Characteristics and cardiovascular disease event rates among African Americans and whites who meet the Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk (FOURIER) trial inclusion criteria

Abstract

Purpose:

Determine the risk for cardiovascular disease (CVD) events among adults with clinically evident CVD who meet the inclusion criteria for the FOURIER clinical trial on PCSK9 inhibition in a real-world database.

Methods:

We analyzed data from 2,072 African-American and 2,972 white REasons for Geographic And Racial Differences in Stroke (REGARDS) study participants 45-85 years of age with clinically evident CVD. Study participants meeting the FOURIER inclusion criteria (one major or two minor cardiovascular risk factors, fasting LDL cholesterol ≥70 mg/dL or non-HDL cholesterol ≥100 mg/dL, triglycerides ≤400 mg/dL, and taking statin) were followed for CVD events (myocardial infarction, stroke, coronary revascularization and CVD death) from baseline in 2003-2007 through 2014.

Results:

Overall, 771 (37.2%) African Americans and 1,200 (40.4%) whites met the FOURIER inclusion criteria. The CVD event rate per 1,000 person-years was 60.6 (95%CI 53.6-67.6) among African Americans and 63.5 (95%CI 57.7-69.3) among whites. The risk for CVD events among adults meeting the FOURIER inclusion criteria was higher for those with a history of multiple cardiovascular events (hazard ratios among African Americans and whites 1.34 [95%CI 1.05-1.71] and 1.34 [1.10-1.63], respectively), a prior coronary revascularization (1.44 [1.13-1.84] and 1.23 [1.00-1.52], respectively), diabetes (1.38 [1.08-1.76] and 1.41 [1.15-1.72], respectively), reduced glomerular filtration rate (1.63 [1.26-2.11] and 1.29 [1.03-1.62], respectively) and albuminuria (1.77 [1.37-2.27] and 1.33 [1.07-1.65], respectively).

Conclusions:

The CVD event rate is high among African Americans and whites meeting the FOURIER inclusion criteria. Characteristics associated with a higher CVD risk may inform the decision to initiate PCSK9 inhibition.

Introduction

In the Further Cardiovascular Outcomes Research with Proprotein Convertase Subtilisin–Kexin Type 9 (PCSK9) Inhibition in Subjects with Elevated Risk (FOURIER) clinical trial, evolocumab reduced the risk for cardiovascular disease (CVD) events compared with placebo among adults 40 to 85 years of age with clinically evident CVD [1]. In addition to having clinically evident CVD, eligibility criteria for the FOURIER trial included one major or two minor cardiovascular risk factors (as described below) and having suboptimal cholesterol levels while taking an optimized regimen of statin therapy.

Prior studies suggest that it may not be cost-effective to treat all individuals who meet the FOURIER trial inclusion criteria with a PCSK9 inhibitor [2–4]. Therefore, determining the CVD event rate and characteristics associated with a higher risk for CVD events among individuals meeting the FOURIER inclusion criteria can contribute to identify sub-groups who may have a larger absolute risk reduction through PCSK9 inhibition. The objectives of the current analysis were to estimate the proportion of adults with clinically evident CVD who meet the inclusion criteria for the FOURIER trial, calculate their CVD event rate and determine characteristics associated with having a higher CVD risk. To achieve these objectives, we used data from a prospective population-based US cohort of African Americans and whites, the REasons for Geographic And Racial Differences in Stroke (REGARDS) study [5]. African Americans have a higher prevalence of cardiovascular risk factors, including hypertension, diabetes and obesity, a lower use of statin therapy, and a higher risk for CVD events compared with whites [6,7]. Therefore, analyses were conducted among African Americans and whites, separately.

Methods

Study population

The REGARDS study enrolled 30,239 African American and white men and women ≥45 years of age who were recruited from all 48 contiguous US states and the District of Columbia between January 2003 and October 2007 [5]. REGARDS study participants have been prospectively followed for CVD events and mortality [5,8]. The REGARDS study protocol was approved by the Institutional Review Boards governing research in human subjects at the participating centers and all participants provided written informed consent.

For the current analysis, we included REGARDS study participants 45 to 85 years of age with clinically evident CVD (as defined below) at baseline who had baseline measurements of total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides, and follow-up for CVD events and mortality. We excluded REGARDS study participants >85 years of age at baseline because this population was not eligible for the FOURIER trial. Participants with missing data on total cholesterol, HDL cholesterol or triglycerides at baseline were excluded as we cannot determine whether they meet the FOURIER trial inclusion criteria.

Baseline assessment in the REGARDS study

Computer-assisted telephone interviews were administered at baseline by trained staff and used to collect participant information including their age, race, gender, cigarette smoking status, prior diagnosed co-morbid conditions (e.g., myocardial infarction [MI], stroke, diabetes, high blood pressure [BP]), history of vascular interventions (e.g., coronary and lower extremity revascularization procedures), and use of oral hypoglycemic and antihypertensive medications, aspirin and insulin.

After completion of the interview, trained health professionals conducted in-home examinations following standardized protocols. Procedures included two BP measurements which were averaged, weight and waist circumference measurements, an electrocardiogram and collection of blood and urine samples. Using blood samples, serum total cholesterol, HDL cholesterol, triglycerides, glucose, high-sensitive C-reactive protein (hsCRP) and creatinine were measured. For participants with fasting triglycerides <400 mg/dL, low-density lipoprotein (LDL) cholesterol was calculated using the Friedewald equation [9]. Non-HDL cholesterol was calculated as total minus HDL cholesterol. Estimated glomerular filtration rate (eGFR) was calculated using information on age, race, gender, and serum creatinine and the Chronic Kidney Disease Epidemiology equation [10]. Albumin and creatinine were measured in urine samples and used to calculate the albumin-to-creatinine ratio (ACR).

Clinically evident CVD at baseline was defined by a history of MI, stroke or peripheral artery disease (PAD) to match the FOURIER trial definition [11]. History of MI was defined by self-report of a prior diagnosis of MI, or evidence of a previous MI on the baseline electrocardiogram. History of stroke was defined by self-report of a prior diagnosis of stroke. PAD was defined by self-report of a prior revascularization procedure in the lower extremities. Hypertension was defined by a mean systolic BP ≥140 mm Hg, mean diastolic BP ≥90 mm Hg or self-report of a prior diagnosis of high BP with concurrent use of antihypertensive medication.

During the in-home examination, staff conducted a review of containers for prescription and over-the-counter medications that participants took during the prior two weeks. Use of a statin, non-statin lipid-lowering medication (ezetimibe, fibrate, niacin and cholestyramine), P2Y₁₂ inhibitor (antiplatelet agent), beta-blocker and renin-angiotensin-aldosterone system (RAAS) inhibitor were defined using medication names recorded during the medication review. Medication dosages were not recorded. The intensity of statin therapy (Supplemental Table 1) was assessed in a subsample of REGARDS study participants who had Medicare pharmacy coverage and a statin prescription fill within 120 days prior to their medication review [12,13].

FOURIER trial inclusion criteria

Inclusion criteria for the FOURIER trial, as operationalized using REGARDS baseline data, included:

1. Having one major or two minor cardiovascular risk factors (Supplemental Table 2).

2. Fasting LDL cholesterol ≥70 mg/dL or non-HDL cholesterol ≥100 mg/dL.

3. Triglycerides ≤400 mg/dL.

4. Statin use.

Data collected during follow-up

Following baseline assessment, REGARDS study participants or proxy respondents were contacted every six months via telephone to determine vital status and the occurrence of coronary revascularization (i.e., coronary artery bypass grafting or percutaneous coronary intervention), stroke and MI events. Suspected events were confirmed using medical records as described previously [8,14]. The main underlying cause of deaths identified during follow-up was determined through interviews with next-of-kin or proxies, medical records in the last year of life, death certificates and autopsy reports [8,15].

We analyzed two composite outcomes to resemble those used in the FOURIER trial. The primary outcome included cardiovascular death, stroke, definite or probable MI or coronary revascularization. The primary outcome in the FOURIER trial also included hospitalization due to unstable angina, but these events were not assessed in the REGARDS study. The secondary outcome included cardiovascular death, stroke or definite or probable MI. Adjudicated CVD events and all-cause mortality in REGARDS were available through December 31, 2014.

Statistical analysis

We calculated the proportion of African Americans and whites 45 to 85 years of age with clinically evident CVD who met the FOURIER inclusion criteria outlined above. The FOURIER trial included participants taking a moderate- or high-intensity statin. Therefore, we calculated the proportion of African American and white REGARDS study participants meeting the FOURIER inclusion criteria who were taking a moderate- or high-intensity statin in the subsample with Medicare pharmacy coverage and a statin prescription fill within 120 days prior to their medication review.

Among REGARDS study participants who met the FOURIER inclusion criteria, we calculated baseline characteristics for African Americans and whites, separately and determined the statistical significance of differences using chi-squared tests, t-tests or Wilcoxon rank sum tests, as appropriate. We also compared the characteristics of participants who met versus did not meet the FOURIER inclusion criteria among African Americans and whites, separately.

The following analyses were restricted to REGARDS study participants who met the FOURIER inclusion criteria. We calculated the cumulative incidence and incidence rate for the primary and secondary composite outcomes, all-cause mortality and each component of the primary outcome (i.e., cardiovascular death, stroke, MI and coronary revascularization) among African Americans and whites, separately. Proportional hazard regression models with progressive adjustment were used to calculate hazard ratios and 95% confidence intervals (CI) for the primary and secondary outcomes, all-cause mortality and each primary outcome component comparing African Americans versus whites. Model 1 included adjustment for age, gender, region of residence, income and education. Model 2 included adjustment for the variables in Model 1 and history of MI, stroke, PAD, a history of multiple cardiovascular events, a prior coronary revascularization, smoking, diabetes, hypertension, metabolic syndrome, body weight, total and HDL cholesterol, triglycerides, eGFR <60 mL/min/1.73 m², ACR ≥30 mg/g, hsCRP >2.0 mg/L, and medication use (i.e., non-statin lipid-lowering medication, aspirin or P2Y₁₂ inhibitor, beta-blocker, and RAAS inhibitor).

We used proportional hazard regression models to calculate hazard ratios and 95% CIs for the primary and secondary outcomes associated with baseline characteristics among African Americans and whites, separately. Many of these characteristics are highly correlated. Therefore, regression models only included adjustment for age, gender, region of residence, income and education to allow to identify baseline characteristics that could be used as risk markers in clinical practice. The statistical significance in the difference between hazard ratios for African Americans and whites was determined by including interaction terms between race and each participant characteristic.

For time-to-event analyses, participants were followed from baseline through their first outcome event, death, loss to follow-up, or December 31, 2014, whichever occurred first. The calculation of cumulative incidence functions and hazard ratios for all outcomes except all-cause mortality were conducted accounting for the competing risk for death [16]. All analyses were conducted using SAS v9.4 (SAS Institute Inc., Cary NC). Statistical significance was defined as a p-value <0.10 for interactions and <0.05 for all other analyses.

Results

Overall, 2,072 Africans American and 2,972 white REGARDS study participants were 45 to 85 years of age, and had clinically evident CVD (Supplemental Figure 1). Of these participants, 771 (37.2%) African Americans and 1,200 (40.4%) whites met the FOURIER trial inclusion criteria (Figure 1). The main reason for not meeting the FOURIER inclusion criteria was not taking a statin, particularly among African Americans. In the subset of REGARD study participants meeting the FOURIER inclusion criteria who had Medicare pharmacy coverage and a statin prescription fill within 120 days prior to their medication review (n=95), 91% were taking a high- or moderate-intensity statin.

Figure 1.

Inclusion criteria for the FOURIER trial among REGARDS study participants with clinically evident cardiovascular disease.

FOURIER: Further cardiovascular OUtcomes Research with proprotein convertase subtilisin/kexin type 9 Inhibition in subjects with Elevated Risk; HDL: high-density lipoprotein; LDL: low-density lipoprotein; REGARDS: REasons for Geographic And Racial Differences in Stroke.

Inclusion criteria for the FOURIER trial were applied sequentially and include:

1. Having one major or two minor cardiovascular risk factors (as defined in Supplemental Table 2).
2. Fasting LDL cholesterol ≥70 mg/dL or non-HDL cholesterol ≥100 mg/dL.
3. Triglycerides ≤400 mg/dL.
4. Statin use.

Among REGARDS study participants meeting the FOURIER inclusion criteria, African Americans were younger and less likely to be male, have a history of MI, a prior coronary revascularization, low HDL cholesterol, be taking a non-statin lipid-lowering medication, and aspirin or a P2Y₁₂ inhibitor compared to whites (Table 1, left panel). African Americans were more likely than whites to have <$25,000 annual household income, less than high school education, a history of stroke, diabetes, a history of multiple cardiovascular events, hsCRP >2.0 mg/L, metabolic syndrome, hypertension, and ACR ≥30 mg/g. Also, African Americans were more likely to be a smoker, have a higher body weight, total, HDL, LDL and non-HDL cholesterol, and triglycerides, and to be taking a RAAS inhibitor compared with whites.

Table 1.

Characteristics of REGARDS study participants with clinically evident cardiovascular disease meeting the FOURIER inclusion criteria (left panel) and FOURIER trial participants (right panel).

	REGARDS study participants			FOURIER trial participantsh
Characteristics	White	African Americans	p-valueg	Evolocumab	Placebo
N	1,200	771		13,784	13,780
Demographics
Age, years, mean (SD)	69.4 (7.9)	66.9 (8.0)	<0.001	62.5 (9.1)	62.5 (8.9)
Men, n (%)	830 (69.2)	361 (46.8)	<0.001	10,397 (75.4)	10,398 (75.5)
Geographic region of residence,a n (%)
Belt	433 (36.1)	233 (30.2)		-	-
Buckle	269 (22.4)	127 (16.5)	<0.001	-	-
Non-belt region	498 (41.5)	411 (53.3)		-	-
<$25,000 annual household income, n (%)	364 (32.0)	410 (56.7)	<0.001	-	-
Less than high school education, n (%)	155 (12.9)	216 (28.0)	<0.001	-	-
History of cardiovascular disease
History of myocardial infarction, n (%)	926 (77.6)	508 (66.7)	<0.001	11,145 (80.9)	11,206 (81.3)
History of stroke,b n (%)	325 (27.2)	322 (41.9)	<0.001	2,686 (19.5)	2,651 (19.2)
Peripheral artery disease, n (%)	139 (11.6)	96 (12.5)	0.55	1,858 (13.5)	1,784 (12.9)
Major cardiovascular risk factors
Diabetes, n (%)	377 (31.5)	378 (49.1)	<0.001	5,054 (36.7)	5,027 (36.5)
Age ≥65 years, n (%)	864 (72.0)	469 (60.8)	<0.001	-	-
Current smoking, n (%)	200 (16.7)	157 (20.4)	0.03	3,854 (28.0)	3,923 (28.5)
Multiple cardiovascular events,c n (%)	383 (31.9)	281 (36.4)	0.04	-	-
Minor cardiovascular risk factors
Coronary revascularization, n (%)	733 (61.4)	305 (40.0)	<0.001	-	-
Low HDL cholesterol,d n (%)	582 (48.5)	322 (41.8)	0.003	-	-
hsCRP >2.0 mg/L, n (%)	595 (50.6)	487 (64.9)	<0.001	-	-
High LDL or non-HDL cholesterol,e n (%)	129 (10.8)	157 (20.4)	<0.001	-	-
Metabolic syndrome,f n (%)	655 (54.6)	469 (60.8)	0.006	-	-
Other characteristics
Weight, Kg, mean (SD)	85.9 (17.9)	90.2 (19.9)	<0.001	85.0 (17.3)	85.5 (17.4)
Hypertension, n (%)	852 (71.4)	682 (88.7)	<0.001	11,045 (80.1)	11,039 (80.1)
Total cholesterol, mg/dL, median (25th-75th pctl)	166 (150, 184)	173 (153, 197)	<0.001	168 (151, 188)	168 (151, 189)
HDL cholesterol, mg/dL, median (25th-75th pctl)	42 (36, 51)	47 (39, 56)	<0.001	44 (37, 53)	44 (37, 53)
LDL cholesterol, mg/dL, median (25th-75th pctl)	91 (79, 105)	99 (85, 121)	<0.001	92 (80, 109)	92 (80, 109)
Non-HDL cholesterol, mg/dL, median (25th-75th pctl)	121 (107, 139)	123 (107, 146)	0.02	-	-
Triglycerides, mg/dL, median (25th-75th pctl)	138 (101, 200)	106 (81, 142)	<0.001	134 (101, 183)	133 (99, 181)
eGFR <60 mL/min/1.73 m2, n (%)	268 (22.3)	181 (23.5)	0.56	-	-
ACR ≥30 mg/g, n (%)	250 (21.6)	241 (33.6)	<0.001	-	-
Medication use
Statin type, n (%)
Atorvastatin	443 (36.9)	310 (40.2)	0.14	-	-
Fluvastatin	39 (3.3)	17 (2.2)	0.17	-	-
Lovastatin	94 (7.8)	75 (9.7)	0.14	-	-
Pravastatin	95 (7.9)	57 (7.4)	0.67	-	-
Rosuvastatin	51 (4.3)	37 (4.8)	0.56	-	-
Simvastatin	480 (40.0)	284 (36.8)	0.16	-	-
Any non-statin lipid-lowering medication, n (%)	221 (18.4)	68 (8.8)	<0.001	-	-
Ezetimibe, n (%)	98 (8.2)	47 (6.1)	0.09	726 (5.3)	714 (5.2)
Fibrate, n (%)	65 (5.4)	7 (0.9)	<0.001	-	-
Niacin, n (%)	75 (6.3)	17 (2.2)	<0.001	-	-
Cholestyramine	1 (0.1)	0 (0.0)	0.42	-	-
Aspirin or P2Y12 inhibitor, n (%)	1,002 (83.5)	591 (76.7)	<0.001	12,766 (92.7)	12,666 (92.0)
Beta-blocker, n (%)	618 (51.5)	404 (52.4)	0.70	10,441 (75.8)	10,374 (75.4)
RAAS inhibitor, n (%)	703 (58.6)	520 (67.4)	<0.001	10,803 (78.4)	10,730 (77.9)

ACR: albumin-to-creatinine ratio; eGFR: estimated glomerular filtration rate; FOURIER: Further cardiovascular OUtcomes Research with proprotein convertase subtilisin/kexin type 9 Inhibition in subjects with Elevated Risk; HDL: high-density lipoprotein; hsCRP: high-sensitivity C reactive protein; LDL: low-density lipoprotein; pctl: percentile; RAAS: renin-angiotensin-aldosterone system; REGARDS: REasons for Geographic And Racial Differences in Stroke; SD: standard deviation.

^aStroke buckle includes coastal North Carolina, South Carolina and Georgia. Stroke belt includes the remaining parts of the stroke buckle states and Tennessee, Mississippi, Alabama, Louisiana and Arkansas.

^bThe FOURIER trial only included nonhemorrhagic stroke.

^cMultiple cardiovascular events is defined by a history of ≥2 myocardial infarction and/or stroke events, or peripheral artery disease in addition to a history of myocardial infarction or stroke.

^dLow HDL cholesterol is defined as HDL cholesterol <40 mg/dL for men and <50 mg/dL for women.

^eHigh LDL cholesterol is defined as fasting LDL cholesterol ≥130 mg/dL. High non-HDL cholesterol is defined as fasting or non-fasting non-HDL cholesterol ≥160 mg/dL.

^f

Metabolic syndrome is defined by having ≥3 of the following components:

• Elevated blood pressure, defined as systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg, or self-reported use of antihypertensive medication.
• Abdominal obesity, defined as waist circumference ≥88 cm among women and ≥102 cm among men.
• Impaired glucose, defined by a fasting serum glucose ≥100 mg/dL (or non-fasting serum glucose ≥140 mg/dL), or self-report of a prior diagnosis of diabetes with current use of insulin or oral hypoglycemic medications.
• Low HDL cholesterol, defined as <40 mg/dL among men and <50 mg/dL among women.
• High triglycerides, defined as fasting triglycerides ≥150 mg/dL.

^gP-value comparing white and African American REGARDS study participants meeting the FOURIER inclusion criteria.

^hCharacteristics of FOURIER trial participants were adapted from Sabatine et al. [1].

African Americans and whites in the REGARDS study who met the FOURIER inclusion criteria were older and less likely to be male, a smoker, and to be taking aspirin or a P2Y₁₂ inhibitor, a beta-blocker and a RAAS inhibitor compared to previously reported characteristics of FOURIER trial participants (Table 1, right panel). African Americans and whites who met the FOURIER inclusion criteria in the REGARDS study were also more likely to have a history of stroke and to be taking ezetimibe compared with FOURIER trial participants. The characteristics of REGARDS study participants 45 to 85 years of age with clinically evident CVD who met versus those who did not meet the FOURIER inclusion criteria by race is shown in Supplemental Table 3.

Risk for outcomes among participants meeting the FOURIER inclusion criteria

The cumulative incidence of the primary and secondary outcomes, all-cause mortality, cardiovascular death, stroke, and MI was similar among African Americans and whites who met the FOURIER inclusion criteria (Figure 2). The cumulative incidence of coronary revascularization was lower among African Americans compared with whites. At 2.2 years of follow-up, African Americans and whites meeting FOURIER trial inclusion criteria in the REGARDS study had higher cumulative incidence of the primary and secondary outcomes compared to previously reported results among FOURIER trial participants. After multivariable adjustment, the risk for the primary and secondary outcomes, all-cause mortality, cardiovascular death, stroke and MI were not statistically significantly different between African Americans and whites meeting the FOURIER inclusion criteria (Table 2). The risk for coronary revascularization was statistically significantly lower among African Americans compared with whites.

Figure 2.

Cumulative incidence for outcomes among REGARDS study participants with clinically evident cardiovascular disease who met the FOURIER inclusion criteria.

FOURIER: Further cardiovascular OUtcomes Research with proprotein convertase subtilisin/kexin type 9 Inhibition in subjects with Elevated Risk; REGARDS: REasons for Geographic And Racial Differences in Stroke.

^a Includes cardiovascular death, stroke, myocardial infarction or coronary revascularization.

^b Also includes hospitalization for unstable angina.

^c Includes cardiovascular death, stroke or myocardial infarction.The cumulative incidence of cardiovascular outcomes at 2.2 years among FOURIER trial participants were adapted from Sabatine et al. [1].

Table 2.

Rates and hazard ratios for cardiovascular outcomes and all-cause mortality among REGARDS study participants who met the FOURIER inclusion criteria.

	Whites	African Americans
	(n=1,200)	(n=771)
Primary outcomea
Events	460	286
Event rate (95% CI) per 1,000 person-years	63.5 (57.7, 69.3)	60.6 (53.6, 67.6)
Hazard ratio (95% CI)
Model 1	1 (reference)	0.98 (0.83, 1.15)
Model 2	1 (reference)	0.92 (0.76, 1.12)
Secondary outcomeb
Events	368	250
Event rate (95% CI) per 1,000 person-years	46.8 (42.0, 51.6)	50.5 (44.3, 56.8)
Hazard ratio (95% CI)
Model 1	1 (reference)	1.08 (0.90, 1.29)
Model 2	1 (reference)	0.99 (0.80, 1.23)
All-cause mortality
Events	488	298
Event rate (95% CI) per 1,000 person-years	55.7 (50.7, 60.6)	54.3 (48.1, 60.5)
Hazard ratio (95% CI)
Model 1	1 (reference)	1.01 (0.86, 1.19)
Model 2	1 (reference)	0.89 (0.74, 1.08)
Cardiovascular death
Events	188	139
Event rate (95% CI) per 1,000 person-years	21.4 (18.4, 24.5)	25.3 (21.1, 29.5)
Hazard ratio (95% CI)
Model 1	1 (reference)	1.18 (0.93, 1.50)
Model 2	1 (reference)	1.00 (0.75, 1.35)
Stroke
Events	117	92
Event rate (95% CI) per 1,000 person-years	13.9 (11.4, 16.4)	17.6 (14.0, 21.2)
Hazard ratio (95% CI)
Model 1	1 (reference)	1.34 (0.99, 1.82)
Model 2	1 (reference)	1.06 (0.73, 1.54)
Myocardial infarction
Events	177	95
Event rate (95% CI) per 1,000 person-years	21.6 (18.5, 24.8)	18.4 (14.7, 22.1)
Hazard ratio (95% CI)
Model 1	1 (reference)	0.88 (0.67, 1.16)
Model 2	1 (reference)	0.87 (0.63, 1.20)
Coronary revascularization
Events	220	89
Event rate (95% CI) per 1,000 person-years	28.6 (24.8, 32.4)	17.6 (13.9, 21.3)
Hazard ratio (95% CI)
Model 1	1 (reference)	0.67 (0.52, 0.87)
Model 2	1 (reference)	0.73 (0.54, 0.99)

CHD: coronary heart disease; CI: confidence interval; FOURIER: Further cardiovascular OUtcomes Research with proprotein convertase subtilisin/kexin type 9 Inhibition in subjects with Elevated Risk; REGARDS: REasons for Geographic And Racial Differences in Stroke.

The median (maximum) follow-up for mortality is 7.9 (11.8) years.

^aIncludes cardiovascular death, stroke, myocardial infarction or coronary revascularization.

^bIncludes cardiovascular death, stroke or myocardial infarction.

Model 1 adjusts for age, gender, geographic region of residence, income and education.

Model 2 adjusts for variables in Model 1 plus history of myocardial infarction, history of stroke, history of peripheral artery disease, multiple cardiovascular events, history of coronary revascularization, smoking, diabetes, hypertension, metabolic syndrome, body weight, total and high-density lipoprotein cholesterol, triglycerides, estimated glomerular filtration rate <60 mL/min/1.73 m², albumin-to-creatinine ratio ≥30 mg/g, highly-sensitive C-reactive protein >2.0 mg/L, and medication use, including any non-statin lipid-lowering medication, aspirin or P2Y₁₂ inhibitor, beta-blocker, and renin-angiotensin-aldosterone system inhibitor.

Characteristics associated with the primary and secondary outcomes

A history of multiple cardiovascular events, a prior coronary revascularization, diabetes, eGFR <60 mL/min/1.73 m², and ACR ≥30 mg/g were associated with a higher risk for the primary outcome among African Americans and whites (Figure 3). Among African Americans, the risk for the primary outcome was also higher among participants taking aspirin or P2Y₁₂ inhibitor, or a beta-blocker. Among whites, the risk for the primary outcome was also higher among men, those with hsCRP >2.0 mg/L and those taking a non-statin lipid-lowering medication. Higher levels of HDL cholesterol were associated with a lower risk for the primary outcome among African Americans and whites. Hazard ratios for the primary outcome associated with participant characteristics were consistent for African Americans and whites (all interaction p-values >0.10). Characteristics associated with the secondary outcome among African Americans and whites meeting the FOURIER inclusion criteria are shown in Supplemental Figure 2.

Figure 3.

Hazard ratios for the primary outcome associated with participant characteristics among African Americans and whites who met the FOURIER inclusion criteria.

ACR: albumin-to-creatinine ratio; eGFR: estimated glomerular filtration rate; FOURIER: Further cardiovascular OUtcomes Research with proprotein convertase subtilisin/kexin type 9 Inhibition in subjects with Elevated Risk; HDL: high-density lipoprotein; hsCRP: high-sensitivity C reactive protein; LDL: low-density lipoprotein; RAAS: renin-angiotensin-aldosterone system.

Hazard ratios adjust for age, gender, geographic region of residence, income and education.

^a Multiple cardiovascular events is defined by a history of ≥2 myocardial infarction and/or stroke events, or peripheral artery disease in addition to a history of myocardial infarction or stroke.

^b Metabolic syndrome is defined by having ≥3 of the following components:

• Elevated blood pressure, defined as systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg, or self-reported use of antihypertensive medication.
• Abdominal obesity, defined as waist circumference ≥88 cm among women and ≥102 cm among men.
• Impaired glucose, defined by a fasting serum glucose ≥100 mg/dL (or non-fasting serum glucose ≥140 mg/dL), or self-report of a prior diagnosis of diabetes with current use of insulin or oral hypoglycemic medications.
• Low HDL cholesterol, defined as <40 mg/dL among men and <50 mg/dL among women.
• High triglycerides, defined as fasting triglycerides ≥150 mg/dL.

^cP-values comparing the hazard ratio among African Americans versus whites.

The primary outcome includes cardiovascular death, stroke, myocardial infarction or coronary revascularization.

Discussion

In the current analysis of a large nationwide cohort study, a substantial proportion of African Americans and whites 45 to 85 years of age with clinically evident CVD met the inclusion criteria for the FOURIER trial. The main reason for not meeting the FOURIER trial inclusion criteria for both African Americans and whites was not being on statin therapy. There were substantial differences in baseline characteristics between African American and white REGARDS study participants who met the inclusion criteria for the FOURIER trial, including demographics, prior history of MI and stroke, and cardiovascular risk factors. Despite these differences, the risk for CVD events was similar among African Americans and whites who met the FOURIER inclusion criteria. Factors associated with a higher risk for CVD events among individuals meeting the FOURIER inclusion criteria were identified and could be used to inform the decision to initiate PCSK9 inhibition in this population.

In the FOURIER trial, PCSK9 inhibition reduced the risk for the primary and secondary outcomes by 15% and 20%, respectively, compared with placebo among patients with clinically evident CVD taking an optimized regimen of statin therapy [1]. In the current study, a large proportion of African Americans and whites with clinically evident CVD were not taking statins and therefore did not meet the FOURIER inclusion criteria. This finding is consistent with prior studies showing suboptimal use of statin therapy and high rates of statin discontinuation following a CVD event, especially among African Americans [17–20]. Results from the current study suggest that despite the suboptimal use of statin therapy, approximately 40% of African Americans and whites with clinically evident CVD may meet the inclusion criteria for the FOURIER trial. Therefore, results from the FOURIER trial is likely generalizable to a large proportion of African Americans and whites with clinically evident CVD.

There were substantial differences in the baseline characteristics of African Americans and whites who met the FOURIER inclusion criteria in the REGARDS study. Findings from the current analysis are consistent with prior studies showing that African Americans are more likely to have a history of stroke [21,22], hypertension [23,24] and albuminuria [25], and less likely to have lower HDL cholesterol [26] compared with whites. African Americans have a higher risk for dying after their first MI compared with whites [27]. This could explain the lower prevalence of a history of MI among African Americans versus whites found in the current analysis. There were also some differences in the characteristics of African American and white REGARDS study participants who met the FOURIER inclusion criteria versus FOURIER trial participants. For example, African Americans and whites who met the FOURIER inclusion criteria in the REGARDS study were more likely to have a history of stroke compared to FOURIER trial participants [1]. This could be explained by FOURIER participants having been recruited using a convenience sampling approach while the REGARDS study enrolled a population-based sample [1,5].

African Americans were underrepresented in the FOURIER clinical trial [1,11]. However, there is no reason to anticipate that results from the FOURIER trial may not be generalizable to African Americans. In the current analysis, African Americans and whites who met the FOURIER inclusion criteria had similar rates for the primary and secondary composite cardiovascular outcomes, all-cause mortality, cardiovascular death, stroke and MI. However, the rate of coronary revascularization was lower for African American versus white REGARDS study participants who met the FOURIER inclusion criteria. This finding is consistent with prior studies showing that African Americans are less likely to receive a coronary revascularization compared with whites, although the reasons behind this finding are not clearly understood [28–31]. Compared with FOURIER trial participants, African Americans and whites in the REGARDS study who met the FOURIER inclusion criteria had higher CVD event rates. In an analysis conducted by Fonarow et al., patients with commercial health insurance or Medicare Supplemental insurance in the Truven MarketScan database who had a history of CVD also had a higher CVD event rate compared with FOURIER trial participants [3]. Along with the current study, the results from the MarketScan database suggest that the absolute risk reduction associated with PCSK9 inhibition among African Americans and whites with clinically evident CVD in the general population may be larger as compared with what could be expected based on event rates in the FOURIER trial [32].

In clinical practice, it may not be cost-effective to treat all individuals who meet the FOURIER trial inclusion criteria with a PCSK9 inhibitor [2–4]. In the current study, having a history of multiple CVD events, a prior coronary revascularization, diabetes, eGFR <60 mL/min/1.73 m² and ACR ≥30 mg/g were associated with a higher risk for CVD events among African American and white REGARDS study participants meeting the FOURIER inclusion criteria. Directing PCSK9 inhibition towards populations with very high CVD risk could result in a larger absolute risk reduction [32]. This could also translate into a lower cost per quality-adjusted life-year saved.

The current study has several strengths including using data from a large population-based cohort of African Americans and whites who resided in all 48 contiguous US states and the District of Columbia with long-term follow-up and a rigorous CVD event adjudication process. Despite these strengths, the current study has potential and known limitations. The FOURIER trial included participants with a history of MI, nonhemorrhagic stroke or PAD. In the current analysis, a history of stroke was defined by self-report of a prior cerebrovascular event regardless of whether this was ischemic or hemorrhagic as this information is not available in the REGARDS study. This could have contributed to the higher prevalence of history of stroke among REGARDS study participants meeting the FOURIER inclusion criteria versus FOURIER trial participants. Also, the definition of PAD in REGARDS was based on self-report of a prior revascularization procedure in the lower extremities. In the FOURIER trial, PAD was defined as intermittent claudication with ankle-brachial index <0.85, peripheral artery revascularization, or amputation due to atherosclerotic disease. Therefore, the current analysis may have underrepresented the population with PAD as compared with the FOURIER trial. While FOURIER subjects were on an optimized statin regime, the statin dosage taken by REGARDS study participants at baseline was unavailable and we included all study participants regardless of dose. In a subgroup analysis of participants meeting the FOURIER inclusion criteria who had Medicare pharmacy coverage, over 90% were taking a high- or moderate-intensity statin. Few REGARDS study participants had Medicare pharmacy coverage at baseline as this program started in 2006 and only provides pharmacy benefits to adults ≥65 years of age and those with end-stage renal disease or disability [33].

In conclusion, a substantial proportion of African Americans and whites with clinically evident CVD met the FOURIER inclusion criteria. Despite differences in baseline characteristics, African Americans and whites meeting the FOURIER inclusion criteria had a similar risk for CVD events. Among African Americans and whites who met the FOURIER inclusion criteria, those with a history of multiple CVD events, a prior coronary revascularization, diabetes, eGFR <60 mL/min/1.73 m² and ACR ≥30 mg/g had a higher risk for CVD events. The results of the current study help to identify subgroups of African Americans and whites who may receive a larger absolute CVD risk reduction with PCSK9 inhibition.