Skip main navigation

Coronary Artery Calcium for Personalized Allocation of Aspirin in Primary Prevention of Cardiovascular Disease in 2019

The MESA Study (Multi-Ethnic Study of Atherosclerosis)
Originally publishedhttps://doi.org/10.1161/CIRCULATIONAHA.119.045010Circulation. 2020;141:1541–1553

Abstract

Background:

Recent American College of Cardiology/American Heart Association Primary Prevention Guidelines recommended considering low-dose aspirin therapy only among adults 40 to 70 years of age who are at higher atherosclerotic cardiovascular disease (ASCVD) risk but not at high risk of bleeding. However, it remains unclear how these patients are best identified. The present study aimed to assess the value of coronary artery calcium (CAC) for guiding aspirin allocation for primary prevention by using 2019 aspirin meta-analysis data on cardiovascular disease relative risk reduction and bleeding risk.

Methods:

The study included 6470 participants from the MESA Study (Multi-Ethnic Study of Atherosclerosis). ASCVD risk was estimated using the pooled cohort equations, and 3 strata were defined: <5%, 5% to 20%, and >20%. All participants underwent CAC scoring at baseline, and CAC scores were stratified as =0, 1 to 99, ≥100, and ≥400. A 12% relative risk reduction in cardiovascular disease events was used for the 5-year number needed to treat (NNT5) calculations, and a 42% relative risk increase in major bleeding events was used for the 5-year number needed to harm (NNH5) estimations.

Results:

Only 5% of MESA participants would qualify for aspirin consideration for primary prevention according to the American College of Cardiology/American Heart Association guidelines and using >20% estimated ASCVD risk to define higher risk. Benefit/harm calculations were restricted to aspirin-naive participants <70 years of age not at high risk of bleeding (n=3540). The overall NNT5 with aspirin to prevent 1 cardiovascular disease event was 476 and the NNH5 was 355. The NNT5 was also greater than or similar to the NNH5 among estimated ASCVD risk strata. Conversely, CAC≥100 and CAC≥400 identified subgroups in which NNT5 was lower than NNH5. This was true both overall (for CAC≥100, NNT5=140 versus NNH5=518) and within ASCVD risk strata. Also, CAC=0 identified subgroups in which the NNT5 was much higher than the NNH5 (overall, NNT5=1190 versus NNH5=567).

Conclusions:

CAC may be superior to the pooled cohort equations to inform the allocation of aspirin in primary prevention. Implementation of current 2019 American College of Cardiology/American Heart Association guideline recommendations together with the use of CAC for further risk assessment may result in a more personalized, safer allocation of aspirin in primary prevention. Confirmation of these findings in experimental settings is needed.

Clinical Perspective

What Is New?

  • The value of the coronary artery calcium (CAC) score to inform the allocation of aspirin in primary prevention had not been evaluated after publication of updated 2019 meta-analysis data on the efficacy and safety of aspirin and of the 2019 American College of Cardiology/American Heart Association primary prevention guidelines.

  • The pooled cohort equations and, in particular, an estimated cardiovascular risk threshold of >20% failed to identify optimal candidates for aspirin therapy.

  • The CAC score was able to identify subgroups of individuals (overall and within estimated risk strata) in which aspirin would yield net benefit (CAC≥100) and in which aspirin would yield net harm (CAC=0).

What Are the Clinical Implications?

  • In primary prevention individuals considered potentially good candidates for low-dose aspirin therapy (age <70 years, no high bleeding risk, believed to be at higher risk), clinicians may want to quantify the CAC score to guide personalized aspirin allocation.

  • Individuals with CAC≥100 and, in particular, CAC≥400 may be good candidates for aspirin therapy for primary prevention, although the net expected benefit will likely be modest.

  • In the presence of CAC=0, the risk of bleeding is greater than the potential benefit, and aspirin therapy for primary prevention should be avoided.

Introduction

The role of aspirin in the primary prevention of atherosclerotic cardiovascular disease (ASCVD) events is currently controversial.1,2 Since 2018, 3 landmark randomized controlled trials and 2 large meta-analyses have suggested a limited benefit of low-dose aspirin for the primary prevention of ASCVD events, which on average is offset by an increased risk of bleeding in elderly individuals, and in those at higher baseline hemorrhagic risk, as well.3–7 Still, in the largest primary prevention meta-analysis available to date, a nonnegligible 11% relative risk reduction (RRR) in ASCVD events was observed for aspirin.7

Given these updated data, the 2019 American College of Cardiology/American Heart Association (ACC/AHA) guidelines for primary prevention were updated to convey recommendations for aspirin use for primary prevention, from a previous class I to a much more tentative IIb recommendation “among select adults 40 to 70 years of age who are at higher ASCVD risk but not at increased bleeding risk.”8 It remains unclear, however, how clinicians should best identify patients who are likely to derive a net benefit from aspirin therapy in routine primary prevention.

Coronary artery calcium (CAC) is a robust marker of coronary atherosclerotic plaque burden9,10 and a powerful ACC/AHA guideline–endorsed prognostic tool for prediction of ASCVD events and personalized allocation of statin therapy.8,11 It has been suggested that CAC may also have value guiding the allocation of other preventive pharmacotherapies, including aspirin, across estimated ASCVD risk groups.12–14 A 2014 analysis in the MESA study (Multi-Ethnic Study of Atherosclerosis) suggested that a CAC score ≥100 could help identify patients most likely to derive net benefit from chronic aspirin therapy, whereas in patients with CAC=0, aspirin would yield net harm.14 However, the study used 2009 meta-analysis data on aspirin safety and efficacy15 and did not use observed bleeding data, instead relying on a fixed aspirin-related bleeding risk for all studied subgroups.

The present study thus aimed to (1) describe the implications of the recent 2019 ACC/AHA primary prevention guidelines8 on aspirin eligibility for primary ASCVD prevention purposes and (2) assess the potential value of CAC for guiding allocation of aspirin therapy by using the most recent, highest quality meta-analysis data available on aspirin-related ASCVD RRR and bleeding risk.7

Methods

A detailed description of the research methods used is subsequently presented. Requests to access the study data set from qualified researchers trained in human subject confidentiality protocols may be sent to the Collaborative Health Studies Coordinating Center, University of Washington, at . The analyses that support the findings of the present study are available from the corresponding author on reasonable request.

Study Design

Funded by the National Heart, Lung, and Blood Institute, MESA is an ongoing, observational, community-based, prospective cohort study of 6814 men and women of 4 self-identified racial/ethnic groups (non-Hispanic whites, blacks, Hispanics, Chinese) living in the United States. MESA participants were recruited between 2000 and 2002 in 6 field centers: Wake Forest University in Winston-Salem, NC; Columbia University in New York, NY; The Johns Hopkins University in Baltimore, MD; University of Minnesota in Minneapolis; Northwestern University in Chicago, IL; and University of California in Los Angeles.16 The study was approved by the institutional review committee of each site, and all participants provided written informed consent at study entry. The age range at baseline was 45 to 84 years, and participants had to be free of clinically overt cardiovascular conditions (including ASCVD) to be eligible for inclusion. Since the start of the study, participants have been followed for incident events for a median of 14 years. Further details on the MESA study design have been described elsewhere.10,12,14,16

Study Population

All MESA participants were considered for inclusion in the present study. Participants with missing information on CAC burden, aspirin use, variables used by the pooled cohort equations (PCEs; age, sex, race/ethnicity, systolic and diastolic blood pressure, medication use for hypertension, serum cholesterol levels, history of diabetes mellitus, and tobacco use), and longitudinal follow-up were excluded from the analyses (Figure 1).

Figure 1.

Figure 1. Flow of MESA participants included in the study. *Participants included in the aspirin benefit/harm calculations. MESA indicates Multi-Ethnic Study of Atherosclerosis.

To assess the value of CAC to inform the allocation of aspirin therapy under the 2019 ACC/AHA primary prevention guideline recommendations, additional exclusions were implemented for benefit/harm calculations: participants under chronic treatment with aspirin at baseline (defined as any aspirin dose taken ≥3 times per week), participants ≥70 years of age, and those with evidence of high bleeding risk features8,17,18 at MESA visit 1. The latter included a history of renal failure, severe liver disease, concurrent anticoagulation drug use, and uncontrolled hypertension, which was defined as either a systolic blood pressure >160 mm Hg, or a diastolic blood pressure >100 mm Hg at baseline (MESA visit 1). This defined the “aspirin-naive, <70 years, not high bleeding risk” subpopulation for the benefit/harm calculations. Information on history of upper gastrointestinal pain, gastrointestinal ulcers, bleeding disorders, thrombocytopenia, or chronic use of nonsteroidal anti-inflammatory drugs8,17,18 was not available.

Baseline Clinical Evaluation in MESA

All MESA participants underwent a detailed baseline assessment as part of study visit 1.16 This included assessment of demographics, family history of cardiovascular disease (CVD), tobacco use, personal history of other cardiovascular risk factors and medical conditions, and medication use, which were evaluated by using standardized questionnaires. Standardized measurements of body mass index and systolic and diastolic blood pressure were also performed, and blood samples were drawn from all participants to measure nonfasting levels of total and high-density lipoprotein cholesterol levels.16

Ten-Year ASCVD Risk Estimation

For the purposes of the present study, 10-year ASCVD risk was estimated using the PCE for each participant by using baseline demographics and relevant risk factor information. Three 10-year ASCVD risk strata were defined: <5%, 5% to 20%, and >20%.

CAC Measurement

All MESA participants underwent computed tomographic scanning for CAC assessment as part of visit 1. Full details on the MESA computed tomographic scanning protocol have been reported elsewhere.10,12,14,16 CAC was scored using the Agatston method.19 The following clinically relevant CAC groups were defined: CAC=0, CAC 1 to 99, and CAC≥100. In exploratory analyses, CAC≥400 was also evaluated.

Outcome Definitions and Event Ascertainment

The outcomes of interest for the present study were CVD events (defined as a composite of fatal/nonfatal myocardial infarction [MI], fatal/nonfatal stroke, and other CVD death) and major bleeding events. The rates of coronary heart disease (CHD) events (defined as a composite of fatal/nonfatal MI and CHD deaths) were also calculated and reported. Major bleeding events were identified using International Classification of Diseases, NinthRevision and Tenth Revision (ICD) codes for any bleeding in participants requiring a hospitalization during follow-up (Table I in the Data Supplement). In MESA, study personnel conduct active surveillance to systematically identify hospitalizations and deaths of study participants. When a hospitalization is identified, the full list of ICD codes is abstracted from the discharge record. Validation studies in other databases have shown that a strategy using ICD codes to identify bleeding events accurately rules out major bleeding cases.20

Statistical Analyses

The number and proportion of study participants in whom aspirin could be considered for primary prevention of ASCVD according to the 2019 ACC/AHA guideline recommendations (ie, individuals <70 years of age with no high bleeding risk features) were calculated by using a >20% estimated 10-year ASCVD risk-based definition of higher risk.8

The following calculations were conducted in the subpopulation of aspirin-naive participants <70 years of age and not at high risk of bleeding. First, their baseline characteristics were described, overall and by baseline CAC burden. Chi-square tests were used to compare categorical variables across CAC strata, and analysis of variance and Kruskal-Wallis tests were used to compare normally and nonnormally distributed continuous variables, respectively. The interplay between estimated 10-year ASCVD risk categories and baseline CAC burden was described graphically. Crude 5- and 10-year incidence rates of CVD, CHD, and major bleeding events were also computed, overall, and by estimated 10-year ASCVD risk and baseline CAC, as well. Cox proportional hazards regression models were used to compare the risk of CVD, CHD, and major bleeding events, respectively, for increasing CAC categories, overall and by estimated ASCVD risk strata. Regression analyses were adjusted for study site, the risk factors included in the PCE, and statin use, and used 10-year follow-up data to maximize statistical power and allow for adequate adjustment for potential confounders.

Using post hoc estimates by Zheng et al7 pooled from primary prevention randomized trials of low-dose aspirin versus placebo (excluding studies of participants with peripheral artery disease), a 12% expected 5-year RRR in CVD events with chronic low-dose aspirin therapy was applied to the observed 5-year incidence proportion of CVD events, overall and in strata defined by baseline CAC burden and by baseline estimated ASCVD risk, as well. Using the reciprocal of the absolute risk reduction, the number needed to treat at 5 years (NNT5) was calculated for each group. Similarly, using the observed 5-year rates of major bleeding events in MESA overall and in each CAC and ASCVD risk strata, and assuming a 42% 5-year relative risk increase (RRI) in major bleeding events,7 the absolute risk increase was calculated, and the number needed to harm at 5 years (NNH5) was computed as its reciprocal. The NNT5 and NNH5 were then compared graphically to characterize the benefit/harm balance of aspirin therapy, overall and among baseline CAC strata and estimated ASCVD risk, respectively.

Subgroup analyses by sex were also conducted. To increase the number of observed bleeding events included in the analyses and to evaluate the robustness of the 5-year results, a sensitivity analysis was conducted using 10-year follow-up CVD and bleeding event data, and the Altman-Anderson method was applied to scale back to 5 years for NNT5 and NNH5 calculations.21 A second sensitivity analysis was conducted using meta-analytic RRR and RRI estimates restricted to aspirin primary prevention trials published after 2000.7 These had to be recalculated replicating the methodology by Zheng et al,7 but excluding 2 studies of participants with peripheral artery disease (POPADAD [Prevention of Progression of Arterial Disease and Diabetes]22 and AAA [Aspirin for Asymptomatic Atherosclerosis]23), yielding a pooled RRR of 9% and a pooled RRI of 37%. In a third sensitivity analysis, interim aspirin users during the first 5 years of follow-up were excluded from the benefit/harm calculations.

In a post hoc analysis, the benefit/harm calculations were replicated among MESA participants who were using aspirin at baseline, were <70 years of age, and had no high bleeding risk features. For these calculations, CVD events without aspirin were estimated to be 12% higher than those observed with aspirin, and bleeding events without aspirin were estimated to be 42% lower than those observed with aspirin. Last, in another post hoc analysis, the number and proportion of study participants in whom aspirin could be considered for primary prevention purposes were recalculated by using a ≥100 CAC–based definition of higher risk.8

All statistical analyses were performed using Stata version 15. A P value <0.05 was used as threshold of statistical significance.

Results

Study Participants

Of the 6814 MESA participants, 6470 had data on baseline CAC scores, chronic aspirin use at baseline, the risk factors used by the PCE, and incident events during follow-up (Figure 1). Of these, 1287 were already using aspirin at baseline, 1233 were ≥70 years of age, and 410 had at least 1 high bleeding risk feature. The remaining 3540 participants defined the “aspirin-naive, <70-year-old, not high bleeding risk” subpopulation.

Participants Qualifying for Aspirin Therapy Consideration Using a >20% ASCVD Risk Definition of Higher Risk

Overall, 316 of the 6470 participants (4.9%) were <70 years of age, had no evidence of high bleeding risk features, and had an estimated 10-year ASCVD risk >20%. The majority of these individuals were men (n=250, 79.1%), and their age was typically in the 60- to 69-year range (n=261, 82.6%). Although CAC≥100 was common (n=133, 42.1%), the prevalence of CAC=0 (n=100, 31.7%) and of CAC 1 to 99 (n=83, 26.3%) was also high. Figure I in the Data Supplement displays the proportion of participants qualifying for consideration of aspirin therapy (using a >20% ASCVD risk definition of higher risk) by key baseline characteristics.

Baseline Characteristics of the “Aspirin-Naive, <70 Year, Not High Bleeding Risk” Subpopulation

The mean age of the participants included in the benefit/harm calculations (n=3540) was 56.5 years, 55% were women, and the median estimated 10-year ASCVD risk using the PCE was 5.1% (Table 1). A total of 353 participants (10%) used statins at baseline. The higher the baseline CAC burden, the older the mean age, the lower the proportion of women, and the worse the cardiovascular risk profile of the participants.

Table 1. Baseline Characteristics of Study Participants Not Using Aspirin at Baseline, Age <70 Years, and With No High Bleeding Risk Features

CharacteristicsCAC Score
All01–99≥100
N3540 (100.0)2219 (62.7)871 (24.6)450 (12.7)
Age, y56.5 (7.2)54.9 (6.9)58.3 (6.9)61.2 (6.4)
Women1952 (55.1)1412 (63.6)393 (45.1)147 (32.7)
Race or ethnicity
 Non-Hispanic white1215 (34.3)693 (31.2)318 (36.5)204 (45.3)
 Black1029 (29.1)692 (31.2)240 (27.6)97 (21.6)
 Hispanic857 (24.2)566 (25.5)198 (22.7)93 (20.7)
 Chinese439 (12.4)268 (12.1)115 (13.2)56 (12.4)
Body mass index, kg/m228.5 (5.7)28.4 (5.8)28.6 (5.4)28.8 (5.5)
Current smoker544 (15.4)310 (14.0)151 (17.3)83 (18.4)
Family history of atherosclerotic cardiovascular disease1340 (39.9)763 (36.0)372 (45.1)220 (48.8)
Diabetes mellitus357 (10.1)173 (7.8)100 (11.5)84 (18.7)
Systolic blood pressure, mm Hg120 (17)118 (17)123 (16)126 (17)
Diastolic blood pressure, mm Hg72 (10)71 (10)73 (9)74 (9)
Medication use for hypertension961 (27.2)510 (23.0)274 (31.5)177 (39.3)
Low-density lipoprotein cholesterol, mg/dL120 (32)118 (30)123 (34)122 (33)
High-density lipoprotein cholesterol, mg/dL51 (15)52 (15)49 (14)48 (15)
Statin use at baseline353 (10.0)163 (7.4)110 (12.6)80 (17.8)
Estimated 10-year atherosclerotic cardiovascular disease risk*5.1 (2.4–10.0)3.7 (1.6–7.6)7.1 (4.0–12.1)10.7 (6.2–17.1)

Values are presented as n (%), mean (SD), or median (interquartile range). All P values for comparisons across CAC categories were <0.001, with the exception of body mass index (P=0.03). CAC indicates coronary artery calcium.

*The pooled cohort equations were used.

Figure 2 displays the interplay between baseline estimated 10-year ASCVD risk and CAC burden in these individuals. The higher the CAC burden, the more frequent intermediate and high 10-year ASCVD risk estimations, and vice versa. CAC=0 was highly prevalent across all estimated ASCVD risk strata, including a 32% prevalence in the >20% estimated 10-year risk group. In individuals at 5% to 20% estimated risk, 49% had CAC=0, and 18% CAC≥100.

Figure 2.

Figure 2. Interplay between 10-year estimated ASCVD risk and baseline CAC in participants who were aspirin-naive, <70 years of age and not at high bleeding risk. Values are presented in percentages. The 10-year ASCVD risk was estimated by using the pooled cohort equations. ASCVD indicates atherosclerotic cardiovascular disease; and CAC, coronary artery calcium.

Incident Events During Follow-Up

Overall crude rates per 1000 person-years of CVD, CHD, and major bleeding events were 3.52 (95% CI, 2.73–4.53), 2.34 (95% CI, 1.72–3.19), and 1.34 (95% CI, 0.89–2.02), respectively (Table 2). For CVD and CHD events, the higher the baseline CAC burden the higher the event rates, with a high of 12.4 per 1000 person-years CVD events in the CAC≥100 group. The same was true for CVD events by CAC within estimated ASCVD risk categories. Similar trends were apparent using 10-year follow-up data, although event rates were generally higher (Table II in the Data Supplement).

Table 2. Crude 5-Year Incidence Rates of Cardiovascular Disease, Coronary Heart Disease, and Major Bleeding Events

RiskCAC Score
All01–99≥100
All
Cardiovascular disease3.52 (2.73–4.53)1.39 (0.84–2.30)4.58 (2.92–7.18)12.35 (8.41–18.14)
Coronary heart disease2.34 (1.72–3.19)0.65 (0.31–1.36)2.88 (1.64–5.08)9.92 (6.47–15.22)
Major bleeding1.34 (0.89–2.02)0.83 (0.43–1.60)2.89 (1.64–5.09)0.93 (0.23–3.70)
ASCVD risk <5%
Cardiovascular disease1.07 (0.56–2.05)0.60 (0.23–1.60)2.16 (0.70–6.69)5.58 (1.40–22.33)
Coronary heart disease0.71 (0.32–1.59)0.45 (0.15–1.39)0.72 (0.10–5.08)5.58 (1.40–22.33)
Major bleeding0.59 (0.25–1.43)0.30 (0.08–1.20)0.72 (0.10–5.08)5.61 (1.40–22.41)
ASCVD risk 5%–20%
Cardiovascular disease5.80 (4.32–7.80)2.64 (1.42–4.91)6.13 (3.69–10.16)14.04 (8.95–22.01)
Coronary heart disease3.94 (2.76–5.64)1.05 (0.40–2.81)4.08 (2.19–7.57)11.76 (7.20–19.19)
Major bleeding2.09 (1.28–3.42)1.85 (0.88–3.88)3.67 (1.91–7.05)*
ASCVD risk >20%
Cardiovascular disease6.69 (3.19–14.03)2.90 (0.41–20.61)3.23 (0.46–22.96)12.72 (5.29–30.56)
Coronary heart disease3.80 (1.43–10.13)*3.23 (0.46–22.96)7.55 (2.43–23.40)
Major bleeding1.90 (0.48–7.59)*6.58 (1.64–26.29)*

Value are presented as incidence rates per 1000 person-years. The 95% CIs were calculated using the quadratic approximation to the Poisson log likelihood for the log-rate parameter. Follow-up was censored at 5 years. The 10-year ASCVD risk was estimated by using the pooled cohort equations. ASCVD indicates atherosclerotic cardiovascular disease; and CAC, coronary artery calcium.

*There were no 5-year events in this group.

In multivariable regression analyses, CAC≥100 was independently associated with an increased risk of CVD events in comparison with CAC=0 (hazard ratio, 3.9 [95% CI, 2.5–6.1]) and with CHD events (hazard ratio, 4.9 [95% CI, 2.8–8.5]; Table II in the Data Supplement). Conversely, no association was observed between CAC burden and bleeding events regardless of whether 5- or 10-year follow-up data were used. Although an association between estimated ASCVD risk and bleeding events was not evident using 5-year follow-up data, 10-year follow-up data demonstrated such an association (Table II in the Data Supplement).

NNT5 and NNH5 Analyses

Overall, the estimated NNT5 with low-dose aspirin was 476 (Table 3). The higher the baseline CAC burden, the lower the NNT5, with a low of NNT5=100 among individuals with CAC≥400. The overall NNH5 with aspirin was 355, and there was no clear association between CAC burden and the NNH5 (Table 4). Although the NNT5 exceeded the NNH5 in the overall study population, CAC≥100 and particularly CAC≥400 identified individuals likely to derive a net benefit from aspirin therapy (NNT5 lower than NNH5, Figure 3A). Also, CAC=0 identified individuals in whom aspirin would likely yield net harm (NNT5 greater than NNH5). The NNH5 could not be estimated for participants with CAC≥400 because of the lack of observed bleeding events in this subgroup. The same trends were present in analyses stratified by sex (Figure 3B).

Table 3. Number Needed to Treat With Aspirin During 5 Years to Prevent 1 Cardiovascular Disease Event

RiskObserved Events at 5 yAspirin (Assuming 12% Relative Risk Reduction)
NumberIncidence (%)Incidence (%)Absolute Risk Reduction (%)Number Needed to Treat
Overall
 All601.751.540.21476
 By CAC score
  0150.700.620.081190
  1–99192.282.010.27365
  ≥100265.945.230.71140
  ≥400128.337.331.00100
ASCVD risk <5%
 All90.540.480.061543
 By CAC score
  040.300.260.042778
  1–9931.080.950.13772
  ≥10022.782.450.33300
  ≥40018.337.331.00100
ASCVD risk 5%–20%
 All442.852.510.34292
 By CAC score
  0101.311.150.16636
  1–99153.042.680.36274
  ≥100196.705.900.80124
  ≥40088.607.571.0397
ASCVD risk >20%
 All73.322.920.40251
 By CAC score
  011.541.360.18541
  1–9911.671.470.20499
  ≥10056.145.400.74136
  ≥40037.776.840.93107

Values are presented as number or percentage. Follow-up was censored at 5 years. The 10-year ASCVD risk was estimated by using the pooled cohort equations. ASCVD indicates atherosclerotic cardiovascular disease; and CAC, coronary artery calcium.

Table 4. Number Needed to Treat With Aspirin During 5 Years to Cause 1 Major Bleeding Event (Number Needed to Harm)

CAC Score and RiskObserved Events at 5 yAspirin (Assuming 42% Relative Risk Increase)
NumberIncidence (%)Incidence (%)Absolute Risk Increase (%)Number Needed to Harm
Overall230.670.950.28355
By CAC score
 090.420.600.18567
 1–99121.432.030.60167
 ≥10020.460.650.19518
 ≥40000.00***
By ASCVD risk
 <5%50.300.430.13794
 5%–20%161.041.480.44229
 >20%20.931.320.39256

Value are presented as number or %. Follow-up was censored at 5 years. The 10-year ASCVD risk was estimated by using the pooled cohort equations. ASCVD indicates atherosclerotic cardiovascular disease; and CAC, coronary artery calcium.

*The value could not be computed.

Figure 3.

Figure 3. Number needed to treat with low-dose aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event by baseline CAC score, overall (A) and by sex (B). Values are presented as number of persons. Follow-up was censored at 5 years. Red horizontal lines represent NNH thresholds. Participants with CAC≥400 had zero bleeding events, and the NNH could not be computed. The exploratory NNT for participants with CAC≥400 was computed only overall. CAC indicates coronary artery calcium; CVD, cardiovascular disease; NNH, number needed to harm; and NNT, number needed to treat.

In analyses by estimated 10-year ASCVD risk, the NNT with aspirin was 1543 among individuals at <5% risk, 292 among individuals at 5% to 20% risk, and 251 among individuals at >20% risk (Table 3). The corresponding NNH5 values are presented in Table 4. The NNT5 exceeded the NNH5 in individuals at low and intermediate estimated risk and were very similar in individuals at high estimated risk (Figure 4). In all 3 risk strata, CAC≥100 and CAC≥400 identified subgroups with lower NNT5, and CAC=0 identified individuals in whom aspirin would likely yield net harm.

Figure 4.

Figure 4. Number needed to treat with aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event, by estimated ASCVD risk and CAC. Values are presented as number of persons. Follow-up was censored at 5 years. Ten-year ASCVD risk was estimated by using the pooled cohort equations. The red horizontal lines represent the NNH threshold for each ASCVD risk stratum. ASCVD indicates atherosclerotic cardiovascular disease; CAC, coronary artery calcium; CVD, cardiovascular disease; and NNH, number needed to harm.

Sensitivity Analyses

Similar results were observed in sensitivity analyses using 10-year follow-up data (Tables III and IV in the Data Supplement). However, the NNT5 and NNH5 were closer together in individuals with estimated ASCVD risk >20% and high CAC burden (Figure II in the Data Supplement). The results of sensitivity analyses using alternative meta-analytic RRR (9%) and RRI (37%) estimates based on trials published after 2000 were also consistent with those from the main analyses, the NNT5 being consistently lower than the NNH5 for participants with CAC≥100 and, in particular, with CAC≥400, overall and across all estimated ASCVD risk strata (Tables V and VI and Figure III in the Data Supplement).

Consistent results were also observed in sensitivity analyses excluding 1011 interim aspirin users, which yielded a subcohort with a lower incidence of ASCVD and bleeding events than that included in the main analyses (Tables VII and VIII in the Data Supplement). The NNH5 could not be estimated for participants with CAC≥100 because of the lack of observed bleeding events in this subgroup.

Post Hoc Analyses Among Baseline Aspirin Users

A total of 694 MESA participants were prevalent aspirin users at baseline, <70 years of age and had no high-risk bleeding features. These individuals were older, a higher number of them were male, and had a higher prevalence of diabetes mellitus, treated hypertension, and statin use than those included in the main benefit/harm analyses (Table IX in the Data Supplement). Although there were only 7 bleeding events in this subpopulation, the results of the NNT/NNH analyses were qualitatively consistent with those from the main analyses (Tables X and XI in the Data Supplement).

Participants Qualifying for Aspirin Therapy Consideration Using a CAC≥100 CAC Definition of Higher Risk

Figure IV in the Data Supplement displays the analyses assessing the proportion of participants who would qualify for consideration of aspirin therapy using a CAC-based definition of higher risk. Overall, 9.9% participants in MESA would qualify, including 41.8% of those with a CAC score ≥100.

Discussion

In a contemporary, multiethnic population of individuals who were free of clinically overt CVD at baseline, <70 years of age, and with no high bleeding risk features, bleeding risk was closely associated with estimated ASCVD risk but not with baseline CAC burden. In these individuals, ASCVD risk estimations and, in particular, an ASCVD risk threshold of >20% did not appear to identify individuals expected to derive a net benefit from chronic aspirin therapy. Conversely, CAC≥100 and ≥400 consistently identified subgroups of individuals in which the NNT5 was lower than NNH5 (ie, likely to derive net benefit from aspirin), overall and across estimated ASCVD risk. However, the NNT5 with aspirin to prevent 1 CVD was relatively high (≥97) even among individuals with high CAC burden. Also, CAC=0 consistently identified individuals in whom chronic aspirin therapy would likely yield net harm. These trends were robust in various sensitivity and post hoc analyses and in subgroup analyses by sex.

Recent randomized controlled trials such as ASPREE (Aspirin in Reducing Events in the Elderly),3 ARRIVE (A Study to Assess the Efficacy and Safety of Enteric-Coated Acetylsalicylic Acid in Patients at Moderate Risk of Cardiovascular Disease),4 and ASCEND (A Study of Cardiovascular Events in Diabetes)5 raised concerns regarding the potential for bleeding harm and limited efficacy of aspirin in the primary prevention of ASCVD. In ASPREE, which included 19 114 primary prevention elderly participants (median age 74 years), in comparison with placebo daily treatment with low-dose aspirin, did not prolong disability-free survival or reduce the incidence of a composite secondary cardiovascular end point comprising fatal CHD, nonfatal MI, fatal or nonfatal stroke, or hospitalization for heart failure over 4.7 years of follow-up. On the other hand, there was a significantly higher rate of major hemorrhagic events in the aspirin arm. In ARRIVE, among 12 546 primary prevention participants without diabetes mellitus aged ≥55 years (men) or ≥60 years (women) at moderate cardiovascular risk, low-dose aspirin did not significantly reduce the incidence of the primary study end point (a composite of cardiovascular death, MI, unstable angina, stroke, or transient ischemic attack) in comparison with placebo, but was associated with a 2.1-fold increased risk of gastrointestinal bleeding events, most of which were classified as mild. In ASCEND, among 15 480 patients with diabetes mellitus free of known ASCVD (mean age, 63 years), low-dose aspirin was associated with a 12% reduction in a composite end point comprising MI, stroke, transient ischemic attack, or death from any vascular cause (excluding any confirmed intracranial hemorrhage) but was also associated with a 29% increase in first major bleeding events.

A subsequent meta-analysis pooling these and 10 prior trials comparing low-dose aspirin with placebo among individuals without a history of MI or stroke reported a pooled 11% RRR in a composite cardiovascular outcome combining cardiovascular mortality, nonfatal MI, and nonfatal stroke (hazard ratio, 0.89 [95% CI, 0.84–0.94]), and a 43% RRI in major bleeding events (hazard ratio, 1.43 [95% CI, 1.30–1.56]), as well. The pooled estimates were 12% and 42%, respectively, after excluding studies in which participants had peripheral artery disease at baseline (POPADAD22 and AAA23), and were 9% and 39%, respectively, when only studies published after 2000 were considered.7 Given this updated evidence, in 2019, the ACC/AHA primary prevention guidelines recommended consideration of low-dose aspirin therapy only in “select adults 40 to 70 years of age who are at higher ASCVD risk but not at increased bleeding risk.”8 Although the wording used in the guidelines intentionally left room for personalized medicine, incorporation of patient preferences, and shared decision making, it also left clinicians with limited objective guidance regarding which higher risk patients might receive the greatest benefit and least harm with low-dose aspirin therapy.

In this context of uncertainty, the present study suggests that ASCVD risk estimations using the PCE and, in particular, a 10-year ASCVD threshold of >20%, fail to identify individuals who would derive net benefit from aspirin therapy for primary prevention. Conversely, these results suggest that CAC may be a useful decision-making tool informing aspirin allocation among individuals who are willing to consider therapy to further reduce their ASCVD risk.1,2,24 CAC may allow maximizing the benefits of aspirin while minimizing harms, and future ACC/AHA guidelines could consider specifying CAC thresholds to define higher risk in aspirin allocation decision making.8 However, experimental studies are needed to confirm these observations.

It must be noted that, in the present study, which comprised a relatively low ASCVD risk population (mean age, 56 years; 10% participants with diabetes mellitus; median estimated ASCVD risk, 5%), the calculated NNT5 was relatively high even among individuals with CAC≥400 (NNT5 ≥97). This further reinforces the notion that the value of low-dose aspirin for primary ASCVD prevention may be currently modest in the general primary prevention population. Future studies comprising individuals with a higher prevalence of clinical features associated with increased ASCVD risk (eg, diabetes mellitus) and very high CAC scores may help further identify optimal candidates for aspirin therapy in primary prevention, who would maximize the 9% to 12% RRR of CVD events reported in the recent meta-analysis by Zheng et al.7

In participants with CAC=0, the NNTs were clearly outweighed by the NNHs, even among groups at intermediate and high estimated ASCVD risk. CAC=0 was a frequent finding, and was associated with very low CVD event rates in all risk strata, similar to prior analyses in MESA and in other cohorts, as well.25–29 The 2019 ACC/AHA primary prevention guideline authors indicated that “those with CAC scores of zero will have event rates <7.5%, which can help guide shared decision making about statins or potentially even aspirin.” However, this was based on 2014 MESA data,14 and no formal recommendation for the use of CAC=0 in aspirin decision making was listed. In this context, the present results provide further, updated observational support to include a formal recommendation to use CAC=0 to avoid aspirin therapy in individuals otherwise thought to be higher risk. This is important, because high ASCVD risk estimations may often signal a concomitant high risk of bleeding, in particular, in the long term.

Last, the present results also provide a rationale for future randomized trials of aspirin in which CAC could be used to identify primary prevention populations at very high ASCVD risk. Experimental studies suggest that proton-pump inhibitors reduce gastrointestinal bleeding events in patients treated with aspirin, although the benefit may be limited in the absence of gastrointestinal lesions.30,31 Therefore, a strategy combining aspirin and proton-pump inhibitors could be specifically evaluated in future CAC-based trials, as this might reduce harm while not affecting benefit.

Study Strengths

The present study has important strengths. First, it was informed by the most updated, highest quality meta-analysis data on the efficacy and safety of aspirin in primary prevention settings.7 Also, it expands prior analyses in MESA,14 using actual 5-year rates of major bleeding events, which were computed overall and by CAC burden and estimated ASCVD risk strata. This allowed for a more granular and informative evaluation of the benefit/harm balance than prior MESA analyses on this topic.14 In addition, in comparison with prior analyses, the PCE were now used to estimate 10-year ASCVD risk, which is considered the standard in cardiovascular risk prediction as of 2019. Overall, the study allowed evaluating the implications of the 2019 ACC/AHA primary prevention guideline recommendations relevant to aspirin use in primary preventive routine clinical practice and to fully characterize the role that CAC may have in this setting using the most updated evidence available.

Study Limitations

Some limitations are worth discussing. First, the populations pooled by Zheng et al7 were heterogeneous, and there are differences between those and MESA. This may limit the transferability of the pooled RRR and RRI to the MESA population. In this context, although the calculated NNTs and NNHs were somewhat sensitive to the characteristics of the subset of MESA participants included in the benefit/harm calculations, the robust set of main, subgroup, sensitivity, and post hoc analyses conducted yielded qualitatively consistent findings.

Second, our ability to identify individuals with high bleeding risk was limited because certain features such as history of gastrointestinal ulcers were not assessed at MESA visit 1. Consequently, the proportion of individuals in whom aspirin might be considered may have been slightly overestimated. Nonetheless, such conditions would be expected to be very infrequent in the overall healthy MESA population; therefore, the extent of this potential bias should be small.

Third, MESA was not originally designed to capture bleeding events, and the ascertainment of hospitalized bleeding cases using ICD codes may have led to some false-positive events, as reported by Delate and colleagues.20 Moreover, some minor bleeding events occurring in patients hospitalized for another reason may have been listed in the hospital discharge reports, which would have inflated the observed rates of bleeding events. Nondetection of out-of-hospital deaths attributable to bleeding is also a concern, although such events would be expected to be very infrequent in a population with very low hospitalized bleeding events and very low 5-year death rates. Altogether, this could have biased the bleeding rates (which would most likely be overestimated) and the NNHs (which would most likely be underestimated). Nevertheless, this is unlikely to invalidate the compelling results observed in participants with CAC=0, and would further strengthen the findings of a favorable NNT/NNH balance in participants with high CAC scores. Also, the very low rates of bleeding events observed even in this setting suggest that, when aspirin is used in individuals younger than 70 years of age with no high bleeding risk features (as recommended in recent ACC/AHA guidelines), the overall baseline risk of bleeding is low.

Fourth, subgroup analyses by baseline statin use were precluded by the sample size of the statin user subgroup in MESA and the number of events. Further studies including larger populations of statin users are needed to better understand the potential value of CAC guiding the allocation of aspirin therapy in the setting of concomitant statin use. Zheng and colleagues7 observed that restricting their meta-analysis to randomized trials published later than 2000, which included increasing populations of statin users, still showed a 9% RRR in CVD events. The results of the sensitivity analyses using this more conservative RRR were consistent with those from the main analyses.

Fifth, although, in the benefit/harm analyses, atherothrombotic and bleeding events were assumed to be equivalent, clinicians and particularly patients may have their own perceptions and priorities. Efforts should be made to fully incorporate these into clinician-patient discussions involving consideration of aspirin therapy for primary prevention purposes.

Last, given the observational nature of the study, the present findings should be considered hypothesis generating. Evaluation in a randomized trial setting of the potential value of CAC for guiding a safe, effective allocation of aspirin for primary ASCVD prevention is needed.

Conclusions

The present study suggests that ASCVD risk estimations using the PCE and, in particular, an estimated ASCVD risk threshold of >20% may fail to identify individuals who would derive a net benefit from chronic aspirin therapy. CAC may be a valuable tool for aspirin therapy allocation in the context of current ACC/AHA primary prevention guidelines, potentially aiding clinicians in the implementation of those recommendations. Specifically, detection of CAC≥100 and, in particular, of CAC≥400 might be used to identify asymptomatic individuals (<70 years of age and with no high bleeding risk features) in whom the benefit/harm balance of aspirin is likely to be favorable. However, the 5-year NNT with aspirin would be relatively high in populations at low overall baseline ASCVD risk such as that included in the present study. In addition, detection of CAC=0 may be used to avoid aspirin therapy for primary prevention among individuals with high estimated ASCVD risk. Although studies in populations at higher baseline risk, with higher prevalence of very high CAC scores and greater baseline statin use are needed, the present results suggest that future guidelines could consider including CAC-based recommendations aimed at facilitating personalized, safe aspirin therapy allocation in routine clinical practice. This is particularly true for the inclusion of formal recommendations to consider avoiding aspirin therapy when CAC=0. Confirmation of these findings in experimental studies is warranted.

Acknowledgments

The authors thank the other investigators, staff, and participants of the MESA study (Multi-Ethnic Study of Atherosclerosis) for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.

Footnotes

Sources of Funding, see page 1552

https://www.ahajournals.org/journal/circ

Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.

The Data Supplement is available with this article at https://www.ahajournals.org/doi/suppl/10.1161/CIRCULATIONAHA.119.045010.

Miguel Cainzos Achirica, MD, MPH, PhD, Ciccarone Center for the Prevention of Cardiovascular Disease, The Johns Hopkins Hospital, Blalock 501, 600 North Wolfe Street, Baltimore, MD 21287. Email

References

  • 1. Fernandes A, McEvoy JW, Halvorsen S. “Doctor, should I keep taking an aspirin a day?”N Engl J Med. 2019; 380:1967–1970. doi: 10.1056/NEJMclde1903004CrossrefMedlineGoogle Scholar
  • 2. Raber I, McCarthy CP, Vaduganathan M, Bhatt DL, Wood DA, Cleland JGF, Blumenthal RS, McEvoy JW. The rise and fall of aspirin in the primary prevention of cardiovascular disease.Lancet. 2019; 393:2155–2167. doi: 10.1016/S0140-6736(19)30541-0CrossrefMedlineGoogle Scholar
  • 3. McNeil JJ, Woods RL, Nelson MR, Reid CM, Kirpach B, Wolfe R, Storey E, Shah RC, Lockery JE, Tonkin AM, et al; ASPREE Investigator Group. Effect of aspirin on disability-free survival in the healthy elderly.N Engl J Med. 2018; 379:1499–1508. doi: 10.1056/NEJMoa1800722CrossrefMedlineGoogle Scholar
  • 4. Gaziano JM, Brotons C, Coppolecchia R, Cricelli C, Darius H, Gorelick PB, Howard G, Pearson TA, Rothwell PM, Ruilope LM, et al; ARRIVE Executive Committee. Use of Aspirin to Reduce Risk of Initial Vascular Events in patients at Moderate Risk of Cardiovascular Disease (ARRIVE): a randomised, double-blind, placebo-controlled trial.Lancet. 2018; 392:1036–1046. doi: 10.1016/S0140-6736(18)31924-XCrossrefMedlineGoogle Scholar
  • 5. The ASCEND Study Collaborative Group. Effects of aspirin for primary prevention in persons with diabetes mellitus.N Engl J Med. 2018; 379:1529–1539.doi: 10.1056/NEJMoa1804988CrossrefMedlineGoogle Scholar
  • 6. Mahmoud AN, Gad MM, Elgendy AY, Elgendy IY, Bavry AA. Efficacy and safety of aspirin for primary prevention of cardiovascular events: a meta-analysis and trial sequential analysis of randomized controlled trials.Eur Heart J. 2019; 40:607–617. doi: 10.1093/eurheartj/ehy813CrossrefMedlineGoogle Scholar
  • 7. Zheng SL, Roddick AJ. Association of aspirin use for primary prevention with cardiovascular events and bleeding events: a systematic review and meta-analysis.JAMA. 2019; 321:277–287. doi: 10.1001/jama.2018.20578CrossrefMedlineGoogle Scholar
  • 8. Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd-Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.Circulation. 2019; 140:e563–e595. doi: 10.1161/CIR.0000000000000677LinkGoogle Scholar
  • 9. Rumberger JA, Simons DB, Fitzpatrick LA, Sheedy PF, Schwartz RS. Coronary artery calcium area by electron-beam computed tomography and coronary atherosclerotic plaque area: a histopathologic correlative study.Circulation. 1995; 92:2157–2162. doi: 10.1161/01.cir.92.8.2157LinkGoogle Scholar
  • 10. Detrano R, Guerci AD, Carr JJ, Bild DE, Burke G, Folsom AR, Liu K, Shea S, Szklo M, Bluemke DA, et al. Coronary calcium as a predictor of coronary events in four racial or ethnic groups.N Engl J Med. 2008; 358:1336–1345. doi: 10.1056/NEJMoa072100CrossrefMedlineGoogle Scholar
  • 11. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, Braun LT, de Ferranti S, Faiella-Tommasino J, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines.Circulation. 2019; 139:e1082–e1143. doi: 10.1161/CIR.0000000000000625LinkGoogle Scholar
  • 12. Silverman MG, Blaha MJ, Budoff MJ, Rivera JJ, Raggi P, Shaw LJ, Berman D, Callister T, Rumberger JA, Rana JS, et al. Potential implications of coronary artery calcium testing for guiding aspirin use among asymptomatic individuals with diabetes.Diabetes Care. 2012; 35:624–626. doi: 10.2337/dc11-1773CrossrefMedlineGoogle Scholar
  • 13. Hecht H, Blaha MJ, Berman DS, Nasir K, Budoff M, Leipsic J, Blankstein R, Narula J, Rumberger J, Shaw LJ. Clinical indications for coronary artery calcium scoring in asymptomatic patients: expert consensus statement from the Society of Cardiovascular Computed Tomography.J Cardiovasc Comput Tomogr. 2017; 11:157–168. doi: 10.1016/j.jcct.2017.02.010CrossrefMedlineGoogle Scholar
  • 14. Miedema MD, Duprez DA, Misialek JR, Blaha MJ, Nasir K, Silverman MG, Blankstein R, Budoff MJ, Greenland P, Folsom AR. Use of coronary artery calcium testing to guide aspirin utilization for primary prevention: estimates from the Multi-Ethnic Study of Atherosclerosis.Circ Cardiovasc Qual Outcomes. 2014; 7:453–460. doi: 10.1161/CIRCOUTCOMES.113.000690LinkGoogle Scholar
  • 15. Antithrombotic Trialists’ (ATT) Collaboration. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials.Lancet. 2009; 373:1849–1860.doi: 10.1016/S0140-6736(09)60503-1CrossrefMedlineGoogle Scholar
  • 16. Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, Greenland P, Jacob DR, Kronmal R, Liu K, et al. Multi-Ethnic Study of Atherosclerosis: objectives and design.Am J Epidemiol. 2002; 156:871–881. doi: 10.1093/aje/kwf113CrossrefMedlineGoogle Scholar
  • 17. Guirguis-Blake JM, Evans CV, Senger CA, O’Connor EA, Whitlock EP. Aspirin for the primary prevention of cardiovascular events: a systematic evidence review for the U.S. Preventive Services Task Force.Ann Intern Med. 2016; 164:804–813. doi: 10.7326/M15-2113CrossrefMedlineGoogle Scholar
  • 18. Bibbins-Domingo K; U.S. Preventive Services Task Force. Aspirin use for the primary prevention of cardiovascular disease and colorectal cancer: U.S. Preventive Services Task Force Recommendation Statement.Ann Intern Med. 2016; 164:836–845. doi: 10.7326/M16-0577CrossrefMedlineGoogle Scholar
  • 19. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography.J Am Coll Cardiol. 1990; 15:827–832. doi: 10.1016/0735-1097(90)90282-tCrossrefMedlineGoogle Scholar
  • 20. Delate T, Jones AE, Clark NP, Witt DM. Assessment of the coding accuracy of warfarin-related bleeding events.Thromb Res. 2017; 159:86–90. doi: 10.1016/j.thromres.2017.10.004CrossrefMedlineGoogle Scholar
  • 21. Altman DG, Andersen PK. Calculating the number needed to treat for trials where the outcome is time to an event.BMJ. 1999; 319:1492–1495. doi: 10.1136/bmj.319.7223.1492CrossrefMedlineGoogle Scholar
  • 22. Belch J, MacCuish A, Campbell I, Cobbe S, Taylor R, Prescott R, Lee R, Bancroft J, MacEwan S, Shepherd J, et al; Prevention of Progression of Arterial Disease and Diabetes Study Group; Diabetes Registry Group; Royal College of Physicians Edinburgh. The Prevention of Progression of Arterial Disease and Diabetes (POPADAD) trial: factorial randomised placebo controlled trial of aspirin and antioxidants in patients with diabetes and asymptomatic peripheral arterial disease.BMJ. 2008; 337:a1840. doi: 10.1136/bmj.a1840CrossrefMedlineGoogle Scholar
  • 23. Fowkes FG, Price JF, Stewart MC, Butcher I, Leng GC, Pell AC, Sandercock PA, Fox KA, Lowe GD, Murray GD; Aspirin for Asymptomatic Atherosclerosis Trialists. Aspirin for prevention of cardiovascular events in a general population screened for a low ankle brachial index: a randomized controlled trial.JAMA. 2010; 303:841–848. doi: 10.1001/jama.2010.221CrossrefMedlineGoogle Scholar
  • 24. Marquis-Gravel G, Roe MT, Harrington RA, Muñoz D, Hernandez AF, Jones WS. Revisiting the role of aspirin for the primary prevention of cardiovascular disease.Circulation. 2019; 140:1115–1124. doi: 10.1161/CIRCULATIONAHA.119.040205LinkGoogle Scholar
  • 25. Nasir K, Bittencourt MS, Blaha MJ, Blankstein R, Agatson AS, Rivera JJ, Miedema MD, Miemdema MD, Sibley CT, Shaw LJ, et al. Implications of coronary artery calcium testing among statin candidates according to American College of Cardiology/American Heart Association Cholesterol Management Guidelines: MESA (Multi-Ethnic Study of Atherosclerosis).J Am Coll Cardiol. 2015; 66:1657–1668. doi: 10.1016/j.jacc.2015.07.066CrossrefMedlineGoogle Scholar
  • 26. Mortensen MB, Fuster V, Muntendam P, Mehran R, Baber U, Sartori S, Falk E. Negative risk markers for cardiovascular events in the elderly.J Am Coll Cardiol. 2019; 74:1–11. doi: 10.1016/j.jacc.2019.04.049CrossrefMedlineGoogle Scholar
  • 27. Blaha M, Budoff MJ, Shaw LJ, Khosa F, Rumberger JA, Berman D, Callister T, Raggi P, Blumenthal RS, Nasir K. Absence of coronary artery calcification and all-cause mortality.JACC Cardiovasc Imaging. 2009; 2:692–700. doi: 10.1016/j.jcmg.2009.03.009CrossrefMedlineGoogle Scholar
  • 28. Valenti V, Hartaigh BÓ, Cho I, Schulman-Marcus J, Gransar H, Heo R, Truong QA, Shaw LJ, Knapper J, Kelkar AA, et al. Absence of coronary artery calcium identifies asymptomatic diabetic individuals at low near-term but not long-term risk of mortality: a 15-year follow-up study of 9715 patients.Circ Cardiovasc Imaging. 2016; 9:e003528. doi: 10.1161/CIRCIMAGING.115.003528LinkGoogle Scholar
  • 29. Sarwar A, Shaw LJ, Shapiro MD, Blankstein R, Hoffmann U, Hoffman U, Cury RC, Abbara S, Brady TJ, Budoff MJ, et al. Diagnostic and prognostic value of absence of coronary artery calcification.JACC Cardiovasc Imaging. 2009; 2:675–688. doi: 10.1016/j.jcmg.2008.12.031CrossrefMedlineGoogle Scholar
  • 30. Vaduganathan M, Bhatt DL, Cryer BL, Liu Y, Hsieh WH, Doros G, Cohen M, Lanas A, Schnitzer TJ, Shook TL, et al; COGENT Investigators. Proton-pump inhibitors reduce gastrointestinal events regardless of aspirin dose in patients requiring dual antiplatelet therapy.J Am Coll Cardiol. 2016; 67:1661–1671. doi: 10.1016/j.jacc.2015.12.068CrossrefMedlineGoogle Scholar
  • 31. Moayyedi P, Eikelboom JW, Bosch J, Connolly SJ, Dyal L, Shestakovska O, Leong D, Anand SS, Störk S, Branch KRH, et al. Pantoprazole to prevent gastroduodenal events in patients receiving rivaroxaban and/or aspirin in a randomized, double-blind, placebo-controlled trial.Gastroenterology. 2019; 157:403–412.e5.CrossrefMedlineGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.