10-year Cardiovascular Disease Event Research Articles

ASCVD risk scores versus a novel cancer-specific Machine Learning-based calculator among patients with breast, colorectal, lung or prostate cancer

Abstract Background Patients with cancer have up to a 3-fold higher risk for cardiovascular disease (CVD) than the general population. Consequently, traditional risk scores developed to predict CVD in the general population (i.e., Pooled Cohort Equations (PCE), Predicting Risk of cardiovascular disease EVENTs [PREVENT]) may provide less accurate risk prediction for this population. Purpose Compare cancer-specific Machine Learning (ML)-based score to PCE and PREVENT to predict 10-year CVD risk in patients with breast cancer (BC), colorectal cancer (CRC), lung cancer (LC), or prostate cancer (PC). Methods Patients aged ≥18 years, diagnosed with BC, CRC, LC, or PC between 2005-2012 at a large hybrid academic-community practice in Northeast Ohio, United States, were included. An ML Extreme Gradient Boosting (XGBoost) algorithm, adapted for survival modeling, was developed using a training subset of the cohort (55% train + 20% test) and ranked 40 to 50 covariates (including social determinants of health, cancer treatment) for predicting 10-year CVD risk for each cancer type using SHAP (SHapley Additive exPlanations) values. The top 10 ML predictors were utilized to create a predictive equation using logistic regression models. This equation was then tested in the cohort validation subset (25%) and subsequently compared to the PCE and PREVENT (both simple and enhanced versions) using the area under the curve (AUC) of the time-dependent receiver operating characteristic curve. The CVD events were identified using ICD codes corresponding to Acute Coronary Syndrome, Ischemic Stroke, and Heart Failure. CVD risk was assessed through Kaplan-Meier. Results We included 10,240 patients (Table 1). The actual 10-year CVD events were: BC 21% (95% CI 19-23), CRC 10% (95% CI 8-12), LC 27% (95% CI 24-30), and PC 20% (95% CI 17-23). The ACC/AHA PCE predicted mean 10-year CVD risks of 16% (95% CI 15-18) for BC, 19% (95% CI 17-21) for CRC, 28% (95% CI 25-32) for LC, and 19% (95% CI 19-20) for PC, with corresponding AUCs of 0.75, 0.65, 0.76, and 0.61, respectively. The PREVENT simple version predicted mean 10-year CVD risks of 13% (95% CI 13-14) for BC, 13% (95% CI 12-13) for CRC, 15% (95% CI 14-15) for LC, and 16% (95% CI 12-19) for PC, with AUCs of 0.55, 0.55, 0.55, and 0.53, respectively. The PREVENT enhanced equation predicted mean 10-year CVD risks of 17% (95% CI 17-18) for BC, 15% (95% CI 14-15) for CRC, 15% (95% CI 15-16) for LC, and 17% (95% 16-19) for PC, with AUCs of 0.75, 0.62, 0.79, and 0.63, respectively. The predictive equation derived from the top 10 ML predictors (Table 2) achieved AUCs of 0.84 for BC, 0.76 for CRC, 0.83 for LC, and 0.71 for PC. Conclusion(s) Conventional CVD equations inadequately assess the risk in patients with BC, CRC, LC, or PC, often resulting in underestimation or overestimation. Cancer-specific ML-derived equations show good performance and emphasize the importance of integrating cancer-related covariates for precise prediction.Table 1Table 2

Validation of Systematic Coronary Risk Evaluation 2 (SCORE2) and SCORE2-Older Persons in the EPIC-Norfolk prospective population cohort.

The European Systematic Coronary Risk Evaluation 2 (SCORE2) and SCORE2-Older Persons (OP) models are recommended to identify individuals at high 10-year risk for cardiovascular disease (CVD). Independent validation and assessment of clinical utility is needed. This study aims to assess discrimination, calibration, and clinical utility of low-risk SCORE2 and SCORE2-OP. Validation in individuals aged 40-69 years (SCORE2) and 70-79 years (SCORE2-OP) without baseline CVD or diabetes from the European Prospective Investigation of Cancer (EPIC) Norfolk prospective population study. We compared 10-year CVD risk estimates with observed outcomes (cardiovascular mortality, non-fatal myocardial infarction, and stroke). For SCORE2, 19 560 individuals (57% women) had 10-year CVD risk estimates of 3.7% [95% confidence interval (CI) 3.6-3.7] vs. observed 3.8% (95% CI 3.6-4.1) [observed (O)/expected (E) ratio 1.0 (95% CI 1.0-1.1)]. The area under the curve (AUC) was 0.75 (95% CI 0.74-0.77), with underestimation of risk in men [O/E 1.4 (95% CI 1.3-1.6)] and overestimation in women [O/E 0.7 (95% CI 0.6-0.8)]. Decision curve analysis (DCA) showed clinical benefit. Systematic Coronary Risk Evaluation 2-Older Persons in 3113 individuals (58% women) predicted 10-year CVD events in 10.2% (95% CI 10.1-10.3) vs. observed 15.3% (95% CI 14.0-16.5) [O/E ratio 1.6 (95% CI 1.5-1.7)]. The AUC was 0.63 (95% CI 0.60-0.65) with underestimation of risk across sex and risk ranges. Decision curve analysis showed limited clinical benefit. In a UK population cohort, the SCORE2 low-risk model showed fair discrimination and calibration, with clinical benefit for preventive treatment initiation decisions. In contrast, in individuals aged 70-79 years, SCORE2-OP demonstrated poor discrimination, underestimated risk in both sexes, and limited clinical utility.

Abstract 11309: Lipoprotein(a) and the Pooled Cohort Equations for Risk Prediction: The Multi-Ethnic Study of Atherosclerosis

Introduction Lipoprotein(a) [Lp(a)] is an independent risk factor for atherosclerotic cardiovascular disease (ASCVD) but is not included in the ACC/AHA ASCVD Risk Estimator based on the Pooled Cohort Equations (PCE). Methods We evaluated the association between Lp(a) >50 mg/dL and risk for cardiovascular disease (CVD) events using multivariable Cox proportional hazards models stratified by ASCVD risk category in individuals free of baseline ASCVD with elevated Lp(a) in the Multi-Ethnic Study of Atherosclerosis (MESA). We then evaluated actual versus PCE-predicted risk by comparing 10-year CVD event rates in those with Lp(a) >50 mg/dL versus Lp(a) ≤50 mg/dL using Poisson regression with adjustment for the components of the PCE by strata of risk (<5%, 5-7.5%, 7.5-20%, ≥20%). We then evaluated the change in the C-index and continuous net reclassification improvement (NRI) with the addition of Lp(a) to the PCE. Results: In participants with borderline risk (5-7.5%), elevated Lp(a) was associated with greater risk for CVD events (HR 2.59, 95% CI 1.41, 4.78). Among patients with elevated Lp(a), the greatest discrimination of risk was present in those with borderline (5-7.5%) predicted risk, where the PCE significantly underestimated risk (observed rate: 8.9% for Lp(a) >50 mg/dL vs 4.1% for Lp(a)≤50 mg/dL, risk ratio 2.54, 95% CI 1.35-4.62, Figure ). In this group, addition of Lp(a) to the PCE resulted in C-index improvement of 0.063. The continuous NRI with the addition of Lp(a) for CVD was 0.0386 (95% CI -0.0115, 0.1087) overall, but 0.3142 (95% CI 0.0350, 0.6083) among those at borderline predicted risk. Conclusions: The PCE underestimates risk in individuals with borderline predicted risk and elevated Lp(a). The addition of Lp(a) to the PCE improves risk prediction in individuals with borderline risk, for whom the greatest opportunity to guide treatment decisions exists.

SCORE2-OP risk prediction algorithms: estimating incident cardiovascular event risk in older persons in four geographical risk regions.

The aim of this study was to derive and validate the SCORE2-Older Persons (SCORE2-OP) risk model to estimate 5- and 10-year risk of cardiovascular disease (CVD) in individuals aged over 70 years in four geographical risk regions. Sex-specific competing risk-adjusted models for estimating CVD risk (CVD mortality, myocardial infarction, or stroke) were derived in individuals aged over 65 without pre-existing atherosclerotic CVD from the Cohort of Norway (28 503 individuals, 10 089 CVD events). Models included age, smoking status, diabetes, systolic blood pressure, and total- and high-density lipoprotein cholesterol. Four geographical risk regions were defined based on country-specific CVD mortality rates. Models were recalibrated to each region using region-specific estimated CVD incidence rates and risk factor distributions. For external validation, we analysed data from 6 additional study populations {338 615 individuals, 33 219 CVD validation cohorts, C-indices ranged between 0.63 [95% confidence interval (CI) 0.61-0.65] and 0.67 (0.64-0.69)}. Regional calibration of expected-vs.-observed risks was satisfactory. For given risk factor profiles, there was substantial variation across the four risk regions in the estimated 10-year CVD event risk. The competing risk-adjusted SCORE2-OP model was derived, recalibrated, and externally validated to estimate 5- and 10-year CVD risk in older adults (aged 70 years or older) in four geographical risk regions. These models can be used for communicating the risk of CVD and potential benefit from risk factor treatment and may facilitate shared decision-making between clinicians and patients in CVD risk management in older persons.

The association between serum uric acid levels and 10-year cardiovascular disease incidence: results from the ATTICA prospective study.

Limited data suggests possible gender-specific association between serum uric acid (SUA) and cardiovascular disease (CVD) incidence. The aim of the present analysis was to evaluate the association between SUA levels and 10-year CVD incidence (2002-2012) in the ATTICA study participants. Overall, 1687 apparently healthy volunteers, with SUA measurements, residing in the greater metropolitan Athens area (Greece), were included. Multivariable Cox-regression models were used to estimate the hazard ratios for SUA in relation to 10-year CVD incidence. Receiver operating curve analysis was conducted to detect optimal SUA cut-off values. Participants in the 2nd and 3rd SUA tertile had 29 and 73% higher 10-year CVD incidence compared with those in the 1st tertile (p < 0.001). In gender-specific analysis, only in women SUA was independently associated with CVD incidence; women in the 3rd SUA tertile had 79% greater 10-year CVD event risk compared to their 1st tertile counterparts. Obese in the 3rd SUA tertile had 2-times higher CVD incidence compared to those in the 1st tertile. Similar findings were observed in metabolically healthy (vs. unhealthy) and metabolically healthy obese. SUA thresholds best predicting 10-year CVD incidence was 5.05 and 4.15 mg/dL (0.30 and 0.25 mmol/L) in men and women, respectively. In conclusion, increased SUA levels were independently related to 10-year CVD event rate in women, obese and metabolically healthy individuals. SUA could predict 10-year CVD incidence even at low levels. Further studies are warranted to identify SUA cut-off values that may improve the detection of individuals at higher CVD risk in clinical practice.

Coronary Artery Calcium Scoring: New Insights into Clinical Interpretation-Lessons from the CAC Consortium.

Coronary artery calcium (CAC) is a highly specific marker for coronary atherosclerosis. The CAC Consortium, a multicenter, retrospective, real-world cohort study, was established to investigate the association between CAC and long-term, cause-specific mortality. This review summarizes findings from CAC Consortium studies published between 2016 and 2020, aiming to demystify CAC as a clinical decision-guiding tool and push the limits of who might benefit from CAC in clinical practice. CAC has been shown to effectively stratify cardiovascular risk across ethnicities irrespective of age, sex, and risk factor burden. In comparison to other widely used risk scores, CAC appears to be most consistent in its ability to add to cardiovascular disease (CVD) event prediction. Beyond risk stratification, CAC has been shown to identify high-risk patient subgroups. While currently recommended only for patients at borderline or intermediate risk by the American College of Cardiology/American Heart Association (10-year atherosclerotic CVD event risk, 5% to < 20%), CAC scoring may also provide value in select young patients aged 30-49 years and in low-risk patients with a family history. While new studies emphasize that patients with a CAC greater than or equal to 1000 be considered a distinct patient group, a CAC of 0 has additionally emerged to be a reliable negative risk factor, identifying patients at low risk of both CVD and non-CVD mortality. © RSNA, 2020.

P1533The sex-based effect of skeletal muscle mass on 10-year cardiovascular disease prognosis of patients with acute coronary syndrome: the mediating effect of systemic inflammation

Abstract Background/Introduction Predictive and prognostic ability of muscle mass in CVD settings is increasingly discussed. Purpose The sex-specific effect of skeletal muscle mass index (SMI) on 10-year recurrent fatal/non fatal cardiovascular disease (CVD) event of acute coronary syndrome (ACS) patients was evaluated. Methods In 2006–2009, n=1,000 consecutive patients (n=222 females), hospitalized at a cardiology clinic with ACS diagnosis and with symptoms and left ventricular function indicative of heart failure were selected. SMI was created to reflect skeletal muscle mass through appendicular skeletal muscle mass (indirectly calculated through population formulas) divided by body mass index (BMI). Results In 10-year follow-up (2016), 55% of ACS patients exhibited recurrent fatal/non fatal CVD events (53% in females vs. 62% in males, p=0.04). Patients in 2nd SMI tertile (mostly overweight) had 10% lower risk for CVD recurrence (female:male rate ratio=0.87) over their counterparts at 1st (mostly normalweight) and 3rd tertile (mostly obese). Multivariate analysis revealed that ACS patients in 2nd SMI tertile presented 46% and 85% lower CVD event risk over their counterparts in 1st tertile (Hazard Ratio (HR)=0.54, 95% Confidence Interval (95% CI) 0.30, 0.96, p=0.002) and 3rd tertile (HR=1.85, 95% CI 1.05, 2.94, p=0.03). Sex-related sensitivity analyses revealed that even if U-shape association was retained in both males and females in crude and age- adjusted models, further adjustment for clinical, lifestyle and sociodemographic factors resulted in a significant trend only in females (HR=1.85, 95% CI 1.05, 2.94, p=0.03). Mean (95% confidence interval) survival (free-of-CVD) time (years) in females patients at 1st, 2nd and 3rd tertile was 7.4 (6.5, 8.2), 8.2 (7.1, 9.2) and 6.7 (5.5, 7.9), respectively. Inflammatory markers i.e. C-reactive protein (CRP) and white blood cells (WBC) had strong confounding effect on the association between SMI and 10-year recurrent CVD event rate in females. A significant inverse linear association between the examined factors was observed only in females (ρ(CRP)=-0.490, p&lt;0.001 &amp; ρ(WBC)=-0.380, p=0.05). Unadjusted analysis revealed that females of 2nd tertile had significantly lower mean values of inflammatory markers compared with their 1st tertile counterparts (all ps&lt;0.05). Additionally, females of 3rd tertile presented slightly higher inflammation levels compared with females of 2nd tertile yet this difference was not significant. In males no significant trend was observed. Conclusion AU-shape association between SMI and 10-year CVD event especially in females was highlighted. This work reveals sex-specific remarks for “obesity-lean paradox” in secondary prevention, implying that high muscle mass accompanied by obesity and excess adiposity may not guarantee better prognosis. Acknowledgement/Funding None

P4410Normal weight central obesity and 10-year cardiovascular disease onset in apparently healthy males and females: the interacting effect of sex

Abstract Background/Introduction The notion that standard weight–height tables are the proper way to determine high-risk groups for obesity-related disorders has been largely challenged, yet scarce data exist from the standpoint of sex. Purpose To evaluate the effect of combined weight and waist-circumference (WC) status on 10-year first fatal/non fatal cardiovascular disease (CVD) event in apparently healthy males and females. Methods A prospective study (2001–2012) was conducted in Greece with n=1,514 males and n=1,528 females (aged &gt;18 years old) free of CVD. Follow-up assessment of CVD event (2011–2012) was achieved in n=2,020 participants (317 cases). Normal weight (NW) status was defined as body mass index (BMI) from 18.5kg/m2 to 25kg/m2 while central obesity as WC≥102cm for males and ≥88cm for females. NW central obesity (NWO) was defined as BMI within the normal range and abnormal WC. Results The prevalence of NWO reached 10% (n=302) with female:male ratio being close to 4:1 (p&lt;0.001). Within NW group ∼25% of participants presented increased WC. Ten-year CVD incidence in NWO subjects reached 14.1% which was twice as high as the respective rate for NW participants without central obesity (7.4%) and followed by the 15.4% and 24.1% in overweight/obese participants without or with central obesity, respectively (p&lt;0.001). The female:male CVD event rate ratio was 0.37, 0.77, 0.42, 0.64 ranking from NW participants without central obesity to NWO and overweight/obese without or with central obesity (p&lt;0.001). Cox regression analysis revealed that NWO status was independently associated with 10-year CVD event (Hazard Ratio (HR)=1.87, 95% Confidence Interval (95% CI) 1.04, 3.37, p=0.03) when compared with their NW counterparts with normal WC, after adjusting for sociodemographic, clinical, lifestyle and biochemical factors (i.e. insulin resistance and systemic inflammation). When NWO subjects were set as reference group and compared with their overweight/obese counterparts with or without central obesity, no significant trends were observed (all ps&gt;0.05). A significant interaction between sex and the combined weight and central obesity status was observed (p for interaction&lt;0.001). Stratified analysis revealed that the aforementioned associations were retained only in NWO females (HR=2.35, 95% CI 1.15, 4.79, p=0.01) when compared with their NW counterparts without central obesity; in males the association did not reach the level of significance. C-statistics [C-index (95% CI)] revealed that within the NW range, WC-adjusted model had a better discriminative ability in females [0.810 (0.778, 0.842)] compared with males [0.789 (0.763, 0.810)] while a BMI-adjusted model seemed to discriminate better the CVD event in males [0.801 (0.772, 0.831)] than in females [0.762 (0.731, 0.798)]. Conclusion Studies addressing the complex interaction between fat and BMI as well as its effect on CVD are demanded, yet with the potential to highlight sex-specific remarks. Acknowledgement/Funding The ATTICA study is supported by research grants from the Hellenic Cardiology Society [HCS2002] and the Hellenic Atherosclerosis Society [HAS2003].

P1923Deep and machine learning models to improve risk prediction of cardiovascular disease using data extraction from electronic health records

Abstract Abstract Background Advances in precision medicine will require an increasingly individualized prognostic evaluation of patients in order to provide the patient with appropriate therapy. The traditional statistical methods of predictive modeling, such as SCORE, PROCAM, and Framingham, according to the European guidelines for the prevention of cardiovascular disease, not adapted for all patients and require significant human involvement in the selection of predictive variables, transformation and imputation of variables. In ROC-analysis for prediction of significant cardiovascular disease (CVD), the areas under the curve for Framingham: 0.62–0.72, for SCORE: 0.66–0.73 and for PROCAM: 0.60–0.69. To improve it, we apply for approaches to predict a CVD event rely on conventional risk factors by machine learning and deep learning models to 10-year CVD event prediction by using longitudinal electronic health record (EHR). Methods For machine learning, we applied logistic regression (LR) and recurrent neural networks with long short-term memory (LSTM) units as a deep learning algorithm. We extract from longitudinal EHR the following features: demographic, vital signs, diagnoses (ICD-10-cm: I21-I22.9: I61-I63.9) and medication. The problem in this step, that near 80 percent of clinical information in EHR is “unstructured” and contains errors and typos. Missing data are important for the correct training process using by deep learning &amp; machine learning algorithm. The study cohort included patients between the ages of 21 to 75 with a dynamic observation window. In total, we got 31517 individuals in the dataset, but only 3652 individuals have all features or missing features values can be easy to impute. Among these 3652 individuals, 29.4% has a CVD, mean age 49.4 years, 68,2% female. Evaluation We randomly divided the dataset into a training and a test set with an 80/20 split. The LR was implemented with Python Scikit-Learn and the LSTM model was implemented with Keras using Tensorflow as the backend. Results We applied machine learning and deep learning models using the same features as traditional risk scale and longitudinal EHR features for CVD prediction, respectively. Machine learning model (LR) achieved an AUROC of 0.74–0.76 and deep learning (LSTM) 0.75–0.76. By using features from EHR logistic regression and deep learning models improved the AUROC to 0.78–0.79. Conclusion The machine learning models outperformed a traditional clinically-used predictive model for CVD risk prediction (i.e. SCORE, PROCAM, and Framingham equations). This approach was used to create a clinical decision support system (CDSS). It uses both traditional risk scales and models based on neural networks. Especially important is the fact that the system can calculate the risks of cardiovascular disease automatically and recalculate immediately after adding new information to the EHR. The results are delivered to the user's personal account.

Exposure to second hand smoke and 10-year (2002–2012) incidence of cardiovascular disease in never smokers: The ATTICA cohort study

BackgroundDespite WHO Framework Convention of Tobacco Control (FCTC) adoption, effective implementation of national smoking bans remains pending in several countries. This study quantified the association of second hand smoke (SHS) exposure and 10-year cardiovascular disease (CVD) among never smokers in such settings. MethodsIn 2001–2002, a sample of 1514 males and 1528 females (range: 18–89 years old) were randomly selected in Greece. Frequency and duration of SHS exposure (i.e. exposure extending >30 min/day) within the home and/or workplace were assessed by interview. Following a 10-year follow-up period (2002–2012), incidence of non-fatal and fatal CVD (ICD-10) was evaluated among n = 2020 participants. The analytic study sample consisted of all never smokers (n = 910). ResultsDespite national smoking ban implementation (2009), 44.6% (n = 406) of never smokers reported SHS exposure. While SHS exposed never smokers exhibited a more favorable profile of CVD-related risk factors at baseline, they subsequently developed similar 10-year CVD incidence rates, at a younger mean age (p = 0.001), than their non-exposed counterparts. Following adjustment for several lifestyle and clinical factors, SHS exposed never smokers exhibited a two-fold elevated 10-year CVD risk (adj. HR: 2.04, 95% CI: 1.43–2.92), particularly among women (adj. HR: 2.45, 95% CI: 1.45–4.06). SHS exposure accounted for 32% excess Population Attributable Risk (PAR) for 10-year CVD events in never smokers, with highest rates (PAR: 52%) being among those exposed in the workplace. ConclusionThe prevention of SHS associated CVD and related healthcare costs mandates additional strategies for securing the effective implementation of comprehensive WHO FCTC based national smoking bans.

Adherence to a DASH-style diet and cardiovascular disease risk: The 10-year follow-up of the ATTICA study.

Recent findings suggest a protective role of the DASH dietary pattern on cardiovascular disease (CVD) incidence and mortality. In this direction, we aimed at investigating the relationship between adherence to a DASH-style diet and CVD risk in a Greek cohort. This sub-sample from the ATTICA epidemiological study consisted of 669 adults with a complete dietary profile at baseline, adequate to calculate DASH-diet score, and complete 10-year follow-up (2002-2012). Demographic, clinical and lifestyle parameters were thoroughly assessed at baseline and CVD incidence was recorded upon medical records at follow-up. Adherence to the DASH-style diet was assessed by a DASH-style diet score developed for the study (range 9-45). Mean value (SD) of the DASH-diet score was 27.1 (5.1) (range 13-41). Adherence to a DASH-style diet was associated neither with the 10-year CVD risk nor with baseline clinical parameters. Multiple regression analysis revealed that, after appropriate adjustments, only age (46% increase per 5-life-years) and BMI (9.7% increase per unit of BMI) were associated with 10-year CVD events. In this small cohort of a Mediterranean population, a cardioprotective effect of a DASH-style diet was not detected.

Quantifying Unmet Need in Statin-Treated Hyperlipidemia Patients and the Potential Benefit of Further LDL-C Reduction Through an EHR-Based Retrospective Cohort Study.

Statins are effective in helping prevent cardiovascular disease (CVD). However, studies suggest that only 20%-64% of patients taking statins achieve reasonable low-density lipoprotein cholesterol (LDL-C) thresholds. On-treatment levels of LDL-C remain a key predictor of residual CVD event risk. To (a) determine how many patients on statins achieved the therapeutic threshold of LDL-C < 100 mg per dL (general cohort) and < 70 mg per dL (secondary prevention cohort, or subcohort, with preexisting CVD); (b) estimate the number of potentially avoidable CVD events if the threshold were reached; and (c) forecast potential cost savings. A retrospective, longitudinal cohort study using electronic health record data from the Indiana Network for Patient Care (INPC) was conducted. The INPC provides comprehensive information about patients in Indiana across health care organizations and care settings. Patients were aged > 45 years and seen between January 1, 2012, and October 31, 2016 (ensuring study of contemporary practice), were statin-naive for 12 months before the index date of initiating statin therapy, and had an LDL-C value recorded 6-18 months after the index date. Subsequent to descriptive cohort analysis, the theoretical CVD risk reduction achievable by reaching the threshold was calculated using Framingham Risk Score and Cholesterol Treatment Trialists' Collaboration formulas. Estimated potential cost savings used published first-year costs of CVD events, adjusted for inflation and discounted to the present day. Of the 89,267 patients initiating statins, 30,083 (33.7%) did not achieve the LDL-C threshold (subcohort: 58.1%). In both groups, not achieving the threshold was associated with patients who were female, black, and those who had reduced medication adherence. Higher levels of preventive aspirin use and antihypertensive treatment were associated with threshold achievement. In both cohorts, approximately 64% of patients above the threshold were within 30 mg per dL of the respective threshold. Adherence to statin therapy regimen, judged by a medication possession ratio of ≥ 80%, was 57.4% in the general cohort and 56.7% in the subcohort. Of the patients who adhered to therapy, 23.7% of the general cohort and 50.5% of the subcohort had LDL-C levels that did not meet the threshold. 10-year CVD event risk in the at-or-above threshold group was 22.78% (SD = 17.24%) in the general cohort and 29.56% (SD = 18.19%) in the subcohort. By reducing LDL-C to the threshold, a potential relative risk reduction of 14.8% in the general cohort could avoid 1,173 CVD events over 10 years (subcohort: 15.7% and 454 events). Given first-year inpatient and follow-up costs of $37,300 per CVD event, this risk reduction could save about $1,455 per patient treated to reach the threshold (subcohort: $1,902; 2017 U.S. dollars) over a 10-year period. Across multiple health care systems in Indiana, between 34% (general cohort) and 58% (secondary prevention cohort) of patients treated with statins did not achieve therapeutic LDL-C thresholds. Based on current CVD event risk and cost projections, such patients seem to be at increased risk and may represent an important and potentially preventable burden on health care costs. Funding support for this study was provided by Merck (Kenilworth, NJ). Chase and Boggs are employed by Merck. Simpson is a consultant to Merck and Pfizer. The other authors have nothing to disclose.

Skeletal muscle mass in acute coronary syndrome prognosis: Gender-based analysis from Hellenic Heart Failure cohort

Background and aimsPredictive and prognostic ability of muscle mass in CVD settings is increasingly discussed. The gender-specific effect of skeletal muscle mass index (SMI) on 10-year recurrent fatal/non fatal cardiovascular disease (CVD) event of acute coronary syndrome (ACS) patients was evaluated. Methods and resultsIn 2006–2009, n = 1000 consecutive patients (n = 222 women), hospitalized at the First Cardiology Clinic of Athens with ACS diagnosis and with symptoms and left ventricular function indicative of heart failure were selected. SMI was created to reflect skeletal muscle mass through appendicular skeletal muscle mass (indirectly calculated through population formulas) divided by body mass index (BMI). In the 10-year follow-up (2016), 55% of ACS patients experienced recurrent fatal/non fatal CVD events (53% in women vs.62% in men, p = 0.04). Patients in the 2nd SMI tertile (mostly overweight) had 10% lower risk for CVD recurrence (women:men rate ratio = 0.87) over their counterparts in the 1st (mostly normalweight) and 3rd tertile (mostly obese). Multivariate analysis revealed that ACS patients in the 2nd SMI tertile presented 46% and 85% lower CVD event risk over their counterparts in the 1st tertile (Hazard Ratio (HR) = 0.54, 95% Confidence Interval (95% CI) 0.30, 0.96, p = 0.002) and 3rd tertile (HR = 1.85, 95%CI 1.05, 2.94, p = 0.03). Gender-based analysis revealed that this trend remained significant only in women. Inflammatory markers had strong confounding effect. ConclusionA U-shape association between SMI and 10-year CVD event especially in women was highlighted. This work reveals gender-specific remarks for “obesity-lean paradox” in secondary prevention, implying that high muscle mass accompanied by obesity and excess adiposity may not guarantee better prognosis.

Learning from Longitudinal Data in Electronic Health Record and Genetic Data to Improve Cardiovascular Event Prediction

Current approaches to predicting a cardiovascular disease (CVD) event rely on conventional risk factors and cross-sectional data. In this study, we applied machine learning and deep learning models to 10-year CVD event prediction by using longitudinal electronic health record (EHR) and genetic data. Our study cohort included 109, 490 individuals. In the first experiment, we extracted aggregated and longitudinal features from EHR. We applied logistic regression, random forests, gradient boosting trees, convolutional neural networks (CNN) and recurrent neural networks with long short-term memory (LSTM) units. In the second experiment, we applied a late-fusion approach to incorporate genetic features. We compared the performance with approaches currently utilized in routine clinical practice – American College of Cardiology and the American Heart Association (ACC/AHA) Pooled Cohort Risk Equation. Our results indicated that incorporating longitudinal feature lead to better event prediction. Combining genetic features through a late-fusion approach can further improve CVD prediction, underscoring the importance of integrating relevant genetic data whenever available.

Transition from metabolically benign to metabolically unhealthy obesity and 10-year cardiovascular disease incidence: The ATTICA cohort study

Background/ObjectivesMetabolically benign obesity remains a scientific field of considerable debate. The aim of the present work was to evaluate whether metabolically healthy obese (MHO) status is a transient condition which propagates 10-year cardiovascular disease (CVD) onset. MethodsA prospective longitudinal study was conducted during 2001–2012, the ATTICA study studying 1514 (49.8%) men and 1528 (50.2%) women (aged >18 years old) free of CVD and residing in the greater Athens area, Greece. Follow-up assessment of first combined CVD event (2011−2012) was achieved in n = 2020 participants; of them, 317 (15.7%) incident cases were identified. Obesity was defined as body mass index ≥30 kg/m2 and healthy metabolic status as absence of all NCEP ATP III (2005) metabolic syndrome components (excluding waist circumference). ResultsThe MHO prevalence was 4.8% (n = 146) with 28.2% of obese participants presenting metabolically healthy status at baseline. Within this group, 52% developed unhealthy metabolic status during the 10-year follow up. MHO vs. metabolically healthy non-obese participants had a higher likelihood of presenting with 10-year CVD events, yet only the subset of them who lost their baseline status reached the level of significance (Hazard Ratio (HR) = 1.43, 95% Confidence Interval (95% CI) 1.02, 2.01). Sensitivity analyses revealed that MHO status was independently associated with elevated CVD risk in women and participants with low adherence to the Mediterranean diet, low grade inflammation, and insulin resistance. ConclusionsMHO status is a transient condition where weight management is demanded to prevent the establishment of unhealthy cardiometabolic features. The existence of obese persons who remain “longitudinally” resilient to metabolic abnormalities is an emerging area of future research.

Long-Term Association of Low-Density Lipoprotein Cholesterol With Cardiovascular Mortality in Individuals at Low 10-Year Risk of Atherosclerotic Cardiovascular Disease.

The associations of low-density lipoprotein cholesterol (LDL-C) with cardiovascular disease (CVD) and coronary heart disease mortality in an exclusively low estimated 10-year risk group are not well delineated. We sought to determine the long-term associations of various LDL-C and non-high-density lipoprotein cholesterol (HDL-C) thresholds and CVD and coronary heart disease mortality in a large, low 10-year risk cohort. The study sample included participants of the CCLS (Cooper Center Longitudinal Study) without a history of CVD or diabetes mellitus and defined as low risk (<7.5%) for 10-year atherosclerotic CVD events at baseline based on Pooled Cohort Risk Assessment Equations. The associations of fasting LDL-C and non-HDL-C with CVD mortality were tested with Cox proportional hazards models. In 36 375 participants (72% men, median age 42) followed for a median of 26.8 years, 1086 CVD and 598 coronary heart disease deaths occurred. Compared with LDL-C <100 mg/dL, LDL-C categories 100 to 129 mg/dL, 130 to 159 mg/dL, 160 to 189.9 mg/dL, and ≥190 mg/dL were associated with a significantly higher risk of CVD death, with hazard ratios of 1.4 (95% CI, 1.1-1.7), 1.3 (95% CI, 1.1-1.6), 1.9 (95% CI, 1.5-2.4), and 1.7 (95% CI, 1.3-2.3), and mean reductions in years free of CVD death of 1.8, 1.1, 4.3, and 3.9, respectively. After adjustment for atherosclerotic CVD risk factors, LDL-C categories 160 to 189 mg/dL and ≥190 mg/dL remained independently associated with CVD mortality, with hazard ratios of 1.7 (95% CI, 1.4-2.2) and 1.5 (95% CI, 1.2-2.1), respectively. In multivariable-adjusted models using non-HDL-C <130 mg/dL as the reference, non-HDL-C 160 to 189 mg/dL, 190 to 219 mg/dL, and ≥220 mg/dL were significantly associated with CVD death, with hazard ratios of 1.3 (95% CI, 1.1-1.6), 1.8 (95% CI, 1.4-2.2), and 1.5 (95% CI, 1.2-2.0), respectively. Restricting the cohort to those with 10-year risk <5% did not diminish the associations of LDL-C and non-HDL-C with CVD mortality. In a low 10-year risk cohort with long-term follow-up, LDL-C and non-HDL-C ≥160 mg/dL were independently associated with a 50% to 80% increased relative risk of CVD mortality. These findings may have implications for future cholesterol treatment paradigms.

10-Year Cardiovascular Disease Risk Estimation Based on Lipid Profile-Based and BMI-Based Framingham Risk Scores across Multiple Sociodemographic Characteristics: The Malaysian Cohort Project.

Cardiovascular disease (CVD) leads to high morbidity and mortality rate worldwide. Therefore, it is important to determine the risk of CVD across the sociodemographic factors to strategize preventive measures. The current study consisted of 53,122 adults between the ages of 35 and 65 years from The Malaysian Cohort project during recruitment phase from year 2006 to year 2012. Sociodemographic profile and physical activity level were assessed via self-reported questionnaire, whereas relevant CVD-related biomarkers and biophysical variables were measured to determine the Framingham Risk Score (FRS). The main outcome was the 10-year risk of CVD via FRS calculated based on lipid profile and body mass index (BMI) associated formulae. The BMI-based formula yielded a higher estimation of 10-year CVD risk than the lipid profile-based formula in the study for both males (median = 13.2% and 12.7%, respectively) and females (median = 4.3% and 4.2%, respectively). The subgroup with the highest risk for 10-year CVD events (based on both FRS formulae) was the Malay males who have lower education level and low physical activity level. Future strategies for the reduction of CVD risk should focus on screening via BMI-based FRS in this at-risk subpopulation to increase the cost-effectiveness of the prevention initiatives.

Observed versus predicted cardiovascular events and all-cause death in HIV infection: a longitudinal cohort study

BackgroundThe aim of the study was to assess the applicability of an algorithm predicting 10-year cardiovascular disease (CVD) generated in the setting of the Framingham Heart Study to a real-life, contemporary Italian cohort of HIV-positive subjects.MethodsThe study was an observational longitudinal cohort study. The probability for 10-year CVD events according to the Framingham algorithm was assessed in 369 consecutive HIV-positive participants free from overt CVD enrolled in 2004, who were followed for a median of 10.0 years (interquartile range, 9.1-10.1). Cardiovascular events included myocardial infarction, hospitalized heart failure, revascularized angina, sudden cardiac death, stroke, peripheral arterial disease.ResultsOver 3097 person-years of observation, we observed a total of 34 CVD events, whereas Framingham algorithm predicted the occurrence of 34.3 CVD events. CVD event rate was 11.0/1000 person-years of follow-up. In a receiver operating characteristics curve analysis, Framingham risk equation showed an excellent predictive value for incident CVD events (c-statistics, 0.83; 95% confidence interval, 0.76-0.90). In a multivariable Cox analysis, age, smoking and diabetes were independent predictors of CVD events. All-cause death rate was 20.0/1000 person-years of follow-up (n = 62 deaths). Causes of death included liver diseases (18), malignancies (14), AIDS-related (11); cardiovascular (9) and others (10). In a Cox analysis, age, AIDS diagnosis and chronic hepatitis were independent predictors of death.ConclusionsObserved CVD events in HIV-infected patients were well predicted by Framingham algorithm. Established major CVD risk factors are the strongest determinants of CVD morbidity in an Italian contemporary cohort of HIV-positive subjects. Interventions to modify traditional risk factors are urgently needed in HIV people.

Abstract P047: The Effect of Serum C-reactive Protein (CRP) or Handgrip Strength Inclusion on the Framingham Risk Score Model of CVD Risk in a Cohort of Middle-aged to Elderly Australian Men

Introduction: Cardiovascular disease (CVD) risk prediction models are used extensively in clinical practice. Recent reviews have suggested that such models would benefit from the identification of novel predictors to improve specificity and sensitivity. Both serum CRP and lower handgrip strength have been demonstrated to predict incident CVD suggesting its utility as a prognostic CVD marker. Objectives: To determine whether handgrip strength improves global CVD risk prediction models based on the Framingham Risk Score (FRS) (using NCEP-ATP-III criteria), and to validate CRP as a marker adding predictive value. Method: Participants were drawn from a randomly-selected, community-dwelling cohort of men residing in Adelaide, Australia. Of these, n=2162 men without CVD at baseline (2002-5) and covariate data (age, fasting total &amp; HDL cholesterol, smoking status, mean systolic blood pressure, and anti-hypertensive usage) were followed for a median period of 8.3 years. Incident CVD events (n=248) were identified through a state-wide registry of in-hospital events (emergency department visit or hospital separation) and cardiac-specific mortality. Mean handgrip strength (by dynamometry) and serum high-sensitivity CRP were obtained at baseline visit. The added prognostic value of handgrip strength and CRP was assessed using splines and log-likelihood ratio tests of nested Cox proportional models, BIC and AIC. Model discrimination was assessed using Harrell’s c-index. Results: Within the FRS 10-year risk groups 0-&lt;5%, 5-&lt;10%, 10-&lt;20%, and 20+% the estimated 10-year CVD event rates were 3% (95%CI: 2-5%), 7% (5-10%), 15% (13-18%) and 22% (17-27%), respectively. Hand grip strength did not add predictive value above standard Framingham risk factors ( p =0.25). CRP made a contribution to estimating global CVD risk ( p =0.002). Both the BIC and AIC were reduced (4.9 and 7.9 units, respectively) and the c-index increased from 0.727 to 0.732 (SE=0.024). Conclusions: A global risk prediction model that includes serum CRP improves CV risk classification in men.

Sick Populations and Sick Subpopulations: Reducing Disparities in Cardiovascular Disease Between Blacks and Whites in the United States.

Cardiovascular disease (CVD) death rates are much higher in blacks than whites in the United States. It is unclear how CVD risk and events are distributed among blacks versus whites and how interventions reduce racial disparities. We developed risk models for fatal and for fatal and nonfatal CVD using 8 cohorts in the United States. We used 6154 adults who were 50 to 69 years of age in the National Health and Nutrition Examination Survey 1999 to 2012 to estimate the distributions of risk and events in blacks and whites. We estimated the total and disparity impacts of a range of population-wide, targeted, and risk-based interventions on 10-year CVD risks and event rates. Twenty-five percent (95% confidence interval [CI], 22-28) of black men and 12% (95% CI, 10-14) of black women were at ≥6.67% risk of fatal CVD (almost equivalent to 20% risk of fatal or nonfatal CVD) compared with 10% (95% CI, 8-12) of white men and 3% (95% CI, 2-4) of white women. These high-risk individuals accounted for 55% (95% CI, 49-59) of CVD deaths among black men and 42% (95% CI, 35-46) in black women compared with 30% (95% CI, 24-35) in white men and 18% (95% CI, 13-22) in white women. We estimated that an intervention that treated multiple risk factors in high-risk individuals could reduce black-white difference in CVD death rate from 1659 to 1244 per 100 000 in men and from 1320 to 897 in women. Rates of fatal and nonfatal CVD were generally similar between black and white men. In women, a larger proportion of women were at ≥7.5% risk of CVD (30% versus 19% in whites), and an intervention that targeted multiple risk factors among this group was estimated to reduce black-white differences in CVD rates from 1688 to 1197 per 100 000. A substantially larger proportion of blacks have a high risk of fatal CVD and bear a large share of CVD deaths. A risk-based intervention that reduces multiple risk factors could substantially reduce overall CVD rates and racial disparities in CVD death rates.