Author + information
- Received June 26, 2018
- Revision received September 17, 2018
- Accepted September 20, 2018
- Published online November 14, 2018.
- Marcio H. Miname, MD, PhDa,
- Marcio Sommer Bittencourt, MD, PhD, MPHb,c,
- Sérgio R. Moraes, BSca,
- Rômulo I.M. Alves, BSca,
- Pamela R.S. Silva, PhDa,
- Cinthia E. Jannes, PhDa,
- Alexandre C. Pereira, MD, PhDa,
- José E. Krieger, MD, PhDa,
- Khurram Nasir, MD, MSc, MPHd and
- Raul D. Santos, MD, PhDa,e,∗ ()
- aHeart Institute (InCor), University of São Paulo Medical School Hospital, São Paulo, Brazil
- bHospital Israelita Albert Einstein & School of Medicine, Faculdade Israelita de Ciência da Saúde Albert Einstein, São Paulo, Brazil
- cCenter for Clinical and Epidemiological Research, University Hospital & São Paulo State Cancer Institute, University of São Paulo, São Paulo, Brazil
- dCenter for Outcomes Research and Evaluation and Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut
- eHospital Israelita Albert Einstein, São Paulo, Brazil
- ↵∗Address for correspondence:
Dr. Raul D. Santos, Heart Institute of São Paulo Medical School, Unidade Clínica de Lipides InCor HCFMUSP, Av. Dr Eneas C. Aguiar 44, CEP-05403-900, São Paulo, SP, Brazil.
Objectives The aim of this study was to evaluate the role of coronary artery calcium (CAC) as a predictor of atherosclerotic cardiovascular disease (ASCVD) (fatal or not myocardial infarction, stroke, unstable angina requiring revascularization, and elective myocardial revascularization) events in asymptomatic primary prevention molecularly proven heterozygous familial hypercholesterolemia (FH) subjects receiving standard lipid-lowering therapy.
Background FH is associated with premature ASCVD. However, the clinical course of ASCVD in subjects with FH is heterogeneous. CAC score, a marker of subclinical atherosclerosis burden, may optimize ASCVD risk stratification in FH.
Methods Subjects with FH underwent CAC measurement and were followed prospectively. The association of CAC with ASCVD was evaluated using multivariate analysis.
Results A total of 206 subjects (mean age 45 ± 14 years, 36.4% men, baseline and on-treatment low-density lipoprotein cholesterol 269 ± 70 mg/dl and 150 ± 56 mg/dl, respectively) were followed for a median of 3.7 years (interquartile range: 2.7 to 6.8 years). CAC was present in 105 (51%), and 15 ASCVD events (7.2%) were documented. Almost half of events were hard outcomes, and the others were elective myocardial revascularizations. The annualized rates of events per 1,000 patients for CAC scores of 0 (n = 101 [49%]), 1 to 100 (n = 62 [30%]) and >100 (n = 43 [21%]) were, respectively, 0, 26.4 (95% confidence interval: 12.9 to 51.8), and 44.1 (95% confidence interval, 26.0 to 104.1). In multivariate Cox regression analysis, log(CAC score + 1) was independently associated with incident ASCVD events (hazard ratio: 3.33; 95% CI: 1.635 to 6.790; p = 0.001).
Conclusions CAC was independently associated with ASCVD events in patients with FH receiving standard lipid-lowering therapy. This may help further stratify near-term risk in patients who might be candidates for further treatment with newer therapies.
- computed tomography
- coronary calcification
- familial hypercholesterolemia
- risk factors
Familial hypercholesterolemia (FH) is an autosomal-codominant disorder characterized by high low-density lipoprotein cholesterol (LDL-C) levels that is associated with premature atherosclerotic cardiovascular disease (ASCVD) (1,2). Although overall, patients with FH are considered a very high ASCVD risk group, a considerable number of these vulnerable subjects do not develop cardiovascular events, regardless of the elevated low LDL-C levels (2), while others do so despite intensive lipid-lowering therapy with statins and other medications (3), suggesting that the actual risk is heterogeneous.
Recent studies have suggested that similar to varied manifestations of clinical ASCVD, the presence and distribution of coronary artery calcium (CAC), a surrogate of atherosclerotic plaque burden, is heterogeneous in subjects with FH (4–8). Although the favorable prognostic value of absence of CAC has been well established in primary prevention settings (9), the true prevalence of absence of CAC and its ability to reclassify risk among patients with FH remains a knowledge gap (1).
As a result, in the present study we sought to evaluate the role of CAC as a predictor of ASCVD events in a prospective cohort of subjects with molecularly proven asymptomatic heterozygous FH receiving standard lipid-lowering therapies. Optimization of risk stratification might have important clinical implications on appropriate management and judicious selection of novel, efficacious, but more expensive lipid-lowering therapies, such as proprotein convertase subtilisin kexin type 9 (PCSK9) inhibitors in patients with FH (10–13).
The study population consisted of patients with molecularly proven heterozygous FH followed at the lipid clinic of the Heart Institute, University of São Paulo Medical School Hospital, in São Paulo, Brazil. Subjects had been tested for molecular defects in 6 FH-related genes: low-density lipoprotein receptor (LDLR), apolipoprotein B (APOB), PCSK9, low-density lipoprotein receptor–related protein 1, apolipoprotein E, and lysosomal acid lipase as previously described (14). Molecular defects in LDLR were categorized as null, defective, and not classified according to the JoJo genetics database (15).
We included patients >18 years of age without previous manifestations of ASCVD who were totally asymptomatic. We excluded patients with homozygous FH. Subjects underwent CAC determination either as part of 2 previous protocols evaluating the presence of subclinical atherosclerosis in patients with FH (4,5) or as clinical routine at the lipid clinic. Both CAC evaluation and clinical follow-up were approved by the institutional ethics committee, and all subjects provided written informed consent.
Baseline clinical information was obtained from a standardized health questionnaire and from electronic medical records. Blood pressure was obtained in the sitting position. Hypertension was defined as blood pressure ≥140/90 mm Hg or the use of antihypertensive medications. Diabetes mellitus was defined as either fasting blood glucose ≥126 mg/dl or the use of glucose-lowering drugs. Smokers were defined as subjects who consumed tobacco products within the past year. Family history of early coronary heart disease (CHD) was considered if subjects had first-degree relatives of male or female sex who presented with CHD before the age of 55 or 65 years, respectively.
Weight (kilograms) and height (meters) were measured using a standard physician’s scale and a stadiometer, and body mass index was calculated. Achilles tendon xanthomas were clinically determined by 1 experienced observer as focal nodularities or if tendons were diffusely enlarged by a laterolateral examination as previously described (16).
The cholesterol-years score, a marker of exposure to cholesterol during life, was calculated by multiplying the age at FH diagnosis (in years) by pre-treatment total cholesterol (in milligrams per deciliter). Then the total cholesterol level after treatment was multiplied by the years of treatment. Both values were added, and the result was the final cholesterol-years score, as previously described (17).
High-dose statin treatment was defined as atorvastatin dose ≥40 mg or rosuvastatin dose ≥20 mg. Ezetimibe was prescribed at the attending physician’s discretion.
Study subjects were contacted annually by telephone and/or by regular consultations at the lipid clinic. Follow-up time was defined as the period between CAC evaluation and the last patient contact (telephone call or outpatient visit) or the date of first ASCVD event. There was no loss to follow-up.
CAC was assessed using prospective electrocardiographic gating with 400-m gantry rotation, 3-mm collimation, 120-kV tube voltage, and 300-mA tube current using a 64- or 320-multidetector-row computed tomographic scanner (Toshiba). CAC severity was measured using the Agatston method and expressed as coronary calcium scores as previously described (4). CAC severity was categorized as scores of 0, 1 to 100, and >100, as previously done (18).
ASCVD event definition
Major atherosclerotic cardiovascular events (MACE) events were defined as fatal and nonfatal myocardial infarction, fatal and nonfatal stroke, unstable angina requiring revascularization, and elective myocardial revascularization due to angina pectoris or symptom equivalents, positive myocardial stress test results, or severe obstruction on computed tomographic angiography when this was performed. Myocardial infarction was defined according to international guidelines as cardiac troponin elevation plus at least 1 of the following: symptoms of ischemia, new or presumed new significant ST-T-wave changes or left bundle branch block on 12-lead electrocardiography, development of pathological Q waves on electrocardiography, imaging evidence of new or presumed new loss of viable myocardium or regional wall motion abnormality, and intracoronary thrombus detected on angiography (19). Unstable angina was defined as rest angina or new-onset severe angina without troponin elevation but that required hospitalization and coronary revascularization (20). Elective myocardial revascularization was defined as percutaneous coronary intervention or coronary artery bypass grafting as indicated by the treating physician because of stable angina pectoris or equivalent symptoms refractory to medical treatment and/or positive myocardial stress testing. Ischemic stroke was defined as persistent neurological dysfunction with documentation on computed tomography or magnetic resonance imaging. Hard endpoints were defined as fatal or nonfatal myocardial infarction, unstable angina requiring revascularization, and fatal or nonfatal stroke. Soft endpoints were defined as elective myocardial revascularizations. All cardiovascular events were independently adjudicated by 2 cardiologists (M.H.M., R.D.S.) blinded to patients’ CAC results.
Normality was assessed using the Kolmogorov-Smirnov test and visual inspection. Continuous variables are expressed as mean ± SD or median and interquartile range, as appropriate, and categorical variables as absolute numbers and percentages. Associations between categorical variables were analyzed using the chi-square test. Associations between continuous variables were analyzed using the paired Student t test or Mann-Whitney U test as appropriate. To improve data distribution, CAC score was transformed using the formula log(CAC score + 1), using a logarithm of base 10, as previously reported (4). Kaplan-Meier curves were built to show the association of CAC categories 0, 1 to 100, and >100 with ASCVD events, and the groups were compared using the log-rank test. Univariate survival analysis was performed using Cox regression models and expressed as hazard ratios and their 95% confidence intervals (CIs). Multivariate Cox regression was performed with variables that were associated with ASCVD events in univariate analysis using the enter method selection. Because only CAC remained significantly associated with events in the development of the multivariate models, we performed a post hoc sensitivity analysis of the potential influence of each other individual predictor on the association between CAC and events. This was performed by the construction of multivariate models in which each individual predictor significantly associated with events was forced into a model together with CAC score. In addition, for each stratum the annualized incidence of events per 1,000 patients per year was calculated. Additionally, we performed sensitivity analysis of the association of the presence versus absence of CAC for the prediction of ASCVD by stratifying the population across sex, age groups (≤45 and >45 years), median baseline and on-treatment LDL-C, and type of LDLR defect. To test a possible impact of additional LDL-C lowering with PCSK9 inhibitors on ASCVD events according to CAC strata, 5-year numbers needed to treat to prevent 1 event were calculated considering the observed absolute rate of events per CAC stratum and an average 20% relative risk reduction of these events with treatment as was seen in 2 studies with these medications (12,13). All analyses were performed using SPSS Statistics 20.0 (IBM, Armonk, New York) and Stata 14.0 (StataCorp, College Station, Texas), and a p value <0.05 was considered to indicate statistical significance.
Table 1 shows clinical and laboratory characteristics of the entire study population as well as according to CAC categories. The final study population included 206 subjects with a mean age of 45 ± 14 years and 131 women (63.6%). Except for 1 case with a mutation on APOB, all subjects had mutations on LDLR (99.5%). Of the latter, 48.3%, 42.9%, and 8.8% were considered as null, defective, and not classified, respectively. At baseline, 142 subjects (68.9%) were receiving statins (mean use duration 7.7 ± 6.9 years). At the end of follow-up, 199 (96.6%) were receiving statins, 164 (79.6%) at high doses, and 130 (64%) also used ezetimibe. In 7 cases (3.3%) statins were not used (2 because of statin intolerance and 5 because of poor adherence to treatment).
Overall, 101 patients (49%) had CAC scores of 0, whereas 62 (30%) had CAC scores of 1 to 100 and 43 (21%) had CAC scores >100. Patients with FH with mild and moderate CAC were significantly older compared with those with CAC scores of 0, though distribution of age across CAC groups remained heterogeneous. As shown in Table 1, 19% of those with CAC scores of 0 were older than 50 years of age, whereas a similar proportion (16.2%) of those with CAC scores >100 were younger than 50 years of age. Furthermore, there was an ascending gradient regarding the prevalence of hypertension, family history of premature CHD, the prevalence of corneal arcus, cholesterol-year scores, triglycerides, and glucose from CAC scores of 0 to 1 to 100 and >100. Highest baseline LDL-C levels were seen in those with CAC scores >100. High-dose statin and ezetimibe use were higher in subjects with CAC scores >0, but no differences were seen in on-treatment LDL-C values among the groups.
During a median follow-up period of 3.7 years (interquartile range: 2.7 to 6.8 years), a total of 15 MACE (7.2%) were observed, representing 16.6 events per 1,000 person-years. Almost half of the events (46.6%) were hard endpoints (3 nonfatal and 2 fatal myocardial infarctions, 1 unstable angina requiring revascularization, and 1 fatal stroke), and the others were elective myocardial revascularizations (see Online Table 1 for full description of events).
Figure 1 shows the Kaplan-Meier curves for MACE according to CAC category. The corresponding ASCVD per 1,000 person-years were 0, 26.44 (95% CI: 12.9 to 51.8), and 44.07 (95% CI: 26.0 to 104.1), respectively, for CAC scores of 0, 0 to 100, and >100. Online Figures 1 and 2 show the Kaplan-Meier curves for hard events only, stratified by the extent or presence of CAC. The association between CAC and events remained significant even after restricting the analysis to hard events. On univariate analysis, ASCVD events were associated with male sex, family history of premature CHD, corneal arcus, high-density lipoprotein cholesterol, and log-transformed CAC (Table 2). In the development of a multivariate Cox regression model, only log-transformed CAC (p = 0.001) remained independently associated with incident ASCVD events. This remained essentially unchanged in a sensitivity analysis in which each of the other variables found to be significantly associated with events were forced into the model with log-transformed CAC (Online Table 2). We performed additional Cox regression models adjusted for the type of LDLR mutation defect, and LDL-C and log-transformed CAC remained associated with incident ASCVD events. Also, we included the cholesterol-year score in a model containing log-transformed CAC, and results remained essentially unchanged.
In sensitivity analyses, the presence of CAC was significantly associated with higher ASCVD event rates across sex, age groups (≤45 and >45 years), and median baseline and on-treatment LDL-C (Online Table 3).
Considering a 20% relative risk reduction in ASCVD events with the addition of a PCSK9 inhibitor to standard lipid-lowering therapy as previously seen (12,13) 5-year numbers needed to treat of 38 and 23 would be estimated for CAC scores of 1 to 100 and >100, respectively.
In our study of a sizable population of patients with molecularly proven heterozygous FH, underlying coronary atherosclerosis burden as determined by assessment of CAC, an established surrogate of coronary atherosclerotic disease (21), was significantly heterogeneous. Furthermore, all ASCVD events occurred among those with detectable subclinical coronary atherosclerotic disease, with a graded association of CAC severity with clinical event risk irrespective of underlying risk factor burden. Most important, absence of CAC, encountered in roughly one-half of studied subjects, was associated with no events occurrence during follow-up.
It is well established that FH is associated with elevated relative and absolute lifetime cardiovascular risk. Before lipid-lowering therapy with statins was used routinely, a 3.19-fold excess risk for CHD mortality was reported among a heterozygous FH population in the UK Simon Broome FH registry (22). In the general Danish population, a definite or probable FH diagnosis was associated with adjusted odds ratios of 13.2 (95% CI: 10.0 to 17.4) and 10.3 (95% CI: 7.8 to 13.8) for coronary artery disease in subjects receiving or not receiving lipid-lowering therapy, respectively (23). Data from the Netherlands show that even with statin therapy, treated patients with FH still have a 2-fold greater incidence of coronary artery disease events than their nonaffected relatives (3), and this suggests a need for more aggressive management.
However, at the same time, the risk for ASCVD among patients with FH is variable, and many subjects may not develop cardiovascular disease at all (2,24). Although there are now substantial data highlighting that a significant proportion of those with multiple risk factors, significant dyslipidemia, higher inflammatory burden, and diabetes have no underlying CAC (18,25,26) and generally are at much lower risk than predicted, there is little information on whether a similar pattern of subclinical atherosclerotic disease burden and subsequent outcome exist in this very high risk population.
Our study adds to the current body of research by demonstrating that coronary atherosclerotic disease burden is extremely variable in middle-aged heterozygous FH patients, with nearly half having no detectable CAC despite lifetime exposure to significantly elevated LDL-C, whereas 30% and 20% had mild to moderate and moderate to severe CAC, respectively. Apart from highlighting that similar to the general population, subclinical coronary atherosclerosis is heterogeneously distributed among subjects with FH (4,6,8), to the best of our knowledge this is the first study to assess the value of CAC presence and burden to predict ASCVD events in primary prevention asymptomatic FH patients receiving standard guideline recommended lipid-lowering therapy. This heterogeneity can indeed indicate one’s susceptibility to clinical events onset independently of classic risk factors. Despite this heterogeneity, subjects with FH have a higher prevalence and intensity of CAC compared with normolipidemic matched subjects (4). In this study, a higher CAC burden was associated with characteristics associated with a greater risk for ASCVD events in FH, such as male sex, older age, hypertension presence, and higher baseline LDL-C and glucose levels (27). This association persisted in sensitivity analyses in higher risk subgroups determined by male sex, older age, and more elevated baseline and on-treatment LDL-C levels, higher cholesterol-year score, a surrogate of lifetime exposure to high LDL-C levels, as well as null LDLR defects, which have been previously associated with a greater CAC burden (28).
It is critical to realize that the most important finding of this study was that no events occurred in those subjects without CAC despite the proven diagnosis of FH and the residual elevated on-treatment LDL-C levels. This observation is consistent with prior reports highlighting the value of absence of CAC (power of zero) (21,29). These findings have important implications. There is consensus that subjects with FH must be treated with standard lipid-lowering therapy such as statins and ezetimibe to reduce the usual extremely elevated LDL-C concentrations (1,2,30), and this may significantly reduce ASCVD burden (3). However, despite benefits of the latter, data from a contemporary prospective study show that patients with molecularly proven FH seldom attain these LDL-C levels recommended for ASCVD prevention (31), as only 43% in our study attained LDL-C <130 mg/dl, and 12% attained LDL-C <100 mg/dl, despite maximization of these therapies. The advent of PCSK9 inhibitors has added a very effective tool for additional LDL-C reduction in subjects with FH, and these drugs are now recommended to treat severe patients with FH (2,10,32) who persist with inadequate cholesterol levels. The National Lipid Association recommends PCSK9 inhibitors to further reduce LDL-C in patients with molecularly proven FH ages 40 to 79 years when residual LDL-C and non–high-density lipoprotein cholesterol are still ≥70 and 100 mg/dl and in patients ages 18 to 39 years when LDL-C and non–high-density lipoprotein cholesterol are ≥100 and 130 mg/dl, even in the absence of other uncontrolled risk factors (32). That would be the case for patients in our study. However, the elevated monetary costs of these medications preclude their widespread use (11).
A potential implication of our study findings is that in the current environment of rising health care costs with finite resources, there is evolving consensus to develop strategies for appropriate resource allocation, by accurately identifying who will more likely benefit as well as subjects among whom the yield of expensive pharmacotherapies may be limited. Although we are not recommending cessation of statins for patients with CAC scores of 0, the favorable prognosis in intermediate-term follow-up extending to 4 years may provide a rationale for CAC testing–guided decision making for consideration of novel therapies such as PCSK9 inhibitors. At the patient level, CAC assessment may help further “personalize” risk to weigh expected benefits and costs of PCSK9 inhibitor treatment, whereas at the population level, CAC assessment as an “interventional prevention” tool may help prioritize decisions about PCSK9 inhibitors where their use may produce the highest yield, especially in cost-constrained health care systems.
Our study must be considered within the context of its design. First, the overall population in the study might have limited power for precise estimations. Despite the limited precision of estimates, the results are robust even with the limited sample size. The relatively short follow-up time could limit our conclusion for a longer period of time, but it reinforces that CAC is a robust marker that can indicate relatively short time cardiovascular disease risk. Second, half of our adjudicated events were soft events (coronary revascularization). Nevertheless, those events were driven mostly by a combination of symptoms and positive ischemia on stress testing. Moreover, the results remained essentially unchanged even in a sensitivity analysis with only hard events, though this needs to be read as exploratory because of the exceedingly small number of events. Third, the on-treatment residual LDL-C level of our population was still higher than recommended values for subjects with FH. Still, most of the subjects were treated with high-dose statins, and a significant proportion were also on ezetimibe. In fact, our on-treatment LDL-C levels and occurrence of MACE are in accordance with similar data from the contemporary much larger SAFEHEART registry (31,33). This inability to reach target LDL-C most likely demonstrated the limitation of current strategies in this population rather than issues with the prescription of currently available agents. Fifth, the study subjects were not randomized to CAC evaluation. However, the robust improvement in risk discrimination suggests indeed that evaluation of subclinical atherosclerosis might be useful in risk stratification in FH, as has been suggested (1,2). Not only that, but our multivariate analysis had limited power to include additional risk factors in the adjusted models. Finally, we were not able to calculate the SAFEHEART-RE risk score, the only prospective ASCVD risk score available for patients with FH, although not validated outside Spain (33), from our patients because we did not have lipoprotein(a) values available from most of our subjects.
The presence and absence of CAC are variable among patients with molecularly proven heterozygous FH receiving standard lipid-lowering therapy. More important, higher CAC was strongly associated with ASCVD in primary prevention, with no events reported among those without CAC. These findings may have important implication for advanced risk stratification and allocation of more aggressive lipid-lowering therapy treatments in this population. Longer term follow-up in a greater number of subjects is necessary to confirm our results.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with heterozygous FH have elevated lifetime risk for ASCVD, necessitating aggressive LDL-C-lowering therapy. However, adequate LDL-C levels are seldom attained with high-intensity statins and ezetimibe, and these patients are potential candidates for PCSK9 inhibitors.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: The detection of subclinical coronary atherosclerosis burden might help in stratifying near-term ASCVD risk in subjects with FH, allowing more informed individual management decision making and population-level resource allocation.
TRANSLATIONAL OUTLOOK 1: Overall, nearly half of the middle-aged patients with FH had no underlying CAC, suggesting significant heterogeneity in coronary atherosclerosis in this high-risk group.
TRANSLATIONAL OUTLOOK 2: After a median of 3.7 years (interquartile range: 2.7 to 6.8 years), CAC was independently associated with ASCVD events in patients with heterozygous FH receiving standard lipid-lowering therapy. The absence of CAC was associated with no occurrence of ASCVD events during follow-up. CAC testing might help in stratifying near-term ASCVD risk in subjects with FH, allowing more informed individual management decision making and population-level resource allocation to help identify higher and lower risk patients with FH, allowing cost-effective use of PCSK9 inhibitors in appropriately high-risk patients with FH.
Dr. Santos has received honoraria related to consulting, speaking, and research activities from Akcea, Amgen, AstraZeneca, Biolab, Esperion, Kowa, Merck, Novo-Nordisk, Pfizer, and Sanofi/Regeneron. Dr. Miname has received honoraria for speaking activities from Amgen. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atherosclerotic cardiovascular disease
- coronary artery calcium
- coronary heart disease
- confidence interval
- familial hypercholesterolemia
- low-density lipoprotein cholesterol
- major atherosclerotic cardiovascular event(s)
- proprotein convertase subtilisin kexin type 9
- Received June 26, 2018.
- Revision received September 17, 2018.
- Accepted September 20, 2018.
- 2018 American College of Cardiology Foundation
- Gidding S.S.,
- Ann Champagne M.,
- de Ferranti S.D.,
- et al.
- Santos R.D.,
- Gidding S.S.,
- Hegele R.A.,
- et al.
- Besseling J.,
- Hovingh G.K.,
- Huijgen R.,
- Kastelein J.J.,
- Hutten B.A.
- Perez de Isla L.,
- Alonso R.,
- Muniz-Grijalvo O.,
- et al.
- Mortensen M.B.,
- Falk E.,
- Li D.,
- et al.
- Hlatky M.A.,
- Kazi D.S.
- Ridker P.M.,
- Rose L.M.,
- Kastelein J.J.P.,
- et al.
- Jannes C.E.,
- Santos R.D.,
- de Souza Silva P.R.,
- et al.
- ↵JoJo genetics database. Available at: http://www.jojogenetics.nl/wp/database. Accessed November 2018.
- Mangili L.C.,
- Miname M.H.,
- Silva P.R.S.,
- et al.
- Nozue T.,
- Kawashiri M.A.,
- Higashikata T.,
- et al.
- Roffi M.,
- Patrono C.,
- Collet J.P.,
- et al.
- Braunwald E.,
- Antman E.M.,
- Beasley J.W.,
- et al.
- Nasir K.,
- Bittencourt M.S.,
- Blaha M.J.,
- et al.
- Humphries S.E.,
- Cooper J.A.,
- Seed M.,
- et al.
- Mundal L.,
- Veierod M.B.,
- Halvorsen T.,
- et al.
- Kramer C.K.,
- Zinman B.,
- Gross J.L.,
- et al.
- Vilades Medel D.,
- Leta Petracca R.,
- Carreras Costa F.,
- et al.
- Blaha M.J.,
- Cainzos-Achirica M.,
- Greenland P.,
- et al.
- Nordestgaard B.G.,
- Chapman M.J.,
- Humphries S.E.,
- et al.
- Perez de-Isla L.,
- Alonso R.,
- Watts G.F.,
- et al.
- Orringer C.E.,
- Jacobson T.A.,
- Saseen J.J.,
- et al.
- Perez de Isla L.,
- Alonso R.,
- Mata N.,
- et al.