-adjusted treatment effects for two lipid-lowering therapies (Fig. 1 ). The analysis was conducted between September 2024 and June 2025. Ethical approval was not required, as only deidentified, publicly available data were used. A systematic review identified two pivotal randomized controlled trials—FOURIER (N = 27,564) and ODYSSEY LONG TERM (N = 18,924)—as the sources for trial-based efficacy and subgroup summaries (eMethods in the Supplement). Both studies evaluated PCSK9 inhibitors in adults with established ASCVD and LDL-C ≥ 70 mg/dL, using a randomized, double-blind, placebo-controlled design (eTable 5 in the Supplement). Subgroup hazard ratios and covariate means were extracted from published results and supplementary materials 22 – 24 . Real-world covariate distributions were derived from six population-based studies: the Survey of Health, Ageing and Retirement in Europe (SHARE), the Health and Retirement Study (HRS), the English Longitudinal Study of Ageing (ELSA), the China Health and Retirement Longitudinal Study (CHARLS), the Longitudinal Aging Study in India (LASI), and the Mexican Health and Aging Study (MHAS). Collectively, these studies comprised 872,550 individual records spanning multiple continents and time periods. After deduplication and eligibility screening (history of ASCVD and hypercholesterolemia), 27,413 unique records remained for analysis (eFigure 2 in the Supplement). Target Population and Covariate Selection The target population for this study consists of the patient cohort from the 2017–2018 National Health and Nutrition Examination Survey (NHANES) database in the United States, with the objective of more accurately simulating the effects of the drug in real-world patients (eFigure 3 in the Supplement). Based on expert interviews and the availability of data, the analysis incorporated seven prespecified covariates: age ≥ 65 years, sex, race (White vs non-White), history of stroke, history of myocardial infarction (MI), diabetes status, and elevated LDL-C 25 – 27 (eTable 6 in the Supplement). All covariates were encoded as binary variables to match the format of subgroup summaries from the trials. LDL-C was available only in selected datasets (e.g., CHARLS and NHANES) and was used selectively in primary and sensitivity analyses. Real-World Matching to Trial Cohorts To ensure comparability between the real-world dataset and trial populations, we implemented a greedy matching algorithm based on a least-squares minimization of covariate means. For each trial, we defined a "target mean" vector for the seven covariates using available trial data. From the full real-world sample, we selected a subset whose covariate means closely matched the target means within a ± 10% tolerance margin (eMethods in the Supplement). The algorithm iteratively removed observations that most reduced the total squared distance between the real-world and trial covariate vectors until optimal balance was achieved. BLUP Imputation of Subgroup Covariates Subgroup summaries from trials typically report hazard ratios stratified by only a subset of covariates. To estimate treatment effects for unreported strata, we applied best linear unbiased prediction (BLUP) using real-world correlation matrices 28 . The imputed proportion of an unreported subgroup was calculated as: $$\:\hat {{p}}_{u}={\overline{p}}_{u}+{{\Sigma\:}}_{ut}{{\Sigma\:}}_{tt}^{-1}({p}_{t}-{\overline{p}}_{t})$$ Here, \(\:{p}_{t}\:\) represents the reported subgroup proportions, \(\:{\overline{p}}_{t}\:\) and \(\:{\overline{p}}_{u}\:\) are real-world means, and \(\:\:{\Sigma\:}\) is the estimated covariance matrix of all covariates from the matched real-world cohort. This allowed us to infer the joint distribution of patient characteristics in each trial subgroup without access to IPD. Estimation of Covariate-Adjusted Treatment Effects Using imputed subgroup-level covariate distributions and their associated hazard ratios, we fitted a population-level linear regression model with a log link function to approximate treatment effects. The model took the form: $$\:\text{l}\text{o}\text{g}\left(H{R}_{i}\right)={\beta\:}_{0}+\sum\:_{j=1}^{k}{\beta\:}_{j}{X}_{ij}+{\epsilon\:}_{i}$$ where \(\:H{R}_{i}\) denotes the subgroup hazard ratio and \(\:{X}_{ij}\:\) the mean value of covariate \(\:j\:\) in subgroup \(\:i\) . Regression coefficients \(\:{\beta\:}_{j}\:\) estimate the marginal effect modification by each covariate. We used these coefficients to predict treatment effects in the NHANES population (N = 495) by plugging in the average covariate profile. Sensitivity Analyses We conducted multiple sensitivity analyses to assess the robustness of our findings. These included: (1) repeating the matching procedure using a ± 5% and ± 15% covariate balance threshold; (2) substituting LDL-C correlations from CHARLS to evaluate geographic sensitivity. Full results are reported in the Supplement. Results Study Population Characteristics Compared to the combined real-world cohort (N = 27,413), which had higher proportions of older adults (68.8% aged ≥ 65), women (51.4%), and individuals with prior stroke (28.2%), the matched cohorts more closely reflected the demographic and clinical characteristics of their respective trials. After matching, the final real-world cohorts comprised 11,682 participants aligned with the FOURIER trial population (evolocumab) and 7,189 participants aligned with the ODYSSEY LONG TERM trial population (alirocumab). In the matched cohorts, 44.0% and 26.9% of participants were aged ≥ 65 years, 78.3% and 74.8% were male, and 85.0% and 79.4% identified as White, respectively. The prevalence of prior stroke was 12.0% in the FOURIER-matched group and 5.0% in the ODYSSEY-matched group, while prior myocardial infarction was present in 69.0% and 83.0%, respectively. Diabetes was reported in 25.0% of the FOURIER-matched group and 28.8% of the ODYSSEY-matched group. Compared to the NHANES population (N = 495), which had a higher proportion of older adults (57.6% aged ≥ 65), women (41.6%), non-White individuals (51.3%), and patients with prior stroke (43.4%) or diabetes (39.4%), the matched cohorts were younger, more predominantly male and White, and had a lower burden of cerebrovascular and metabolic comorbidities (Table 1 ). Table 1 Baseline characteristics of the population Variable Variable as measured in study Combined cohort(N = 27413) FOURIER (N = 27,564) ODYSSEY LONG TERM (N = 18924) NHANES(N = 495) Target means Matched means(N = 11682) Target means Matched means(N = 7189) Age Age ≥ 65: n (%) 68.84% 44.00% 46.27% 26.90% 26.92% 57.58% Gender Male: n (%) 48.63% 78.30% 77.20% 74.80% 74.79% 58.38% Race White: n (%) 76.83% 85.00% 82.79% 79.40% 79.37% 48.69% CVD Stroke: n (%) 28.23% 12.00% 13.20% 5.00% 5.49% 43.43% MI: n (%) 61.62% 69.00% 71.43% 83.00% 83.17% 41.62% Metabolic Diabetes: n (%) 32.86% 25.00% 25.22% 28.80% 28.81% 39.39% LDL-C Higher: n (%) NA 24.80% NA 29.80% NA 38.38% CVD: Cardiovascular Disease; MI: Myocardial Infarction; LDL-C: Low-Density Lipoprotein Cholesterol. Covariate Balance and Imputation Accuracy All seven covariates in the matched samples were balanced within ± 7% absolute difference from the original trial-reported subgroup means (eTable 6 in the Supplement). Correlation matrices used for BLUP revealed weak-to-moderate associations between age, cardiovascular history, and lipid levels. The highest observed correlation was between age and male sex (r = 0.28) and between prior MI and diabetes (r = 0.25) (eFigure 4 in the Supplement). These matrices served as the foundation for imputing missing subgroup covariate distributions in the trial data (eTable 7 in the Supplement). Treatment Effects in the NHANES Population Using covariate profiles from the NHANES cohort, meta-interpolation yielded an adjusted HR of 0.73 (95% CI, 0.65–0.82) for evolocumab and 0.78 (95% CI, 0.76–0.80) for alirocumab (Fig. 2 ). These estimates suggest greater real-world effectiveness than those reported in the original trials (HR = 0.85 for both agents). The improvement in apparent effectiveness may reflect adjustment to the older, higher-risk NHANES population, which included more individuals with prior stroke, MI, or elevated LDL-C. Subgroup-Specific Treatment Heterogeneity Subgroup analyses in the NHANES population showed generally stronger treatment effects for both drugs across age, sex, race, stroke history, metabolic status, and baseline LDL-C levels. Alirocumab showed similar efficacy in the MI subgroup between RCT and NHANES populations (HR, 0.89 vs 0.89), but was more effective in the non-MI subgroup in the RCT (HR, 0.70 vs 0.71). In RCT data, evolocumab showed greater benefit in patients < 65 years, without MI, and with diabetes. Alirocumab was more effective in male patients. In the NHANES cohort, evolocumab showed enhanced efficacy in patients ≥ 65 years and those with MI, while its effect was consistent across diabetes subgroups. Alirocumab showed similar efficacy in men and women. Covariate Influence on Treatment Outcomes Evolocumab showed consistent efficacy, with no significant modification by age, sex, metabolic status, or LDL-C level (eTable 8 in the Supplement). Modest variation was observed by race (coefficient, 0.23; P = .049) and stroke history (–0.26; P = .044). In contrast, alirocumab demonstrated greater heterogeneity in treatment effect, with stronger efficacy associated with older age (–0.20; P < .001), male sex (–0.05; P = .013), White race (–0.17; P < .001), higher LDL-C (–0.24; P < .001), and presence of metabolic abnormalities (–0.08; P = .018). Reduced efficacy was observed in patients with prior stroke (0.20; P < .001) and myocardial infarction (0.34; P < .001). Sensitivity Analyses Across all sensitivity analyses—including tighter (± 5%) and looser (± 15%) covariate-balance thresholds and use of LDL-C correlations from CHARLS—the HRs for both drugs remained stable (eTable 9–14 and eFigure 5–10 in the Supplement). For evolocumab, HRs ranged from 0.72 to 0.73; for alirocumab, from 0.78 to 0.80 (Table 2 ). These findings support the robustness of the meta-interpolation framework across varying data assumptions. Table 2 Hazard Ratios for Evolocumab and Alirocumab Under Sensitivity Analyses Using Meta-Interpolation Intervention Basic analysis HR (95% Cl) a ± 5% covariate balance threshold HR (95% Cl) b ± 15% covariate balance threshold HR (95% Cl) b LDL-C correlations from CHARLS HR (95% Cl) c Evolocumab 0.73 (0.65–0.82) 0.73(0.66–0.81) 0.72(0.63–0.81) 0.73(0.65–0.81) Alirocumab 0.78 (0.76–0.80) 0.80(0.78–0.82) 0.80(0.78–0.82) 0.80(0.78–0.82) HR, hazard ratio; CI, confidence interval; LDL-C, low-density lipoprotein cholesterol; CHARLS, China Health and Retirement Longitudinal Study; NHANES, National Health and Nutrition Examination Survey. a The basic analysis represents the primary covariate-adjusted estimate in the NHANES-matched population. b Sensitivity analyses were conducted by applying stricter (± 5%) and more lenient (± 15%) covariate balance thresholds during the matching procedure. c Sensitivity analyses were conducted by using LDL-C correlation structures derived from CHARLS. Discussion In this study, we introduced a novel evidence synthesis framework—meta-interpolation—to estimate the comparative effectiveness of evolocumab and alirocumab in a real-world population with ASCVD. By integrating subgroup-level data from RCTs with covariate distributions derived from large-scale observational datasets, we were able to estimate adjusted treatment effects in the absence of IPD. This approach has both immediate clinical implications for lipid-lowering therapy and broader applications for evidence generation in cardiovascular pharmacotherapy. The comparative findings suggest that evolocumab offers more consistent efficacy across key clinical subgroups, including by age, sex, metabolic status, and prior cardiovascular history. In contrast, alirocumab's efficacy was more variable, with stronger effects observed in younger, male, White patients with elevated LDL-C and no history of stroke or myocardial infarction. These subgroup-specific differences align with prior observations of effect modification in cardiovascular outcomes trials, where patient characteristics influence both absolute and relative risk reductions 20 , 21 . In clinical practice, such heterogeneity may support more personalized drug selection. For instance, in older adults or patients with a history of stroke—populations often underrepresented in RCTs—evolocumab may be a more reliable option due to its stable performance. In contrast, alirocumab may be preferentially considered in male patients with very high LDL-C levels and fewer comorbidities. These insights may be particularly useful in formulary decision-making, shared decision-making, or population-level treatment allocation under resource constraints 29 . Importantly, our results are not intended to favor one drug over another universally, but rather to highlight the need for context-sensitive prescribing based on patient-specific factors. This aligns with the growing emphasis on precision medicine in cardiovascular prevention, where one-size-fits-all recommendations are increasingly being replaced by risk-stratified approaches 30 . Beyond its application to PCSK9 inhibitors, the meta-interpolation method provides a flexible and scalable tool for comparative effectiveness research in settings where IPD is unavailable, head-to-head trials are lacking, or effect modification is suspected. Traditional methods such as MAIC or STC often require restrictive assumptions about shared effect modification or depend on IPD from at least one comparator trial 31 , 32 . In contrast, meta-interpolation can leverage publicly available RCT subgroup data and population-level real-world data to estimate adjusted treatment effects with less reliance on proprietary or inaccessible sources. This approach is particularly relevant as novel cardiovascular therapies continue to expand, including SGLT2 inhibitors, GLP-1 receptor agonists, siRNA-based therapies (e.g., inclisiran), and anti-inflammatory agents (e.g., colchicine, ziltivekimab). These agents are often evaluated in specialized trial populations but deployed across a much broader spectrum of real-world patients, creating an urgent need for methods that can translate trial results into practice 33 , 34 . In addition, meta-interpolation may support HTA, reimbursement modeling, and population-based guideline development, especially in regions or systems where real-world clinical registries are fragmented but survey-based demographic and clinical data are available. Its compatibility with common observational sources like NHANES, CHARLS, or SHARE enhances its utility in global settings. Limitations Several limitations should be noted. First, the accuracy of estimated treatment effects depends on the granularity and quality of subgroup data reported in trials. Most trial publications only report limited subgroup categories and may lack joint distributions, limiting model resolution. Second, the BLUP-based interpolation relies on assumed linear relationships among covariates, which may not fully capture higher-order interactions or nonlinearities. Third, LDL-C values were not directly available in several real-world datasets, requiring assumptions based on external correlation structures, which may introduce imprecision. Finally, the modest sample size in NHANES (n = 495) constrains subgroup-level estimates and may reduce generalizability to more diverse clinical populations. Conclusion This study demonstrates that meta-interpolation is a robust and pragmatic method to estimate adjusted treatment effects when IPD is unavailable. Applied to PCSK9 inhibitors, this framework identified clinically relevant differences in subgroup efficacy that can inform individualized prescribing. More broadly, it offers a scalable strategy to extend trial-based evidence to real-world populations, supporting evidence-informed decisions in cardiovascular care and beyond. Abbreviations RCTs Randomized Controlled Trials PCSK9 Proprotein Convertase Subtilisin/Kexin Type 9 ASCVD Atherosclerotic Cardiovascular Disease HR Hazard Ratio CI Confidence Interval IPD Individual Patient Data MACE Major Adverse Cardiovascular Events NMA Network Meta-Analysis MAIC Matching-Adjusted Indirect Comparisons STC Simulated Treatment Comparisons RWD Real-World Data BLUP Best Linear Unbiased Prediction LDL-C Low-Density Lipoprotein Cholesterol SHARE Survey of Health, Ageing and Retirement in Europe HRS Health and Retirement Study ELSA English Longitudinal Study of Ageing CHARLS China Health and Retirement Longitudinal Study LASI Longitudinal Aging Study in India MHAS Mexican Health and Aging Study NHANES National Health and Nutrition Examination Survey CVDs Cardiovascular Diseases MI Myocardial Infarction SGLT2 Sodium-Glucose Cotransporter-2 GLP-1 Glucagon-Like Peptide-1 siRNA Small Interfering RNA HTA Health Technology Assessment Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable Competing interests The authors declare that they have no competing interests. Acknowledges Not applicable. Funding NA. Author Contribution LY S contributed to the conceptualization of the study, methodology, and formal analysis, and was responsible for writing the original draft and reviewing & editing the manuscript. DC Z assisted with the methodology, data curation, formal analysis, and contributed to writing the review & editing of the manuscript. HF H contributed to the methodology, data curation, and formal analysis. WW D and YM Z was involved in data curation, formal analysis, and reviewing & editing the manuscript. L T contributed to the study conception, and critically revised the manuscript. All authors read and approved the final manuscript. Availability of data and materials Data are available in a public, open access repository. All data relevant to the study are included in the article or uploaded as supplementary information. References Bothwell LE, Greene JA, Podolsky SH, Jones DS. Assessing the Gold Standard — Lessons from the History of RCTs. N Engl J Med. 2016;374(22):2175–81. 10.1056/NEJMms1604593 . Goldstein CE, Weijer C, Brehaut JC, et al. Ethical issues in pragmatic randomized controlled trials: a review of the recent literature identifies gaps in ethical argumentation. BMC Med Ethics. 2018;19(1):14. 10.1186/s12910-018-0253-x . Fernainy P, Cohen AA, Murray E, Losina E, Lamontagne F, Sourial N. Rethinking the pros and cons of randomized controlled trials and observational studies in the era of big data and advanced methods: a panel discussion. BMC Proceedings . 2024;18(2):1. 10.1186/s12919-023-00285-8 Lyman GH, Kuderer NM. The strengths and limitations of meta-analyses based on aggregate data. BMC Med Res Methodol. 2005;5(1):14. 10.1186/1471-2288-5-14 . Kazi DS, Moran AE, Coxson PG, et al. Cost-effectiveness of PCSK9 Inhibitor Therapy in Patients With Heterozygous Familial Hypercholesterolemia or Atherosclerotic Cardiovascular Disease. JAMA. 2016;316(7):743–53. 10.1001/jama.2016.11004 . Fleming TR, DeMets DL. Surrogate End Points in Clinical Trials: Are We Being Misled? Ann Intern Med. 1996;125(7):605–13. 10.7326/0003-4819-125-7-199610010-00011 . Taichman DB, Backus J, Baethge C, et al. Sharing Clinical Trial Data: A Proposal from the International Committee of Medical Journal Editors. PLoS Med. 2016;13(1):e1001950. 10.1371/journal.pmed.1001950 . Mello MM, Francer JK, Wilenzick M, Teden P, Bierer BE, Barnes M. Preparing for Responsible Sharing of Clinical Trial Data. N Engl J Med. 2013;369(17):1651–8. 10.1056/NEJMhle1309073 . Thompson SG, Higgins JPT. How should meta-regression analyses be undertaken and interpreted? Stat Med. 2002;21(11):1559–73. 10.1002/sim.1187 . Baker WL, White CM, Cappelleri JC, Kluger J, Coleman CI. Health Outcomes, Policy, and Economics (HOPE) Collaborative Group. Understanding heterogeneity in meta-analysis: the role of meta-regression. Int J Clin Pract. 2009;63(10):1426–34. 10.1111/j.1742-1241.2009.02168.x . Salanti G, Higgins JPT, Ades AE, Ioannidis JPA. Evaluation of networks of randomized trials. Stat Methods Med Res. 2008;17(3):279–301. 10.1177/0962280207080643 . Dias S, Sutton AJ, Ades AE, Welton NJ. Evidence Synthesis for Decision Making 2. Med Decis Mak. 2013;33(5):607–17. 10.1177/0272989X12458724 . Tonin FS, Rotta I, Mendes AM, Pontarolo R. Network meta-analysis: a technique to gather evidence from direct and indirect comparisons. Pharm Pract (Granada). 2017;15(1):943. 10.18549/PharmPract.2017.01.943 . Caldwell DM, Ades AE, Higgins JPT. Simultaneous comparison of multiple treatments: combining direct and indirect evidence. BMJ. 2005;331(7521):897–900. Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JPT. Evaluating the quality of evidence from a network meta-analysis. PLoS ONE. 2014;9(7):e99682. 10.1371/journal.pone.0099682 . Sterne JA, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol. 2000;53(11):1119–29. 10.1016/s0895-4356(00)00242-0 . Signorovitch JE, Sikirica V, Erder MH, et al. Matching-Adjusted Indirect Comparisons: A New Tool for Timely Comparative Effectiveness Research. Value Health. 2012;15(6):940–7. 10.1016/j.jval.2012.05.004 . Ishak KJ, Proskorovsky I, Benedict A. Simulation and Matching-Based Approaches for Indirect Comparison of Treatments. Pharmacoecon. 2015;33(6):537–49. 10.1007/s40273-015-0271-1 . Harari O, Soltanifar M, Cappelleri JC, et al. Network meta-interpolation: Effect modification adjustment in network meta‐analysis using subgroup analyses. Res Synthesis Methods. 2023;14(2):211–33. 10.1002/jrsm.1608 . Giugliano RP, Pedersen TR, Park JG, et al. Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. Lancet. 2017;390(10106):1962–71. 10.1016/s0140-6736(17)32290-0 . Szarek M, White HD, Schwartz GG, et al. Alirocumab and Cardiovascular Outcomes after Acute Coronary Syndrome. J Am Coll Cardiol. 2019;73(4):387–96. 10.1016/j.jacc.2018.10.039 . Ray KK, Colhoun HM, Szarek M, et al. Effects of alirocumab on cardiovascular and metabolic outcomes after acute coronary syndrome in patients with or without diabetes: a prespecified analysis of the ODYSSEY OUTCOMES randomised controlled trial. Lancet Diabetes Endocrinol. 2019;7(8):618–28. 10.1016/S2213-8587(19)30158-5 . Metabolic risk factors. and effect of alirocumab on cardiovascular events after acute coronary syndrome: a post-hoc analysis of the ODYSSEY OUTCOMES randomised controlled trial. Lancet Diabetes Endocrinol. 2022;10(5):330–40. 10.1016/S2213-8587(22)00043-2 . Jukema JW, Zijlstra LE, Bhatt DL, et al. Effect of Alirocumab on Stroke in ODYSSEY OUTCOMES. Circulation. 2019;140(25):2054–62. 10.1161/CIRCULATIONAHA.119.043826 . Freitag A, Gurskyte L, Sarri G. Increasing transparency in indirect treatment comparisons: is selecting effect modifiers the missing part of the puzzle? A review of methodological approaches and critical considerations. J Comp Eff Res. 2023;12(10):e230046. 10.57264/cer-2023-0046 . Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods. 2012;3(2):80–97. 10.1002/jrsm.1037 . Hancock M, Herbert RD, Maher CG. A guide to interpretation of studies investigating subgroups of responders to physical therapy interventions. Phys Ther. 2009;89(7):698–704. 10.2522/ptj.20080351 . Henderson CR. Best Linear Unbiased Estimation and Prediction under a Selection Model. Biometrics. 1975;31(2):423. 10.2307/2529430 . Garrison LP, Neumann PJ, Erickson P, Marshall D, Mullins CD. Value Health. 2007;10(5):326–35. 10.1111/j.1524-4733.2007.00186.x . Using Real-World Data for Coverage and Payment Decisions: The ISPOR Real-World Data Task Force Report. Natarajan P, Young R, Stitziel NO, et al. Polygenic Risk Score Identifies Subgroup With Higher Burden of Atherosclerosis and Greater Relative Benefit From Statin Therapy in the Primary Prevention Setting. Circulation. 2017;135(22):2091–101. 10.1161/CIRCULATIONAHA.116.024436 . Phillippo DM, Ades AE, Dias S, Palmer S, Abrams KR, Welton NJ. Methods for Population-Adjusted Indirect Comparisons in Health Technology Appraisal. Med Decis Mak. 2018;38(2):200–11. 10.1177/0272989X17725740 . Signorovitch JE, Sikirica V, Erder MH, et al. Matching-adjusted indirect comparisons: a new tool for timely comparative effectiveness research. Value Health. 2012;15(6):940–7. 10.1016/j.jval.2012.05.004 . Ridker PM, Everett BM, Thuren T, et al. Antiinflammatory Therapy with Canakinumab for Atherosclerotic Disease. N Engl J Med. 2017;377(12):1119–31. 10.1056/NEJMoa1707914 . Bhatt DL, Steg PG, Miller M, et al. Cardiovascular Risk Reduction with Icosapent Ethyl for Hypertriglyceridemia. N Engl J Med. 2019;380(1):11–22. 10.1056/NEJMoa1812792 . Additional Declarations No competing interests reported. Supplementary Files Supplement.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7274417","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":499055241,"identity":"119184fc-d61e-402e-80bd-ae2e895fe34c","order_by":0,"name":"Lingyao Sun","email":"","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Lingyao","middleName":"","lastName":"Sun","suffix":""},{"id":499055242,"identity":"9cef692c-9b82-4768-b45d-c033ebb0499f","order_by":1,"name":"Dachuang Zhou","email":"","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Dachuang","middleName":"","lastName":"Zhou","suffix":""},{"id":499055243,"identity":"de9d9881-7ef7-4d33-ba7b-bfa48523c77b","order_by":2,"name":"Hongfei Hu","email":"","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Hongfei","middleName":"","lastName":"Hu","suffix":""},{"id":499055244,"identity":"eefdd44f-ca6f-4e8a-b9a7-def2177f8b46","order_by":3,"name":"Weiwei Ding","email":"","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Weiwei","middleName":"","lastName":"Ding","suffix":""},{"id":499055245,"identity":"c263803d-f75c-4ed4-b4d8-f59773a719c2","order_by":4,"name":"Yimei Zhong","email":"","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":false,"prefix":"","firstName":"Yimei","middleName":"","lastName":"Zhong","suffix":""},{"id":499055246,"identity":"3f20d31b-654c-4a01-a597-13c9fa15daf0","order_by":5,"name":"Lei Tian","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyUlEQVRIie3QsQrCMBCA4ZNCdDjJJhHFvkKKgw6Cr5LuDro7FASnomtLX6KPUDno5BvYoSA4x0U6iXEW2owO+beDfBwXAJfrX1MAU/C8otCNHehFhiB4LLyksS2BLwGc04BZvOdZvKjrfYW8j5qM9PmoaCeiugaRKh84Pgxz2i4hSDPVTqTYGMIIJRmSICh5syJvwjVhTchsSXg0WzwEOyKqcpeEJ0JBTJpPFt238OyQP5sXzfiZ7lo3K59POgiI1tGGuFwul+u3D0xbRB7Yc4sjAAAAAElFTkSuQmCC","orcid":"","institution":"China Pharmaceutical University","correspondingAuthor":true,"prefix":"","firstName":"Lei","middleName":"","lastName":"Tian","suffix":""}],"badges":[],"createdAt":"2025-08-01 21:53:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7274417/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7274417/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89232629,"identity":"b1510177-e2f2-4f68-bc14-174a191e503b","added_by":"auto","created_at":"2025-08-17 14:28:55","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":86030,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eResearch Framework and Data Processing Workflow of the Meta-interpolation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLDL-C, Low-Density Lipoprotein Cholesterol; MACE, Major Adverse Cardiovascular Events; HR, Hazard Ratio; RWD, Real-World Data; NHANES,National Health and Nutrition Examination Survey; CVDs, Cardiovascular Diseases.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7274417/v1/6ff0c8fdc39c84fbdf368823.png"},{"id":89233336,"identity":"4aa029a1-7fee-4685-ad1e-3b6ad8ff2f18","added_by":"auto","created_at":"2025-08-17 14:36:55","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":25525,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eForest plot of regression results\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCVDs: Cardiovascular Diseases; MI: Myocardial Infarction; LDL-C: Low-Density Lipoprotein Cholesterol.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7274417/v1/b50a22a0061ecbc39efd5415.png"},{"id":90081957,"identity":"1fbb252f-a450-483c-9dcf-2c8941aaf9cb","added_by":"auto","created_at":"2025-08-28 09:10:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1074998,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7274417/v1/a1321108-6bb8-4ce8-b600-96ed3376f726.pdf"},{"id":89233337,"identity":"35bed95e-5044-49a4-9822-e20fca72b7f5","added_by":"auto","created_at":"2025-08-17 14:36:55","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":825224,"visible":true,"origin":"","legend":"","description":"","filename":"Supplement.docx","url":"https://assets-eu.researchsquare.com/files/rs-7274417/v1/af9c7f760aedfb8c4c4312a1.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Real-World and Subgroup Data Synthesis to Guide PCSK9 Inhibitor Use in Cardiovascular Disease","fulltext":[{"header":"Introduction","content":"\u003cp\u003eRandomized clinical trials (RCTs) remain the cornerstone for evaluating drug efficacy due to their robust internal validity\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. However, limitations persist in their application to real-world clinical decision-making\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Among these challenges are the lack of head-to-head trials comparing competing interventions and the frequent unavailability of individual patient data (IPD)\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. These gaps are particularly pronounced in the field of cardiovascular disease, where evaluating long-term outcomes like major adverse cardiovascular events (MACE) often requires large-scale and prolonged follow-up studies\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. The cost, ethical challenges, and feasibility constraints of conducting such trials often leave decision-makers reliant on indirect comparisons to guide treatment decisions\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eTo address these gaps, researchers and health technology assessment agencies have increasingly turned to advanced evidence synthesis techniques, including network meta-analysis (NMA), matching-adjusted indirect comparisons (MAIC), and simulated treatment comparisons (STC)\u003csup\u003e\u003cspan additionalcitationids=\"CR10 CR11 CR12 CR13 CR14\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e–\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. While these methods provide valuable insights, their effectiveness is contingent on specific conditions. NMA requires a well-connected network of trials with consistent populations\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e; MAIC and STC rely on access to IPD and assume that treatment effect modifiers are shared across studies\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. When these assumptions are violated or data are unavailable—as is often the case in real-world settings—the utility of these approaches diminishes.\u003c/p\u003e\u003cp\u003eIn this context, we introduced and evaluated a novel method termed meta-interpolation\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. This approach aimed to extend conventional meta-analytic strategies by integrating real-world data (RWD) and subgroup summary data from randomized trials to estimate individualized treatment effects without requiring IPD. Meta-interpolation leveraged the distributional structure of patient covariates observed in large-scale population datasets to impute missing subgroup characteristics in trial data. Using these estimates, we applied regression modeling to predict treatment effects tailored to specific patient populations.\u003c/p\u003e\u003cp\u003eIn this study, we applied the meta-interpolation framework to assess and compare the effectiveness of two PCSK9 inhibitors—evolocumab and alirocumab—in patients with atherosclerotic cardiovascular disease (ASCVD) and elevated low-density lipoprotein cholesterol (LDL-C). By integrating RWD from over 870,000 individuals with trial subgroup summaries from the FOURIER and ODYSSEY LONG TERM studies, we demonstrated how meta-interpolation could generate clinically relevant, covariate-adjusted hazard ratios in the absence of IPD\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. These results may help inform more precise treatment decisions for ASCVD patients, promoting evidence synthesis and advancing precision therapeutics.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cb\u003eStudy Design and Data Sources\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis comparative-effectiveness modeling study integrates subgroup-level trial data with real-world population datasets to estimate covariate-adjusted treatment effects for two lipid-lowering therapies (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The analysis was conducted between September 2024 and June 2025. Ethical approval was not required, as only deidentified, publicly available data were used.\u003c/p\u003e\u003cp\u003eA systematic review identified two pivotal randomized controlled trials—FOURIER (N = 27,564) and ODYSSEY LONG TERM (N = 18,924)—as the sources for trial-based efficacy and subgroup summaries (eMethods in the Supplement). Both studies evaluated PCSK9 inhibitors in adults with established ASCVD and LDL-C ≥ 70 mg/dL, using a randomized, double-blind, placebo-controlled design (eTable 5 in the Supplement). Subgroup hazard ratios and covariate means were extracted from published results and supplementary materials\u003csup\u003e\u003cspan additionalcitationids=\"CR23\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e–\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eReal-world covariate distributions were derived from six population-based studies: the Survey of Health, Ageing and Retirement in Europe (SHARE), the Health and Retirement Study (HRS), the English Longitudinal Study of Ageing (ELSA), the China Health and Retirement Longitudinal Study (CHARLS), the Longitudinal Aging Study in India (LASI), and the Mexican Health and Aging Study (MHAS). Collectively, these studies comprised 872,550 individual records spanning multiple continents and time periods. After deduplication and eligibility screening (history of ASCVD and hypercholesterolemia), 27,413 unique records remained for analysis (eFigure 2 in the Supplement).\u003c/p\u003e\u003cp\u003e\u003cb\u003eTarget Population and Covariate Selection\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe target population for this study consists of the patient cohort from the 2017–2018 National Health and Nutrition Examination Survey (NHANES) database in the United States, with the objective of more accurately simulating the effects of the drug in real-world patients (eFigure 3 in the Supplement).\u003c/p\u003e\u003cp\u003eBased on expert interviews and the availability of data, the analysis incorporated seven prespecified covariates: age ≥ 65 years, sex, race (White vs non-White), history of stroke, history of myocardial infarction (MI), diabetes status, and elevated LDL-C\u003csup\u003e\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e–\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e(eTable 6 in the Supplement). All covariates were encoded as binary variables to match the format of subgroup summaries from the trials. LDL-C was available only in selected datasets (e.g., CHARLS and NHANES) and was used selectively in primary and sensitivity analyses.\u003c/p\u003e\u003cp\u003e\u003cb\u003eReal-World Matching to Trial Cohorts\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo ensure comparability between the real-world dataset and trial populations, we implemented a greedy matching algorithm based on a least-squares minimization of covariate means. For each trial, we defined a \"target mean\" vector for the seven covariates using available trial data. From the full real-world sample, we selected a subset whose covariate means closely matched the target means within a ± 10% tolerance margin (eMethods in the Supplement). The algorithm iteratively removed observations that most reduced the total squared distance between the real-world and trial covariate vectors until optimal balance was achieved.\u003c/p\u003e\u003cp\u003e\u003cb\u003eBLUP Imputation of Subgroup Covariates\u003c/b\u003e\u003c/p\u003e\u003cp\u003eSubgroup summaries from trials typically report hazard ratios stratified by only a subset of covariates. To estimate treatment effects for unreported strata, we applied best linear unbiased prediction (BLUP) using real-world correlation matrices\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. The imputed proportion of an unreported subgroup was calculated as:\u003c/p\u003e\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\hat {{p}}_{u}={\\overline{p}}_{u}+{{\\Sigma\\:}}_{ut}{{\\Sigma\\:}}_{tt}^{-1}({p}_{t}-{\\overline{p}}_{t})$$\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eHere, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{p}_{t}\\:\\)\u003c/span\u003e\u003c/span\u003erepresents the reported subgroup proportions, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\overline{p}}_{t}\\:\\)\u003c/span\u003e\u003c/span\u003eand \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\overline{p}}_{u}\\:\\)\u003c/span\u003e\u003c/span\u003eare real-world means, and\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:{\\Sigma\\:}\\)\u003c/span\u003e\u003c/span\u003e is the estimated covariance matrix of all covariates from the matched real-world cohort. This allowed us to infer the joint distribution of patient characteristics in each trial subgroup without access to IPD.\u003c/p\u003e\u003cp\u003e\u003cb\u003eEstimation of Covariate-Adjusted Treatment Effects\u003c/b\u003e\u003c/p\u003e\u003cp\u003eUsing imputed subgroup-level covariate distributions and their associated hazard ratios, we fitted a population-level linear regression model with a log link function to approximate treatment effects. The model took the form:\u003c/p\u003e\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\text{l}\\text{o}\\text{g}\\left(H{R}_{i}\\right)={\\beta\\:}_{0}+\\sum\\:_{j=1}^{k}{\\beta\\:}_{j}{X}_{ij}+{\\epsilon\\:}_{i}$$\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:H{R}_{i}\\)\u003c/span\u003e\u003c/span\u003e denotes the subgroup hazard ratio and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{X}_{ij}\\:\\)\u003c/span\u003e\u003c/span\u003ethe mean value of covariate \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\:\\)\u003c/span\u003e\u003c/span\u003ein subgroup \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e. Regression coefficients \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\beta\\:}_{j}\\:\\)\u003c/span\u003e\u003c/span\u003eestimate the marginal effect modification by each covariate. We used these coefficients to predict treatment effects in the NHANES population (N = 495) by plugging in the average covariate profile.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSensitivity Analyses\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe conducted multiple sensitivity analyses to assess the robustness of our findings. These included: (1) repeating the matching procedure using a ± 5% and ± 15% covariate balance threshold; (2) substituting LDL-C correlations from CHARLS to evaluate geographic sensitivity. Full results are reported in the Supplement.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003eStudy Population Characteristics\u003c/b\u003e\u003c/p\u003e\u003cp\u003eCompared to the combined real-world cohort (N\u0026thinsp;=\u0026thinsp;27,413), which had higher proportions of older adults (68.8% aged\u0026thinsp;\u0026ge;\u0026thinsp;65), women (51.4%), and individuals with prior stroke (28.2%), the matched cohorts more closely reflected the demographic and clinical characteristics of their respective trials.\u003c/p\u003e\u003cp\u003eAfter matching, the final real-world cohorts comprised 11,682 participants aligned with the FOURIER trial population (evolocumab) and 7,189 participants aligned with the ODYSSEY LONG TERM trial population (alirocumab). In the matched cohorts, 44.0% and 26.9% of participants were aged\u0026thinsp;\u0026ge;\u0026thinsp;65 years, 78.3% and 74.8% were male, and 85.0% and 79.4% identified as White, respectively. The prevalence of prior stroke was 12.0% in the FOURIER-matched group and 5.0% in the ODYSSEY-matched group, while prior myocardial infarction was present in 69.0% and 83.0%, respectively. Diabetes was reported in 25.0% of the FOURIER-matched group and 28.8% of the ODYSSEY-matched group.\u003c/p\u003e\u003cp\u003eCompared to the NHANES population (N\u0026thinsp;=\u0026thinsp;495), which had a higher proportion of older adults (57.6% aged\u0026thinsp;\u0026ge;\u0026thinsp;65), women (41.6%), non-White individuals (51.3%), and patients with prior stroke (43.4%) or diabetes (39.4%), the matched cohorts were younger, more predominantly male and White, and had a lower burden of cerebrovascular and metabolic comorbidities (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBaseline characteristics of the population\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eVariable as measured in study\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eCombined cohort(N\u0026thinsp;=\u0026thinsp;27413)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eFOURIER\u003c/p\u003e\u003cp\u003e(N\u0026thinsp;=\u0026thinsp;27,564)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003eODYSSEY LONG TERM\u003c/p\u003e\u003cp\u003e(N\u0026thinsp;=\u0026thinsp;18924)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eNHANES(N\u0026thinsp;=\u0026thinsp;495)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eTarget means\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMatched means(N\u0026thinsp;=\u0026thinsp;11682)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eTarget means\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMatched means(N\u0026thinsp;=\u0026thinsp;7189)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAge\u0026thinsp;\u0026ge;\u0026thinsp;65: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e68.84%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e44.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e46.27%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e26.90%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e26.92%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e57.58%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMale: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e48.63%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e78.30%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e77.20%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e74.80%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e74.79%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e58.38%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRace\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWhite: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e76.83%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e85.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e82.79%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e79.40%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e79.37%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e48.69%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCVD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eStroke: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28.23%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e12.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e13.20%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e5.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e5.49%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e43.43%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMI: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e61.62%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e69.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e71.43%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e83.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e83.17%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e41.62%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMetabolic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDiabetes: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.86%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e25.00%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e25.22%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e28.80%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e28.81%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e39.39%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLDL-C\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHigher: n (%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e24.80%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e29.80%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e38.38%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eCVD: Cardiovascular Disease; MI: Myocardial Infarction; LDL-C: Low-Density Lipoprotein Cholesterol.\u003c/p\u003e\u003cp\u003e\u003cb\u003eCovariate Balance and Imputation Accuracy\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAll seven covariates in the matched samples were balanced within \u0026plusmn;\u0026thinsp;7% absolute difference from the original trial-reported subgroup means (eTable 6 in the Supplement). Correlation matrices used for BLUP revealed weak-to-moderate associations between age, cardiovascular history, and lipid levels. The highest observed correlation was between age and male sex (r\u0026thinsp;=\u0026thinsp;0.28) and between prior MI and diabetes (r\u0026thinsp;=\u0026thinsp;0.25) (eFigure 4 in the Supplement). These matrices served as the foundation for imputing missing subgroup covariate distributions in the trial data (eTable 7 in the Supplement).\u003c/p\u003e\u003cp\u003e\u003cb\u003eTreatment Effects in the NHANES Population\u003c/b\u003e\u003c/p\u003e\u003cp\u003eUsing covariate profiles from the NHANES cohort, meta-interpolation yielded an adjusted HR of 0.73 (95% CI, 0.65\u0026ndash;0.82) for evolocumab and 0.78 (95% CI, 0.76\u0026ndash;0.80) for alirocumab (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). These estimates suggest greater real-world effectiveness than those reported in the original trials (HR\u0026thinsp;=\u0026thinsp;0.85 for both agents). The improvement in apparent effectiveness may reflect adjustment to the older, higher-risk NHANES population, which included more individuals with prior stroke, MI, or elevated LDL-C.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSubgroup-Specific Treatment Heterogeneity\u003c/b\u003e\u003c/p\u003e\u003cp\u003eSubgroup analyses in the NHANES population showed generally stronger treatment effects for both drugs across age, sex, race, stroke history, metabolic status, and baseline LDL-C levels. Alirocumab showed similar efficacy in the MI subgroup between RCT and NHANES populations (HR, 0.89 vs 0.89), but was more effective in the non-MI subgroup in the RCT (HR, 0.70 vs 0.71).\u003c/p\u003e\u003cp\u003eIn RCT data, evolocumab showed greater benefit in patients\u0026thinsp;\u0026lt;\u0026thinsp;65 years, without MI, and with diabetes. Alirocumab was more effective in male patients. In the NHANES cohort, evolocumab showed enhanced efficacy in patients\u0026thinsp;\u0026ge;\u0026thinsp;65 years and those with MI, while its effect was consistent across diabetes subgroups. Alirocumab showed similar efficacy in men and women.\u003c/p\u003e\u003cp\u003e\u003cb\u003eCovariate Influence on Treatment Outcomes\u003c/b\u003e\u003c/p\u003e\u003cp\u003eEvolocumab showed consistent efficacy, with no significant modification by age, sex, metabolic status, or LDL-C level (eTable 8 in the Supplement). Modest variation was observed by race (coefficient, 0.23; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.049) and stroke history (\u0026ndash;0.26; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.044). In contrast, alirocumab demonstrated greater heterogeneity in treatment effect, with stronger efficacy associated with older age (\u0026ndash;0.20; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), male sex (\u0026ndash;0.05; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.013), White race (\u0026ndash;0.17; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), higher LDL-C (\u0026ndash;0.24; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), and presence of metabolic abnormalities (\u0026ndash;0.08; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.018). Reduced efficacy was observed in patients with prior stroke (0.20; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001) and myocardial infarction (0.34; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001).\u003c/p\u003e\u003cp\u003e\u003cb\u003eSensitivity Analyses\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAcross all sensitivity analyses\u0026mdash;including tighter (\u0026plusmn;\u0026thinsp;5%) and looser (\u0026plusmn;\u0026thinsp;15%) covariate-balance thresholds and use of LDL-C correlations from CHARLS\u0026mdash;the HRs for both drugs remained stable (eTable 9\u0026ndash;14 and eFigure 5\u0026ndash;10 in the Supplement). For evolocumab, HRs ranged from 0.72 to 0.73; for alirocumab, from 0.78 to 0.80 (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). These findings support the robustness of the meta-interpolation framework across varying data assumptions.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eHazard Ratios for Evolocumab and Alirocumab Under Sensitivity Analyses Using Meta-Interpolation\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIntervention\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBasic analysis HR (95% Cl) \u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026plusmn;\u0026thinsp;5% covariate balance threshold HR (95% Cl) \u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026plusmn;\u0026thinsp;15% covariate balance threshold HR (95% Cl) \u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLDL-C correlations from CHARLS HR (95% Cl) \u003csup\u003ec\u003c/sup\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEvolocumab\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.73 (0.65\u0026ndash;0.82)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.73(0.66\u0026ndash;0.81)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.72(0.63\u0026ndash;0.81)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.73(0.65\u0026ndash;0.81)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlirocumab\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.78 (0.76\u0026ndash;0.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.80(0.78\u0026ndash;0.82)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.80(0.78\u0026ndash;0.82)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.80(0.78\u0026ndash;0.82)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eHR, hazard ratio; CI, confidence interval; LDL-C, low-density lipoprotein cholesterol; CHARLS, China Health and Retirement Longitudinal Study; NHANES, National Health and Nutrition Examination Survey.\u003c/p\u003e\u003cp\u003e\u003csup\u003ea\u003c/sup\u003e The basic analysis represents the primary covariate-adjusted estimate in the NHANES-matched population.\u003c/p\u003e\u003cp\u003e\u003csup\u003eb\u003c/sup\u003e Sensitivity analyses were conducted by applying stricter (\u0026plusmn;\u0026thinsp;5%) and more lenient (\u0026plusmn;\u0026thinsp;15%) covariate balance thresholds during the matching procedure.\u003c/p\u003e\u003cp\u003e\u003csup\u003ec\u003c/sup\u003e Sensitivity analyses were conducted by using LDL-C correlation structures derived from CHARLS.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we introduced a novel evidence synthesis framework\u0026mdash;meta-interpolation\u0026mdash;to estimate the comparative effectiveness of evolocumab and alirocumab in a real-world population with ASCVD. By integrating subgroup-level data from RCTs with covariate distributions derived from large-scale observational datasets, we were able to estimate adjusted treatment effects in the absence of IPD. This approach has both immediate clinical implications for lipid-lowering therapy and broader applications for evidence generation in cardiovascular pharmacotherapy.\u003c/p\u003e\u003cp\u003eThe comparative findings suggest that evolocumab offers more consistent efficacy across key clinical subgroups, including by age, sex, metabolic status, and prior cardiovascular history. In contrast, alirocumab's efficacy was more variable, with stronger effects observed in younger, male, White patients with elevated LDL-C and no history of stroke or myocardial infarction. These subgroup-specific differences align with prior observations of effect modification in cardiovascular outcomes trials, where patient characteristics influence both absolute and relative risk reductions\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eIn clinical practice, such heterogeneity may support more personalized drug selection. For instance, in older adults or patients with a history of stroke\u0026mdash;populations often underrepresented in RCTs\u0026mdash;evolocumab may be a more reliable option due to its stable performance. In contrast, alirocumab may be preferentially considered in male patients with very high LDL-C levels and fewer comorbidities. These insights may be particularly useful in formulary decision-making, shared decision-making, or population-level treatment allocation under resource constraints\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eImportantly, our results are not intended to favor one drug over another universally, but rather to highlight the need for context-sensitive prescribing based on patient-specific factors. This aligns with the growing emphasis on precision medicine in cardiovascular prevention, where one-size-fits-all recommendations are increasingly being replaced by risk-stratified approaches\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eBeyond its application to PCSK9 inhibitors, the meta-interpolation method provides a flexible and scalable tool for comparative effectiveness research in settings where IPD is unavailable, head-to-head trials are lacking, or effect modification is suspected. Traditional methods such as MAIC or STC often require restrictive assumptions about shared effect modification or depend on IPD from at least one comparator trial\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In contrast, meta-interpolation can leverage publicly available RCT subgroup data and population-level real-world data to estimate adjusted treatment effects with less reliance on proprietary or inaccessible sources.\u003c/p\u003e\u003cp\u003eThis approach is particularly relevant as novel cardiovascular therapies continue to expand, including SGLT2 inhibitors, GLP-1 receptor agonists, siRNA-based therapies (e.g., inclisiran), and anti-inflammatory agents (e.g., colchicine, ziltivekimab). These agents are often evaluated in specialized trial populations but deployed across a much broader spectrum of real-world patients, creating an urgent need for methods that can translate trial results into practice\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e In addition, meta-interpolation may support HTA, reimbursement modeling, and population-based guideline development, especially in regions or systems where real-world clinical registries are fragmented but survey-based demographic and clinical data are available. Its compatibility with common observational sources like NHANES, CHARLS, or SHARE enhances its utility in global settings.\u003c/p\u003e\u003cp\u003e\u003cb\u003eLimitations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eSeveral limitations should be noted. First, the accuracy of estimated treatment effects depends on the granularity and quality of subgroup data reported in trials. Most trial publications only report limited subgroup categories and may lack joint distributions, limiting model resolution. Second, the BLUP-based interpolation relies on assumed linear relationships among covariates, which may not fully capture higher-order interactions or nonlinearities. Third, LDL-C values were not directly available in several real-world datasets, requiring assumptions based on external correlation structures, which may introduce imprecision. Finally, the modest sample size in NHANES (n\u0026thinsp;=\u0026thinsp;495) constrains subgroup-level estimates and may reduce generalizability to more diverse clinical populations.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study demonstrates that meta-interpolation is a robust and pragmatic method to estimate adjusted treatment effects when IPD is unavailable. Applied to PCSK9 inhibitors, this framework identified clinically relevant differences in subgroup efficacy that can inform individualized prescribing. More broadly, it offers a scalable strategy to extend trial-based evidence to real-world populations, supporting evidence-informed decisions in cardiovascular care and beyond.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eRCTs\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eRandomized Controlled Trials\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003ePCSK9\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eProprotein Convertase Subtilisin/Kexin Type 9\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eASCVD\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eAtherosclerotic Cardiovascular Disease\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eHR\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHazard Ratio\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eCI\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eConfidence Interval\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eIPD\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eIndividual Patient Data\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eMACE\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMajor Adverse Cardiovascular Events\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eNMA\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eNetwork Meta-Analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eMAIC\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMatching-Adjusted Indirect Comparisons\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eSTC\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSimulated Treatment Comparisons\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eRWD\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eReal-World Data\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eBLUP\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eBest Linear Unbiased Prediction\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eLDL-C\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eLow-Density Lipoprotein Cholesterol\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eSHARE\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSurvey of Health, Ageing and Retirement in Europe\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eHRS\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHealth and Retirement Study\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eELSA\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eEnglish Longitudinal Study of Ageing\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eCHARLS\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eChina Health and Retirement Longitudinal Study\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eLASI\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eLongitudinal Aging Study in India\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eMHAS\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMexican Health and Aging Study\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eNHANES\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eNational Health and Nutrition Examination Survey\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eCVDs\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eCardiovascular Diseases\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eMI\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMyocardial Infarction\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eSGLT2\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSodium-Glucose Cotransporter-2\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eGLP-1\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eGlucagon-Like Peptide-1\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003esiRNA\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSmall Interfering RNA\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003e\u003cb\u003eHTA\u003c/b\u003e\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHealth Technology Assessment\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003ch2\u003eAcknowledges\u003c/h2\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eNA.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eLY S contributed to the conceptualization of the study, methodology, and formal analysis, and was responsible for writing the original draft and reviewing \u0026amp; editing the manuscript. DC Z assisted with the methodology, data curation, formal analysis, and contributed to writing the review \u0026amp; editing of the manuscript. HF H contributed to the methodology, data curation, and formal analysis. WW D and YM Z was involved in data curation, formal analysis, and reviewing \u0026amp; editing the manuscript. L T contributed to the study conception, and critically revised the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eAvailability of data and materials\u003c/h2\u003e\n\u003cp\u003eData are available in a public, open access repository. All data relevant to the study are included in the article or uploaded as supplementary information.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBothwell LE, Greene JA, Podolsky SH, Jones DS. Assessing the Gold Standard \u0026mdash; Lessons from the History of RCTs. N Engl J Med. 2016;374(22):2175\u0026ndash;81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMms1604593\u003c/span\u003e\u003cspan address=\"10.1056/NEJMms1604593\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGoldstein CE, Weijer C, Brehaut JC, et al. Ethical issues in pragmatic randomized controlled trials: a review of the recent literature identifies gaps in ethical argumentation. BMC Med Ethics. 2018;19(1):14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12910-018-0253-x\u003c/span\u003e\u003cspan address=\"10.1186/s12910-018-0253-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFernainy P, Cohen AA, Murray E, Losina E, Lamontagne F, Sourial N. Rethinking the pros and cons of randomized controlled trials and observational studies in the era of big data and advanced methods: a panel discussion. \u003cem\u003eBMC Proceedings\u003c/em\u003e. 2024;18(2):1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12919-023-00285-8\u003c/span\u003e\u003cspan address=\"10.1186/s12919-023-00285-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLyman GH, Kuderer NM. The strengths and limitations of meta-analyses based on aggregate data. BMC Med Res Methodol. 2005;5(1):14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/1471-2288-5-14\u003c/span\u003e\u003cspan address=\"10.1186/1471-2288-5-14\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKazi DS, Moran AE, Coxson PG, et al. Cost-effectiveness of PCSK9 Inhibitor Therapy in Patients With Heterozygous Familial Hypercholesterolemia or Atherosclerotic Cardiovascular Disease. JAMA. 2016;316(7):743\u0026ndash;53. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2016.11004\u003c/span\u003e\u003cspan address=\"10.1001/jama.2016.11004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFleming TR, DeMets DL. Surrogate End Points in Clinical Trials: Are We Being Misled? Ann Intern Med. 1996;125(7):605\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7326/0003-4819-125-7-199610010-00011\u003c/span\u003e\u003cspan address=\"10.7326/0003-4819-125-7-199610010-00011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTaichman DB, Backus J, Baethge C, et al. Sharing Clinical Trial Data: A Proposal from the International Committee of Medical Journal Editors. PLoS Med. 2016;13(1):e1001950. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pmed.1001950\u003c/span\u003e\u003cspan address=\"10.1371/journal.pmed.1001950\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMello MM, Francer JK, Wilenzick M, Teden P, Bierer BE, Barnes M. Preparing for Responsible Sharing of Clinical Trial Data. N Engl J Med. 2013;369(17):1651\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMhle1309073\u003c/span\u003e\u003cspan address=\"10.1056/NEJMhle1309073\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eThompson SG, Higgins JPT. How should meta-regression analyses be undertaken and interpreted? Stat Med. 2002;21(11):1559\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/sim.1187\u003c/span\u003e\u003cspan address=\"10.1002/sim.1187\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBaker WL, White CM, Cappelleri JC, Kluger J, Coleman CI. Health Outcomes, Policy, and Economics (HOPE) Collaborative Group. Understanding heterogeneity in meta-analysis: the role of meta-regression. Int J Clin Pract. 2009;63(10):1426\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1742-1241.2009.02168.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1742-1241.2009.02168.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSalanti G, Higgins JPT, Ades AE, Ioannidis JPA. Evaluation of networks of randomized trials. Stat Methods Med Res. 2008;17(3):279\u0026ndash;301. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/0962280207080643\u003c/span\u003e\u003cspan address=\"10.1177/0962280207080643\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDias S, Sutton AJ, Ades AE, Welton NJ. Evidence Synthesis for Decision Making 2. Med Decis Mak. 2013;33(5):607\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/0272989X12458724\u003c/span\u003e\u003cspan address=\"10.1177/0272989X12458724\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTonin FS, Rotta I, Mendes AM, Pontarolo R. Network meta-analysis: a technique to gather evidence from direct and indirect comparisons. Pharm Pract (Granada). 2017;15(1):943. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18549/PharmPract.2017.01.943\u003c/span\u003e\u003cspan address=\"10.18549/PharmPract.2017.01.943\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCaldwell DM, Ades AE, Higgins JPT. Simultaneous comparison of multiple treatments: combining direct and indirect evidence. BMJ. 2005;331(7521):897\u0026ndash;900.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSalanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JPT. Evaluating the quality of evidence from a network meta-analysis. PLoS ONE. 2014;9(7):e99682. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0099682\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0099682\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSterne JA, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol. 2000;53(11):1119\u0026ndash;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s0895-4356(00)00242-0\u003c/span\u003e\u003cspan address=\"10.1016/s0895-4356(00)00242-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSignorovitch JE, Sikirica V, Erder MH, et al. Matching-Adjusted Indirect Comparisons: A New Tool for Timely Comparative Effectiveness Research. Value Health. 2012;15(6):940\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jval.2012.05.004\u003c/span\u003e\u003cspan address=\"10.1016/j.jval.2012.05.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIshak KJ, Proskorovsky I, Benedict A. Simulation and Matching-Based Approaches for Indirect Comparison of Treatments. Pharmacoecon. 2015;33(6):537\u0026ndash;49. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s40273-015-0271-1\u003c/span\u003e\u003cspan address=\"10.1007/s40273-015-0271-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHarari O, Soltanifar M, Cappelleri JC, et al. Network meta-interpolation: Effect modification adjustment in network meta‐analysis using subgroup analyses. Res Synthesis Methods. 2023;14(2):211\u0026ndash;33. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jrsm.1608\u003c/span\u003e\u003cspan address=\"10.1002/jrsm.1608\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGiugliano RP, Pedersen TR, Park JG, et al. Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. Lancet. 2017;390(10106):1962\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s0140-6736(17)32290-0\u003c/span\u003e\u003cspan address=\"10.1016/s0140-6736(17)32290-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSzarek M, White HD, Schwartz GG, et al. Alirocumab and Cardiovascular Outcomes after Acute Coronary Syndrome. J Am Coll Cardiol. 2019;73(4):387\u0026ndash;96. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jacc.2018.10.039\u003c/span\u003e\u003cspan address=\"10.1016/j.jacc.2018.10.039\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRay KK, Colhoun HM, Szarek M, et al. Effects of alirocumab on cardiovascular and metabolic outcomes after acute coronary syndrome in patients with or without diabetes: a prespecified analysis of the ODYSSEY OUTCOMES randomised controlled trial. Lancet Diabetes Endocrinol. 2019;7(8):618\u0026ndash;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S2213-8587(19)30158-5\u003c/span\u003e\u003cspan address=\"10.1016/S2213-8587(19)30158-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMetabolic risk factors. and effect of alirocumab on cardiovascular events after acute coronary syndrome: a post-hoc analysis of the ODYSSEY OUTCOMES randomised controlled trial. Lancet Diabetes Endocrinol. 2022;10(5):330\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S2213-8587(22)00043-2\u003c/span\u003e\u003cspan address=\"10.1016/S2213-8587(22)00043-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJukema JW, Zijlstra LE, Bhatt DL, et al. Effect of Alirocumab on Stroke in ODYSSEY OUTCOMES. Circulation. 2019;140(25):2054\u0026ndash;62. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1161/CIRCULATIONAHA.119.043826\u003c/span\u003e\u003cspan address=\"10.1161/CIRCULATIONAHA.119.043826\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFreitag A, Gurskyte L, Sarri G. Increasing transparency in indirect treatment comparisons: is selecting effect modifiers the missing part of the puzzle? A review of methodological approaches and critical considerations. J Comp Eff Res. 2023;12(10):e230046. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.57264/cer-2023-0046\u003c/span\u003e\u003cspan address=\"10.57264/cer-2023-0046\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSalanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods. 2012;3(2):80\u0026ndash;97. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jrsm.1037\u003c/span\u003e\u003cspan address=\"10.1002/jrsm.1037\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHancock M, Herbert RD, Maher CG. A guide to interpretation of studies investigating subgroups of responders to physical therapy interventions. Phys Ther. 2009;89(7):698\u0026ndash;704. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2522/ptj.20080351\u003c/span\u003e\u003cspan address=\"10.2522/ptj.20080351\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHenderson CR. Best Linear Unbiased Estimation and Prediction under a Selection Model. Biometrics. 1975;31(2):423. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2307/2529430\u003c/span\u003e\u003cspan address=\"10.2307/2529430\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGarrison LP, Neumann PJ, Erickson P, Marshall D, Mullins CD. Value Health. 2007;10(5):326\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1524-4733.2007.00186.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1524-4733.2007.00186.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Using Real-World Data for Coverage and Payment Decisions: The ISPOR Real-World Data Task Force Report.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNatarajan P, Young R, Stitziel NO, et al. Polygenic Risk Score Identifies Subgroup With Higher Burden of Atherosclerosis and Greater Relative Benefit From Statin Therapy in the Primary Prevention Setting. Circulation. 2017;135(22):2091\u0026ndash;101. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1161/CIRCULATIONAHA.116.024436\u003c/span\u003e\u003cspan address=\"10.1161/CIRCULATIONAHA.116.024436\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePhillippo DM, Ades AE, Dias S, Palmer S, Abrams KR, Welton NJ. Methods for Population-Adjusted Indirect Comparisons in Health Technology Appraisal. Med Decis Mak. 2018;38(2):200\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/0272989X17725740\u003c/span\u003e\u003cspan address=\"10.1177/0272989X17725740\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSignorovitch JE, Sikirica V, Erder MH, et al. Matching-adjusted indirect comparisons: a new tool for timely comparative effectiveness research. Value Health. 2012;15(6):940\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jval.2012.05.004\u003c/span\u003e\u003cspan address=\"10.1016/j.jval.2012.05.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRidker PM, Everett BM, Thuren T, et al. Antiinflammatory Therapy with Canakinumab for Atherosclerotic Disease. N Engl J Med. 2017;377(12):1119\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMoa1707914\u003c/span\u003e\u003cspan address=\"10.1056/NEJMoa1707914\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBhatt DL, Steg PG, Miller M, et al. Cardiovascular Risk Reduction with Icosapent Ethyl for Hypertriglyceridemia. N Engl J Med. 2019;380(1):11\u0026ndash;22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMoa1812792\u003c/span\u003e\u003cspan address=\"10.1056/NEJMoa1812792\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"PCSK9 inhibitors, real-world evidence, meta-interpolation, cardiovascular disease, precision medicine","lastPublishedDoi":"10.21203/rs.3.rs-7274417/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7274417/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRandomized controlled trials (RCTs) are the gold standard for assessing the efficacy of medical interventions; however, their findings often do not fully translate into real-world populations due to differences in patient characteristics. In this study, we aimed to estimate real-world populations treatment effects for PCSK9 inhibitors in patients with atherosclerotic cardiovascular disease (ASCVD) by integrating real-world data with subgroup-level data from two major RCTs, FOURIER and ODYSSEY LONG TERM. This approach seeks to improve precision medicine by aligning trial results with real-world data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe combined individual-level data from six large population-based studies (N = 872,550) with subgroup data from the FOURIER and ODYSSEY LONG TERM trials. A meta-interpolation method was applied to estimate treatment effects for two PCSK9 inhibitors, evolocumab and alirocumab, using regression modeling to adjust for covariates such as age, sex, race, and cardiovascular history. Hazard ratios (HRs) for major adverse cardiovascular events were estimated, and sensitivity analyses were conducted to test the robustness of the findings.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn a U.S.-representative population, evolocumab demonstrated consistent efficacy across subgroups with an HR of 0.73 (95% confidence interval [CI], 0.65–0.82). In contrast, alirocumab showed variable efficacy, with a significant difference in treatment effects between men (HR, 0.71) and women (HR, 0.89; P \u0026lt; 0.001). Sensitivity analyses confirmed the robustness of these findings across different covariate balance thresholds and data assumptions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIntegrating real-world data with RCT subgroup data via meta-interpolation offers a reliable method to estimate treatment effects tailored to real-world populations. Our results suggest that evolocumab provides consistent benefit across various patient subgroups, while alirocumab’s efficacy may vary based on factors such as sex and comorbidities. This approach can guide more personalized treatment decisions in ASCVD, advancing the field of precision cardiovascular pharmacotherapy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrial Registration: \u003c/strong\u003eNot applicable.\u003c/p\u003e","manuscriptTitle":"Real-World and Subgroup Data Synthesis to Guide PCSK9 Inhibitor Use in Cardiovascular Disease","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-17 14:28:50","doi":"10.21203/rs.3.rs-7274417/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"40135ea9-be68-408e-8fa4-01e9c8dc6624","owner":[],"postedDate":"August 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-08-28T09:09:34+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-17 14:28:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7274417","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7274417","identity":"rs-7274417","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}