Citations

Introduction: Lp(a) is a causal risk factor for ASCVD that is often elevated in patients with personal or family history (FHx) of ASCVD. The 2019 ESC guidelines recommend universal Lp(a) testing, while the 2019 AHA/ACC guidelines include Lp(a) as a ‘risk enhancer’. We used electronic health records (EHR) to assess utilization of Lp(a) testing in patients with ASCVD, premature ASCVD, and FHx of ASCVD at Mayo Clinic sites in Rochester MN, Phoenix AZ and Jacksonville FL and the Mayo Clinic Health System (MCHS) from 1999-2021. Methods: We ascertained Lp(a) testing using LOINC codes and ASCVD status using validated ICD/CPT-based algorithms. ASCVD included coronary heart disease (CHD), cerebrovascular disease (CVD) (ischemic stroke and/or carotid artery stenosis), peripheral artery disease, and abdominal aortic aneurysm. Premature ASCVD was defined as occurring in males < 55 years and females < 65 years. We ascertained FHx using EHR survey data from 2017-2021. We estimated the proportion of patients with ASCVD, premature ASCVD, or FHx of ASCVD who underwent Lp(a) testing. Results: Between 1999-2021, we identified 310,760 patients with at least 1 ASCVD subtype of whom 66,710 had premature disease. FHx data was available starting 2017. For the timeframe 2017-2021, we identified 267,177 patients with FHx of ASCVD and 90,269 ASCVD patients. As summarized in Table 1, Lp(a) was tested only in a small proportion of patients; 7% of ASCVD patients, 11% of premature ASCVD patients, and 6% of patients with FHx of ASCVD. Premature CHD patients had the highest testing proportion (17%). Conclusions: Our results demonstrate underutilization of Lp(a) testing in patients with personal or FHx of ASCVD. Lp(a) is one of the mediators of “ASCVD residual risk” in patients on statins, and with new Lp(a)-lowering medications in phase III clinical trials, patient and provider education will be needed to increase awareness of Lp(a) as a modulator of ASCVD risk.

Introduction: Hospitalization for COVID-19 is often associated with myocardial injury (elevated cardiac troponin [cTn]). In the initial pandemic waves, patients with myocardial injury had increased likelihood of cardiovascular complications and inpatient mortality; however, the differences in impact of distinct subtypes of myocardial injury on mortality is unclear. In addition, little is known about how these outcomes compare with the subsequent, milder, Omicron wave. Methods: Utilizing Mayo Clinic enterprise patients we developed algorithms based on diagnostic/procedural codes, cTn levels, and temporal relationships to positive COVID-19 PCR tests, to determine 5 common subtypes of myocardial injury: stress cardiomyopathy, myocarditis, pericarditis, pulmonary embolism, and myocardial infarction. We examined hospitalized adults with COVID-19; those with elevated cTn designated as cases and others as controls. We fitted Cox models using the following covariates: spline transformed age, sex, race, ethnicity, index date and clinical comorbidities. Patients were analyzed based on admission dates coinciding with either the Alpha-Delta or Omicron waves. Results: Across both wave cohorts, hospitalized patients with acute COVID-19 complicated by myocardial injury had an increased risk of mortality compared to those without ( Figure ). Patients with one of the 5 subtypes examined tended to have higher risk than those in the “other myocardial injury” category. Admission during the Omicron wave, compared to the Alpha-Delta waves, tended to have higher mortality risk in all subtypes except myocarditis. Conclusions: Our algorithms successfully identified subgroups of myocardial injury in patients hospitalized with acute COVID-19 leading to stratification of mortality risk. Surprisingly, among hospitalized patients during the Omicron wave, myocardial injury of most subtypes was tended to have increased risk of mortality compared to the Alpha-Delta waves.

Introduction: Clonal hematopoiesis of indeterminate potential (CHIP) is associated with an increased risk of heart failure (HF) and stroke, particularly HF with preserved ejection fraction and hemorrhagic stroke (HS). Hypothesis: We hypothesized that CHIP is associated with incident hypertension and this association mediates the effect of CHIP on HF and HS. Methods: In 200K UK-Biobank participants with exome sequencing, we ascertained CHIP as variant allele fraction (VAF) ≥2% and large CHIP as VAF ≥10%. Hypertension, HF, and HS diagnosis were determined based on the ICD-9 and ICD-10 codes. Cox proportional hazards models were used to test the association of both CHIP and the CHIP subtypes based on the most implicated driver genes (i.e., DNMT3A , TET2 , and ASXL1 ) with incident hypertension. Logistic regression models were used for mediation analyses, quantifying the extent to which hypertension mediates the effects of CHIP. Results: After excluding those with hematological malignancy (n=1,025) and pre-existing hypertension (n=14,754), 184,760 participants (56.14±8.12 years, 44.25% male, and 95.33% white ethnicity) were included. During a median follow-up of 12.2 years, 42,049 (22.76%) participants developed hypertension. CHIP mutations were significantly associated with incident hypertension (Table). Mutations in DNMT3A and TET2 but not ASXL1 drove this association. Mediation analysis revealed 17.9% (9.4% - 38.2%, P<0.001) of the effect of CHIP on HF and 5.7% (2.4% - 22.4%, P<0.001) of the effect on HS were mediated through the increased risk of hypertension. Conclusion: CHIP is associated with an increased risk of hypertension. The association of CHIP with HF and HS is partially mediated by hypertension. These findings suggest a new mechanistic pathway for the association of CHIP with cardiovascular diseases.

Background Cardio-ankle vascular index (CAVI) is a measure of arterial stiffness associated with cardiovascular (CV) events and mortality. African American (AA) adults have a higher prevalence of CV risk factors (RFs) and adverse social determinants of health (SDOH) leading to a higher CV mortality risk than White adults. These disparities may be associated with arterial stiffness, however limited research exists on the association of CV RFs, SDOH, and CAVI among AAs. Hypotheses We hypothesized that CV RFs (eg, diabetes) and SDOH (eg, income) would be associated with CAVI among AA adults. Our secondary hypothesis was that AA adults would have a higher prevalence of abnormal CAVI scores (≥8.0) than White adults. Methods In the FAITH! HH+ study, CAVI measurements, CV RFs, and SDOH surveys were obtained. We collected CAVI readings and CV RFs from a secondary cohort of White adults (n=600), referred for echocardiography and without structural cardiac disease. The association of CV RFs, SDOH and CAVI scores among AAs was assessed via ANOVA and univariable linear regression. We compared CV RFs among both cohorts using Chi-squared tests and ANOVA, followed by comparing CAVI scores between cohorts using a multivariable linear regression model. Results In total, 48 AAs were included from the FAITH! HH+ cohort (mean age [SD]: 54.2 [11.5] years, 67.4% women). There was no significant association between CV RFs or SDOH with CAVI scores. In the 600 White adults assessed, mean age was 59.5 [14.7] years and 45.7% were women. Compared to White adults, AA participants had a lower burden of hyperlipidemia (43.5% vs 61.4%; p=0.017), but a higher burden of hypertension (73.9% vs 57.7%; p=0.031), diabetes (28.3% vs 9.3%; p<0.001), and obesity (71.7% obese vs 36.0%; p<0.001). There was no significant difference in CAVI scores between groups when adjusted for age, sex, and CV RFs (AA cohort: mean: 8.0 [1.4] vs White cohort: 8.4 [1.5]; p=0.790). Conclusions In this pilot study, SDOH and CV RFs were not associated with CAVI scores among AAs. There was no significant difference in CAVI scores between AA and White adults, despite a significantly higher burden of CV RFs in the AA cohort. Further studies are warranted to identify factors linked with onset and progression of arterial stiffness among AAs.

Introduction: Aging and inflammation are associated with risk of atrial fibrillation (AF), to which clonal hematopoiesis of indeterminate potential (CHIP) may contribute. We investigated whether CHIP was associated with AF, inflammatory and cardiac biomarkers and cardiac structural changes. Methods: We ascertained CHIP (variant allele frequency, VAF ≥2%) in the Atherosclerosis Risk in Community Study (ARIC visit-5, n = 4,131) and the UK Biobank (UKBB, n = 195,851) participants. Using multivariable Cox regression models, we studied the association of CHIP and the top three implicated genes (i.e., DNMT3A , TET2 , ASXL1 ) with incident AF and conducted a meta-analysis on both cohorts. We also ascertained high-sensitivity C-reactive protein (hs-CRP), interleukin 6 (IL6), IL18, high-sensitivity troponin T and I (hs-TnT and hs-TnI), NT-proBNP, and echocardiography indices in ARIC. We first standardized the natural log of all markers (st-ln) and used the adjusted quantile regression to compare the median in those with and without CHIP. Results: In ARIC, the mean ± SD age was 75.8 ± 5.2 years, 41% were men, 23% were Black, 25% had CHIP, and 11.6% had large CHIP (VAF ≥10%). These measures in UKBB were 56.4 ± 8.1 years, 45%, 2%, 5.8%, and 2.6%, respectively. In the meta-analysis, compared to those without CHIP, the hazard ratios (95% confidence interval) for AF were 1.12 (1.01 - 1.25, P =0.035), 1.31 (1.07 - 1.60, P =0.009), and 1.48 (1.05 - 2.08, P =0.026) for Large-CHIP, Large- TET2, and Large- ASXL1 , respectively. hs-CRP and IL18 levels were similar in those with and without CHIP as well as gene-specific subgroups. Large- TET2 and ASXL1 were associated with higher IL6, which was separately associated with AF. Additionally, Large- ASXL1 was associated with increased left ventricle mass index and NT-proBNP (coefficients in Figure 1 ). Conclusion: Large TET2 and ASXL1 but not DNMT3A CHIP is associated with higher IL6 levels, indices of cardiac remodeling, and an increased risk of AF.

Apart from ancestry, personal or environmental covariates may contribute to differences in polygenic score (PGS) performance. We analyzed the effects of covariate stratification and interaction on body mass index (BMI) PGS (PGS BMI ) across four cohorts of European (N = 491,111) and African (N = 21,612) ancestry. Stratifying on binary covariates and quintiles for continuous covariates, 18/62 covariates had significant and replicable R 2 differences among strata. Covariates with the largest differences included age, sex, blood lipids, physical activity, and alcohol consumption, with R 2 being nearly double between best- and worst-performing quintiles for certain covariates. Twenty-eight covariates had significant PGS BMI –covariate interaction effects, modifying PGS BMI effects by nearly 20% per standard deviation change. We observed overlap between covariates that had significant R 2 differences among strata and interaction effects – across all covariates, their main effects on BMI were correlated with their maximum R 2 differences and interaction effects (0.56 and 0.58, respectively), suggesting high-PGS BMI individuals have highest R 2 and increase in PGS effect. Using quantile regression, we show the effect of PGS BMI increases as BMI itself increases, and that these differences in effects are directly related to differences in R 2 when stratifying by different covariates. Given significant and replicable evidence for context-specific PGS BMI performance and effects, we investigated ways to increase model performance taking into account nonlinear effects. Machine learning models (neural networks) increased relative model R 2 (mean 23%) across datasets. Finally, creating PGS BMI directly from GxAge genome-wide association studies effects increased relative R 2 by 7.8%. These results demonstrate that certain covariates, especially those most associated with BMI, significantly affect both PGS BMI performance and effects across diverse cohorts and ancestries, and we provide avenues to improve model performance that consider these effects.

Introduction: Circulating biomarkers may improve prediction of cardiovascular (CV) events in chronic coronary disease (CCD). The incremental value of biomarkers in CCD after adjustment for clinical factors, inducible ischemia and coronary artery disease (CAD) severity is uncertain. Methods: 756 ISCHEMIA (International Study of Comparative Health Effectiveness with Medical and Invasive Approaches) participants with site-determined moderate-severe ischemia and 9 assayed biomarkers (Figure) were included. Among participants with interpretable stress tests for ischemia and CAD by computed coronary tomography angiography (CCTA) (N=507), associations of biomarkers with the primary (CV death, myocardial infarction (MI), or hospitalization for unstable angina, heart failure, or resuscitated cardiac arrest) or secondary endpoints (CV death or MI) were evaluated using sequentially adjusted multivariable Cox proportional hazards models (Model 1=unadjusted, Model 2=clinical features, Model 3=Model 2 + multivessel CAD by CCTA, and Model 4=Model 3 + severe ischemia). Results: Median (IQR) participant age was 64 (57-70) years, 19% were female, 81% white, and 5% Hispanic. Hypertension (85%), diabetes (46%) and obesity (45%) were common, 28% had an estimated glomerular filtration rate <60 ml/min/1.73m 2 . Severe ischemia was present in 44%, and 39% had multivessel CAD (≥70% stenosis) by CCTA. Over 3.5 (2.4-4.7) years of follow-up there were 146 (19%) primary and 128 (17%) secondary endpoints. Soluble CD40L, GDF-15, NT-proBNP, hsTnT were associated with the primary and secondary endpoints in univariable and multivariable (adjusting for clinical features, multivessel CAD and severe ischemia) models. Conclusion: Biomarkers of inflammation (GDF-15), thrombosis (sCD40L), myocardial stretch (NT-proBNP) and injury (hs-troponin) were associated with CV endpoints in the ISCHEMIA trial independent of clinical features, multivessel CAD and severe ischemia.

Introduction: Polygenic risk scores (PRS) for coronary heart disease (CHD) are associated with incident CHD events. We assessed whether the strength of this association differs over the life course. Methods: We studied 7,958 genotyped participants of European-ancestry without CHD enrolled in the Atherosclerosis Risk in Communities Study (ARIC). Participants were 45-64 years old at the baseline visit occurring 1987-89 with follow up available through 2015. A genome-wide PRS for CHD composed of 1.7 million variants was computed for each participant. Incident CHD was defined as clinically adjudicated myocardial infarction including silent infarction, coronary revascularization, or fatal coronary event. An extended Cox model was constructed with age as the model time scale. The association of PRS with incident CHD was assumed to be log-linear and further modeled as a function of age assuming differing piecewise constant effect sizes across the intervals ≤55, 55-65, 65-80, and ≥80 years. The regression model also included the following covariates: sex, systolic blood pressure, total and HDL cholesterol, history of diabetes, antihypertensive medication use, and smoking status. Results: In the ARIC cohort (median enrollment age 54, interquartile range [IQR] 49-59, female 52%), over a median of 24.5 years (IQR 14.4-28.1), there were 1,872 incident CHD events. In comparison to average polygenic risk for CHD (50th percentile of PRS distribution), harboring high polygenic risk (95th percentile) was associated with a 2.9-fold increased risk of incident CHD (95% CI 2.12-3.95) up to age 55 years, independent of conventional risk factors. The strength of association was attenuated later in life (Hazard ratio, age 55-65, 2.5 [2.14-2.84]; age 65-80, 1.87 [1.68-2.07]; age ≥80, 1.78 [1.38-2.31]) Conclusions: A genome-wide PRS for CHD is more strongly associated with incident coronary events at younger age. The strength of association declined later in life while remaining significant. The age-dependent association of PRS with CHD may need to be accounted for in risk prediction models.

Introduction: Chronic coronary disease (CCD) is a leading cause of death worldwide. Whole blood (WB) transcriptomic (RNA-seq) and methylation ‘omics’ data can discern molecular underpinnings of CCD and create molecular subtypes to aid risk prediction. Methods: Of 1,064 participants with moderate-severe ischemia (Age: 67.1±6.2, 19.4% Female, 17% non-White) from the ISCHEMIA (International Study of Comparative Health Effectiveness with Medical and Invasive Approaches) biorepository, 646 participants had available WB RNA-seq and 737 had WB methylation data ( Fig 1A ). Unsupervised clustering and differential analyses identified omics-derived molecular subtypes. We identified co-regulated gene modules of each subtype, performed subtype-level enrichment of clinical variables, and performed Kaplan-Meier analyses for time to cardiovascular (CV) death or myocardial infarction (MI). Results: Multi-omic analysis revealed 4 RNA-seq-derived subtypes (RS) and 3 methylation-derived subtypes (MS). RNA-seq Subtype 1 (RS1) was enriched for molecular drivers indicative of an adaptive immune response, had clinical enrichment of female sex, statin use, and greater ischemia, and had decreased risk of CV death or MI ( Fig 1B ). RNA-seq Subtype 2 (RS2) was enriched for innate immune response pathway, and patients who were older and White, and was associated with increased risk of CV death or MI. Similarly, methylation subtypes also yielded a high-risk subtype (MS3) and a low-risk subtype (MS2) ( Fig 1C ). Integration of transcriptomic and methylomic subtypes led to further distinction of resilient and vulnerable subgroups. Conclusions: WB RNA-seq and methylation data identified molecularly-defined subtypes that informed the underlying disease pathology of CCD patients in ISCHEMIA. These subgroups were differentially enriched for clinical characteristics and event outcomes, demonstrating the potential use for blood-based omics tests to aid in the prognosis of CCD.

Introduction: Relatively little is known about outcomes following return of secondary findings detected in large scale sequencing projects. We assessed 1-year outcomes following return of pathogenic/likely pathogenic (P/LP) variants associated with familial hypercholesterolemia (FH) in phase III of the eMERGE network (2015-2019). Methods: Adult participants (n=18,544) at 7 sites were enrolled in a prospective cohort study to study the clinical impact of returning results from targeted sequencing of 68 actionable genes, including LDLR , APOB , and PCSK9 . Process, intermediate, and/or clinical outcomes within 1-year after return of results (RoR) were ascertained by electronic health record (EHR) review. A process outcome was defined as referral to a specialist or ordering new tests, including lipid profile and screening tests for coronary heart disease. An intermediate outcome was defined as new diagnosis of FH or atherosclerotic cardiovascular disease. A clinical outcome was defined as initiation or change in lipid-lowering medication. The primary outcomes were a new diagnosis of FH and modification of lipid lowering therapy. Results: Results were returned to 134 individuals, and data for 119 participants were available for the outcomes analysis. Process (86%), intermediate (49%), and/or clinical (26%) outcomes occurred in 92% of participants. Process outcomes included 72% with new tests ordered and 61% referred to a specialist. Before RoR, 86.5% of individuals with FH associated P/LP variants had a diagnosis of hypercholesterolemia but only 17.6% had a diagnosis of FH; 89% were on a statin but only 58% were on a high intensity statin. Following RoR, 44% (52/119) received a new diagnosis of FH. Lipid lowering treatment was modified in 26% (31/119), including initiation of a statin in 5%, increasing statin dose in 13%, and adding ezetimibe or a PCSK9 inhibitor in 12.6%. Conclusions: In a multisite cohort of EHR-linked biobanks, monogenic FH was underdiagnosed (only 17.6% had a specific diagnosis) and undertreated (only 58% were on a high intensity statin); 44% of participants received a new diagnosis of FH and 26% had lipid lowering treatment modified after RoR. These results highlight the clinical utility of returning FH variants in EHR-linked biobanks.