Age-Related Differences in the Noninvasive Evaluation for Possible Coronary Artery Disease: Insights From the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) Trial | Cardiology | JAMA Cardiology | JAMA Network
[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.  Association Between Age and Clinical Outcomes Based on Test Positivity
Association Between Age and Clinical Outcomes Based on Test Positivity

A stress test result was defined as positive if there were ST-segment changes of ischemia or if the test was terminated early (<3 minutes) owing to reproduction of symptoms, arrhythmia, or hypotension during the stress test. For stress echocardiography and stress nuclear, test results were also considered positive with presence of inducible ischemia in at least 1 coronary territory. Any obstructive disease on computed tomography angiography was considered a positive result; obstructive was defined as an at least 50% lesion in left main or at least 70% obstruction in any other coronary artery. Coronary artery calcium (CAC) of at least 100 was considered positive; CAC 0 to 99 was considered negative.

Figure 2.  Time to Event Based on Age and Test Modality
Time to Event Based on Age and Test Modality

A stress test result was defined as positive if there were ST-segment changes of ischemia or if the test was terminated early (<3 minutes) owing to reproduction of symptoms, arrhythmia, or hypotension during the stress test. For stress echocardiography and stress nuclear, test results were also considered positive with presence of inducible ischemia in at least 1 coronary territory. Any obstructive disease on computed tomography angiography was considered a positive result; obstructive was defined as at least 50% lesion in left main or at least 70% obstruction in any other coronary artery. Coronary artery calcium (CAC) of at least 100 was defined as positive; CAC 0 to 99 was considered negative.

Table 1.  Baseline Characteristics by Age Group
Baseline Characteristics by Age Group
Table 2.  Test Characteristics and Results by Age and Test Type
Test Characteristics and Results by Age and Test Type
Table 3.  Association Between Age and Cardiovascular Death/Myocardial Infarction Based on Test Positivity
Association Between Age and Cardiovascular Death/Myocardial Infarction Based on Test Positivity
1.
McClelland  RL, Chung  H, Detrano  R, Post  W, Kronmal  RA.  Distribution of coronary artery calcium by race, gender, and age: results from the Multi-Ethnic Study of Atherosclerosis (MESA).  Circulation. 2006;113(1):30-37. doi:10.1161/CIRCULATIONAHA.105.580696PubMedGoogle ScholarCrossref
2.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
3.
Douglas  PS, Hoffmann  U, Patel  MR,  et al; PROMISE Investigators.  Outcomes of anatomical versus functional testing for coronary artery disease.  N Engl J Med. 2015;372(14):1291-1300. doi:10.1056/NEJMoa1415516PubMedGoogle ScholarCrossref
4.
Douglas  PS, Hoffmann  U, Lee  KL,  et al; PROMISE investigators.  Prospective Multicenter Imaging Study for Evaluation of chest pain: rationale and design of the PROMISE trial.  Am Heart J. 2014;167(6):796-803.e1. doi:10.1016/j.ahj.2014.03.003PubMedGoogle ScholarCrossref
5.
Rich  MW, Chyun  DA, Skolnick  AH,  et al; American Heart Association Older Populations Committee of the Council on Clinical Cardiology, Council on Cardiovascular and Stroke Nursing, Council on Cardiovascular Surgery and Anesthesia, and Stroke Council; American College of Cardiology; and American Geriatrics Society.  Knowledge gaps in cardiovascular care of the older adult population: a scientific statement from the American Heart Association, American College of Cardiology, and American Geriatrics Society.  Circulation. 2016;133(21):2103-2122. doi:10.1161/CIR.0000000000000380PubMedGoogle ScholarCrossref
6.
Alexander  KP, Newby  LK, Cannon  CP,  et al; American Heart Association Council on Clinical Cardiology; Society of Geriatric Cardiology.  Acute coronary care in the elderly, part I: non-ST-segment-elevation acute coronary syndromes: a scientific statement for healthcare professionals from the American Heart Association Council on Clinical Cardiology: in collaboration with the Society of Geriatric Cardiology.  Circulation. 2007;115(19):2549-2569. doi:10.1161/CIRCULATIONAHA.107.182615PubMedGoogle ScholarCrossref
7.
Alexander  KP, Newby  LK, Armstrong  PW,  et al; American Heart Association Council on Clinical Cardiology; Society of Geriatric Cardiology.  Acute coronary care in the elderly, part II: ST-segment-elevation myocardial infarction: a scientific statement for healthcare professionals from the American Heart Association Council on Clinical Cardiology: in collaboration with the Society of Geriatric Cardiology.  Circulation. 2007;115(19):2570-2589. doi:10.1161/CIRCULATIONAHA.107.182616PubMedGoogle ScholarCrossref
8.
Newman  AB, Naydeck  BL, Sutton-Tyrrell  K, Feldman  A, Edmundowicz  D, Kuller  LH.  Coronary artery calcification in older adults to age 99: prevalence and risk factors.  Circulation. 2001;104(22):2679-2684. doi:10.1161/hc4601.099464PubMedGoogle ScholarCrossref
9.
Giallauria  F, Vigorito  C, Ferrara  N, Ferrucci  L.  Cardiovascular calcifications in old age: mechanisms and clinical implications.  Curr Transl Geriatr Exp Gerontol Rep. 2013;2(4):255-267. doi:10.1007/s13670-013-0063-4Google ScholarCrossref
10.
Rai  M, Baker  WL, Parker  MW, Heller  GV.  Meta-analysis of optimal risk stratification in patients >65 years of age.  Am J Cardiol. 2012;110(8):1092-1099. doi:10.1016/j.amjcard.2012.05.048PubMedGoogle ScholarCrossref
11.
Marwick  TH, Mehta  R, Arheart  K, Lauer  MS.  Use of exercise echocardiography for prognostic evaluation of patients with known or suspected coronary artery disease.  J Am Coll Cardiol. 1997;30(1):83-90. doi:10.1016/S0735-1097(97)00148-4PubMedGoogle ScholarCrossref
12.
Herrera  AP, Snipes  SA, King  DW, Torres-Vigil  I, Goldberg  DS, Weinberg  AD.  Disparate inclusion of older adults in clinical trials: priorities and opportunities for policy and practice change.  Am J Public Health. 2010;100(suppl 1):S105-S112. doi:10.2105/AJPH.2009.162982PubMedGoogle ScholarCrossref
13.
Dai  X, Busby-Whitehead  J, Alexander  KP.  Acute coronary syndrome in the older adults.  J Geriatr Cardiol. 2016;13(2):101-108.PubMedGoogle Scholar
14.
Fihn  SD, Gardin  JM, Abrams  J,  et al; American College of Cardiology Foundation; American Heart Association Task Force on Practice Guidelines; American College of Physicians; American Association for Thoracic Surgery; Preventive Cardiovascular Nurses Association; Society for Cardiovascular Angiography and Interventions; Society of Thoracic Surgeons.  2012 ACCF/AHA/ACP/AATS/PCNA/SCAI/STS Guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, and the American College of Physicians, American Association for Thoracic Surgery, Preventive Cardiovascular Nurses Association, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons.  J Am Coll Cardiol. 2012;60(24):e44-e164. doi:10.1016/j.jacc.2012.07.013PubMedGoogle ScholarCrossref
15.
Madhavan  MV, Gersh  BJ, Alexander  KP, Granger  CB, Stone  GW.  Coronary artery disease in patients ≥80 years of age.  J Am Coll Cardiol. 2018;71(18):2015-2040. doi:10.1016/j.jacc.2017.12.068PubMedGoogle ScholarCrossref
16.
Kuller  LH, Lopez  OL, Mackey  RH,  et al.  Subclinical Cardiovascular Disease and Death, Dementia, and Coronary Heart Disease in Patients 80+ Years.  J Am Coll Cardiol. 2016;67(9):1013-1022. doi:10.1016/j.jacc.2015.12.034PubMedGoogle ScholarCrossref
17.
Grundy  SM, Stone  NJ, Bailey  AL,  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(25):e1082-e1143.PubMedGoogle Scholar
18.
Budoff  MJ, Mayrhofer  T, Ferencik  M,  et al; PROMISE Investigators.  Prognostic value of coronary artery calcium in the PROMISE Study (Prospective Multicenter Imaging Study for Evaluation of Chest Pain).  Circulation. 2017;136(21):1993-2005. doi:10.1161/CIRCULATIONAHA.117.030578PubMedGoogle ScholarCrossref
Original Investigation
November 18, 2019

Age-Related Differences in the Noninvasive Evaluation for Possible Coronary Artery Disease: Insights From the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) Trial

Author Affiliations
  • 1Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina
  • 2Department of Cardiovascular Medicine, Mayo Clinic, Rochester, Minnesota
  • 3National Heart, Lung, and Blood Institute, Bethesda, Maryland
  • 4Deborah Heart and Lung Center, Browns Mills, New Jersey
  • 5Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
JAMA Cardiol. 2020;5(2):193-201. doi:10.1001/jamacardio.2019.4973
Key Points

Question  Among patients in the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) trial, did age affect the likelihood of a positive test result for coronary artery disease and its prognostic value, and did this differ according to type of test (anatomic vs functional test)?

Findings  In this prespecified PROMISE trial substudy, test result positivity increased with age regardless of the noninvasive test completed. However, among patients younger than 65 years, anatomic testing provided better prognostic discrimination, whereas among patients 65 years and older, functional testing was able to distinguish future risk, with a significant interaction between age, test type, and prognosis.

Meaning  Positive functional test results vs anatomic test results differ in their association with cardiovascular death or myocardial infarction across patient age. Age-specific approaches to the noninvasive evaluation of coronary artery disease should be further explored.

Abstract

Importance  Although cardiovascular (CV) disease represents the leading cause of morbidity and mortality that increases with age, the best noninvasive test to identify older patients at risk for CV events remains unknown.

Objective  To determine whether the prognostic utility of anatomic vs functional testing varies based on patient age.

Design, Setting, and Participants  Prespecified analysis of the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) study, which used a pragmatic comparative effectiveness design. Participants were enrolled from 193 sites across North America and comprised outpatients without known coronary artery disease (CAD) but with symptoms suggestive of CAD. Data were analyzed between October 2018 and April 2019.

Interventions  Randomization to noninvasive testing with coronary computed tomographic angiography or functional testing.

Main Outcomes and Measures  The composite of CV death/myocardial infarction (MI) over a median follow-up of 25 months.

Results  Among 10 003 PROMISE patients, we included the 8966 who received the noninvasive test to which they were randomized and had interpretable results; 6378 (71.1%) were younger than 65 years, 2062 (23.0%) were between ages 65 and 74 years, and 526 (5.9%) were 75 years and older. More than half of participants were women (4720 of 8966 [52.6%]). Only a minority of patients were of nonwhite race/ethnicity, a proportion that was lower among the older age groups (1071 of 6378 [16.8%] for <65 years; 258 of 2062 [12.5%] for age 65-74 years; 41 of 526 [7.8%] for ≥75 years). Compared with patients younger than 65 years, older patients were more likely to have a positive test result (age 65-74 years: odds ratio, 1.65; 95% CI, 1.42-1.91; age ≥75 years: odds ratio, 2.32; 95% CI, 1.83-2.95), regardless of noninvasive test completed. A positive functional test result was not associated with CV death/MI in patients younger than 65 years (hazard ratio [HR], 1.09; 95% CI, 0.43-2.82) but it was among older patients (age 65-74 years: HR, 3.18; 95% CI, 1.44-7.01; age ≥75 years: HR, 6.55; 95% CI, 1.46-29.35). Conversely, a positive anatomic test result was associated with CV death/MI among patients younger than 65 years (HR, 3.04; 95% CI, 1.46-6.34) but not among older patients (age, 65-74 years: HR, 0.67; 95% CI, 0.15-2.94; age ≥75 years: HR, 1.07; 95% CI, 0.22-5.34; P for interaction = .01). An elevated coronary artery calcium score was predictive of events in patients younger than 65 years (HR, 2.73; 95% CI, 1.31-5.69) but not for older patients (age 65-74 years: HR, 0.44; 95% CI, 0.14-1.42; age ≥75 years: HR, 1.31; 95% CI, 0.25-6.88).

Conclusions and Relevance  Older patients with stable symptoms suggestive of CAD are more likely to have a positive noninvasive test result and more coronary artery calcium. However, only a positive functional test result was associated with risk of CV death/MI. Age-specific approaches to noninvasive evaluation of CAD should be further examined.

Trial Registration  ClinicalTrials.gov identifier: NCT01174550

Introduction

The incidence and prevalence of symptomatic coronary artery disease (CAD) increase with age.1,2 Given the aging population, more older adults are presenting for evaluation of symptoms possibly related to CAD. However, identifying those who are at an increased risk for future cardiovascular (CV) events remains unclear. There are also unique considerations in the performance or interpretation of noninvasive testing for CAD among different age groups of patients. For example, noninvasive testing in older adults may present unique challenges owing to exercise limitations or uninterpretable electrocardiograms (ECGs). Further, the pretest probability of CAD varies with age and affects interpretation of both positive and negative test results.

The Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) randomized outpatients with stable symptoms suggestive of CAD to noninvasive testing with either anatomic (coronary computed tomographic angiography [CTA]) or functional testing (exercise or pharmacologic).3 Over a median 25 months of follow-up, an initial noninvasive testing strategy of anatomic testing was not associated with better clinical outcomes as compared with functional testing. The PROMISE trial provides a unique opportunity to evaluate the prognostic utility of anatomic vs functional testing with age, specifically in older adults. In this prespecified substudy, we hypothesized that the likelihood of a positive test result would be different in older vs younger patients and that the association between a positive test result and a clinical event would be influenced by test type as well as patient age.

Methods
Study Cohort and Design

The PROMISE trial was a pragmatic comparative effectiveness trial that enrolled 10 003 patients at 193 sites in North America. Full details regarding the PROMISE study design, rationale, and results have been previously published.3,4 Briefly, patients were randomized to an initial strategy of anatomic or functional testing. For patients in the functional testing arm, the choice of test modality (exercise ECG, stress echocardiography, or stress nuclear) was left to the discretion of the local clinician. For patients in the anatomic testing arm, the decision to perform coronary artery calcium (CAC) scoring as a part of the coronary CTA was left to discretion of the site. Randomization was stratified by site and prespecified stress test type. Following randomization, all tests were performed and interpreted by the site, and subsequent care was left to the discretion of the clinician. For the PROMISE trial, institutional review boards approved of the study at each site. Each patient was provided with oral and written informed and signed declaration of informed consent. The formal trial protocols can be found in Supplement 1.

Test positivity was defined as follows. For functional testing, an exercise ECG result was defined as positive if there were ST-segment changes of ischemia (ST-segment elevation or ≥1 -mm horizontal or downsloping ST-segment depression) or if the test was terminated early (<3 minutes) owing to reproduction of symptoms, arrhythmia, or hypotension. For functional evaluation, test results were considered positive if inducible ischemia was present in at least 1 coronary territory or similarly if the stress test required early termination (<3 minutes) of exercise. A high-risk test was defined as one with inducible ischemia involving a multivessel territory. A positive CTA result was defined as one with obstructive CAD. Test results were classified according to severity including normal, nonobstructive CAD (stenosis of 1%-69%, <50% left main) and obstructive CAD (stenosis ≥70% or ≥50% left main stenosis). A high-risk result was defined as a proximal left anterior descending stenosis of at least 70%, left main obstruction at least 50%, or multivessel disease. Among patients who also underwent CAC testing, patients were defined as having normal (Agatston score 0), mild (1-99), moderate (100-400), or severe (>400) CAC.

For this analysis, all patients were included if they underwent testing as randomized and had interpretable results. The 8966 patients meeting this criteria were similar in baseline characteristics to the excluded population (n = 1037, eTable 1 in Supplement 2). The primary end point of interest was a composite of CV death and myocardial infarction (MI). This end point was chosen, rather than the primary end point from the PROMISE trial (all cause death/MI/unstable angina hospitalization [UAH]) owing to the higher rates of noncardiovascular death among the older groups of patients. Use of CV death/MI provided a more specific end point for the examination of prognostic capabilities for future cardiovascular events of the 2 noninvasive tests based on age.

Statistical Analyses

Patients were divided into 3 groups: younger than 65 years (n = 6378), age 65 to 74 years (n = 2062), and 75 years and older (n = 526). These age cut points were chosen based on prior literature that identifies the group of patients 75 years and older as a significant population of interest and that also note the utility of harmonization of age subgroups to afford comparability.5-7 Baseline and presentation characteristics were described using median (25th, 75th percentiles) for continuous variables and percentages for categorical variables. Test results were compared based on age group and by type of test. Among patients who underwent CAC scoring, the proportion of patients with normal, mild, moderate, or severe CAC was compared between age groups. The association of continuous CAC score was also evaluated with age in a continuous manner.

We evaluated the effect of age and test type on test positivity. Using younger than 65 years as a reference, the odds of a positive test result were evaluated for ages 65 to 74 years and 75 years and older. We performed a multivariable logistic regression analysis, adjusting for clinically important variables, to determine whether the odds of a positive test result differed by test modality in patients of different age groups. The interaction test between type of noninvasive test and age was also performed.

We assessed the effect of age and randomized test type on the prognostic value of testing. Using age younger than 65 years as a reference, the risk of a clinical outcome (CV death/MI) was evaluated for age 65 to 74 years and 75 years and older. Multivariable Cox regression models were used to examine the association between test type, age group, and clinical outcomes of interest. The interaction between patient age and test type was performed. A multivariable Cox regression model was used to evaluate the prognostic value of a positive test result based on test modality and age group. The interaction between patient age, test result, and test type was performed. A similar analysis was completed using the primary end point from the PROMISE trial: all-cause death/MI/UAH. For both end points, a sensitivity analysis was performed looking at the change in ln (hazard ratio [HR]), which evaluated the prognostic value of a positive test result based on test modality and age as a continuous variable. Risk for each outcome was presented per 10-year increase in age.

Time to event (CV death/MI) analyses were completed based on test result for patients who underwent functional testing, coronary CTA, and CAC scoring in each age group. For CAC, a positive result was defined as CAC ≥100, whereas a score of 0 to 99 was considered negative. As a sensitivity analysis, time to event for each age group was also evaluated based on test type (intention to test) in the intention-to-treat population (all randomized patients) vs those included in this analysis (tested as randomized and with interpretable result).

For all analyses, a P value of less than .05 was considered significant, and all P values were 2-sided. Variables included for adjustment were consistent across analyses and included sex, race, body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), hypertension, diabetes, and site characterization of chest pain (typical, atypical, or noncardiac). Covariates for adjustment were limited given the small number of events in some age groups. The covariates used in this analysis were chosen based on observed differences across age categories and clinical judgement. All statistical calculations were carried out using SAS, version 9.4 (SAS Institute Inc).

Results
Baseline Characteristics

Among the 8966 included patients, 6378 (71.1%) were younger than 65 years, 2062 (23.0%) were between ages 65 and 74 years, and 526 (5.9%) were 75 years and older (eFigure 1 in Supplement 2). Computed tomography angiography was performed in 4500 patients and stress testing in 4466 patients. Among patients randomized to CTA, 4034 also underwent CAC testing (eFigure 2 in Supplement 2). There were significant differences in demographic characteristics and cardiac risk factors based on age. Older patients were more likely to be women, with lower BMIs, less diabetes, and less history of depression, and were more likely to be nonsmokers (Table 1). Older patients were also less likely to present with chest pain and more frequently reported dyspnea and were more likely to have ECG findings that could interfere with the interpretation of an exercise stress test. With increasing age, 10-year atherosclerotic CV disease risk increased; all patients 75 years and older had a score greater than 7.5% (mean score, 36.5%).

Prior to randomization, a stress nuclear study was chosen as the intended functional test by the clinician for most patients, increasing from 65.7% among patients younger than 65 years (4193 of 6378) to 73.8% of patients 75 years and older (388 of 526). Use of pharmacologic tests, as compared with exercise functional testing, also increased with age, from 24.9% of patients younger than 65 years (778 of 3129) to 50.9% of patients 75 years and older (133 of 283) (Table 2).

Age and Test Result

Positive noninvasive test results were more common with older age: 10.6% (n = 678) for younger than 65 years, 15.4% (n = 317) for ages 65 to 74 years, and 19.6% (n = 103) for 75 years and older (age 65-74 years: adjusted odds ratio [OR], 1.65; 95% CI, 1.42-1.91; age ≥75 years: adjusted OR, 2.32; 95% CI, 1.83-2.95). This was also true for those patients who underwent functional testing (11.2% [n = 350 of 3129] vs 15.6% [n = 164 of 1054] vs 17.7% [n = 50 of 283]; P < .001), and among those with a positive functional test, a high-risk result was more common with increasing age (60.6% [n = 212 of 350] vs 65.2% [107 of 164] vs 72.0% [n = 36 of 50]; P < .001). Similarly, older patients who underwent anatomic testing with coronary CTA were more likely to have a positive (obstructive) test result (10.1% [n = 328 of 3249] vs 15.2% [n = 1553 of 1008] vs 21.8% [n = 53 of 243]; P < .001). There was no significant interaction between noninvasive test type and positivity based on patient age (age <65 years: adjusted OR, 1.13; 95% CI, 0.96-1.33; age 65-74 years: adjusted OR, 1.04; 95% CI, 0.82-1.33; age ≥75 years: adjusted OR, 0.74; 95% CI, 0.48-1.15; P for interaction = .20).

Age and CAC

Among those who also had CAC scoring, one-third of patients 75 years and older had CAC of 400 (n = 73 of 219) as compared with only 10% of those patients younger than 65 years (n = 287 of 2900) (P < .001). When evaluated in a continuous manner, CAC was generally higher in men but increased with age for both sexes (eFigure 2 in Supplement 2).

Age, Test Type, and Outcomes

Overall, the rate of CV death/MI increased with age: 1.1% of patients younger than 65 years (n = 73), 2.0% of patients age 65 to 74 years (n = 42), and 2.9% of patients 75 years and older (n = 15). Compared with patients younger than 65 years, older patients were at higher risk for an event after adjustment (age 65-74 years: HR, 1.78; 95% CI, 1.21-2.62; age 75 years and older: HR, 2.21; 95% CI, 1.24-3.91). There was no interaction between the performed test and age group with the clinical outcome.

Among patients younger than 65 years who underwent functional testing, there was no difference in the association of a positive vs negative test result and CV death/MI after adjustment (HR, 1.09; 95% CI, 0.43-2.82). However, with increasing age group (Figure 1), a positive stress test result was significantly associated with the outcome (age 65 to 74 years: HR, 3.18; 95% CI, 1.44-7.01; age ≥75 years: HR, 6.55; 95% CI, 1.46-29.35).

Conversely, among patients who underwent anatomic evaluation with coronary CTA, a positive test result was associated with CV death/MI for patients younger than 65 years (HR, 3.04; 95% CI, 1.46-6.34) but was not predictive of future events in patients ages 65 to 74 years (HR, 0.67; 95% CI, 0.15-2.94) and at least 75 years (HR, 1.07; 95% CI, 0.22-5.34; Figure 1). There was a significant interaction between test results, test modality, and age group (Table 3) on the clinical outcome of CV death or MI, indicating that test result and outcomes were modified jointly by age and test type (Table 3). The sensitivity analysis of the CV death/MI outcome using age as a continuous variable showed similar results (eTable 2 in Supplement 2), with a significant difference in the slopes between CTA and stress testing. As seen in the categorical age models, CTA had a decreasing HR (HR, −0.71; 95% CI, −1.41 to 0.00; P = .05) while stress tests showed a significant increasing trend (HR, 0.67; 95% CI, 0.08-1.26; P = .03).

Among patients younger than 65 years who also underwent CAC evaluation, a high (≥100) CAC was associated with CV death/MI (HR, 2.73; 95% CI, 1.31-5.69). However, among older patients (age 65-74 years or ≥75 years), CAC of at least 100 did not provide additional risk discrimination when compared with CAC of 0 to 99 (age 65-74 years: HR, 0.44; 95% CI, 0.14-1.42; ≥75 years: HR, 1.31; 95% CI, 0.25-6.88). There was also a significant interaction between CAC score and age group on the clinical outcome of CV death or MI (P for interaction = .03).

When an analogous evaluation was performed for the composite outcome of all-cause death/MI/UAH, we found similar results among patients 75 years and older. A positive stress test was significantly associated with an event (HR, 4.08; 95% CI, 1.37-12.17) whereas neither a positive coronary CTA nor CAC score were associated with the outcome (HR, 1.30; 95% CI, 0.41-4.16; HR, 1.90; 95% CI, 0.51-7.07, respectively). However, among age groups younger than 65 years and 65 to 74 years, both stress testing (adjusted HR, 3.82; 95% CI, 2.32-6.30, P <.001 for <65 years; adjusted HR, 3.57; 95% CI, 1.96-6.50; P <.001 for age 65-74 years) and CTA (adjusted HR, 4.87; 95% CI, 3.07-7.72; P <.001 for <65 years; adjusted HR, 3.00; 95% CI, 1.58-5.69, P <.001 for age 65-74 years) were similarly predictive of future events. There was no significant interaction between test results, test modality, and age group (eTable 3 in Supplement 2). The change in ln(HR) per 10-year increase in age was significant for CTA (adjusted HR, −0.45; 95% CI −0.86 to −0.04, P = .03) but not for stress testing (adjusted HR, −0.11; 95% C,I −0.51 to 0.30; P = .61). There was no significant interaction between test results, test modality, and age group for the outcome of all-cause death/MI/UAH using age as a continuous variable (eTable 2 in the Supplement).

Time to First Event Modified by Age, Noninvasive Test, and Test Positivity

Among patients younger than 65 years, those with a positive coronary CTA had an early increase in the probability of having an event compared with those with a negative coronary CTA (HR, 3.04; 95% CI, 1.46-6.34). Similarly, for patients who also had CAC scoring completed, those younger than 65 years with a positive CAC score (≥100) had an increase in events over time compared with those patients with CAC less than 100. However, patients younger than 65 years who underwent stress testing had a near identical risk of event throughout the follow-up regardless of a positive vs negative stress test result (HR, 1.09; 95% CI, 0.43-2.82) (Figure 2).

Conversely, in both the 65 to 74 years and 75 years and older age groups, patients with a positive stress test result had higher risk of having an event as compared with a negative stress test result (HR, 3.18; 95% CI, 1.44-7.01 and HR, 6.55; 95% CI, 1.46-29.35; respectively). However, in both the 65 to 74 years and 75 years and older age groups, those with a positive coronary CTA result or positive CAC score (≥100) had no higher probability of having CV death/MI throughout the study period as compared with those with a negative result (CTA: HR, 0.67; 95% CI, 0.15-2.94 and HR, 1.07; 95% CI, 0.22-5.34; respectively; CAC: HR, 0.44; 95% CI, 0.15-2.94 and HR, 1.31; 95% CI, 0.22-5.34; respectively).

Time to first event was similar when we evaluated functional testing vs CTA among all randomized patients from PROMISE (N = 10 003) in an intention-to-test sensitivity analysis (<65 years: adjusted HR, 1.23; 95% CI, 0.81-1.86; age 65-74 years: adjusted HR, 1.77; 95% CI, 0.99-3.17; ≥75 years: adjusted HR, 0.74; 95% CI, 0.29-1.88; eFigure 3 in Supplement 2).

Discussion

The noninvasive evaluation of stable symptoms in patients without a history of CAD seeks to confirm the diagnosis and provide prognostic information about future risk. In this study of the PROMISE trial, age was associated with an increased prevalence of CAD, as evidenced by a higher proportion of positive functional test results, obstructive disease on coronary CTA, and higher CAC score, with no interaction between test type and diagnostic yield. However, we found a significant interaction between age group and noninvasive test type for the prediction of future CV death or MI events. Specifically, coronary CTA and CAC score better stratified prognosis for patients younger than 65 years, while functional testing was a better determinant of future risk for older adults (≥65 years).

The higher burden of disease among older patients in our study aligns with prior work that has shown that age is independently associated with CAC.8,9 Notably, we observed that the increase in underlying obstructive disease and CAC in older patients was associated with a more limited discriminating ability of noninvasive tests to predict future CV events. Because prevalence is high, these findings should potentially be viewed more as a marker of age rather than an indication of risk for future events. Conversely, we found that functional testing provides effective risk stratification in patients 65 years and older. This aligns with prior literature that highlights the prognostic information gained from functional assessment for coronary artery disease.10,11 However, among patients younger than 65 years old, we observed that while functional testing did not effectively discriminate risk for future events, the presence of CAC is more informative for identifying risk for adverse outcomes.

Despite the well-established higher prevalence of CAD in older patients, there remains a paucity of scientific data to help guide clinical decisions.12 In the setting of evaluation of CV risk, there are unique patient and testing considerations that may limit the generalization of trial results. For example, older patients with possible CAD are more likely to present with atypical symptoms, as exemplified by the nearly one-third of patients 75 years and older in our study who presented with dyspnea. Because older patients frequently have multiple comorbidities, it can be particularly challenging to parse out the etiology of such symptoms.13,14 Each noninvasive test also has unique considerations. Exercise stress tests may be constrained by frailty, exercise limitations, and baseline ECG abnormalities.15 We observed an increase in the use of pharmacologic stress testing with age, and more than 10% of patients 75 years and older had baseline ECG abnormalities that could interfere with the interpretation. Alternatively, coronary CTA may be limited by a higher burden of atrial fibrillation and dense coronary calcification, as we observed in our study.15,16

These potential limitations are important to consider to identify the test most likely to provide the desired information. However, among the patients in our study, we found that older patients are more likely to have a positive noninvasive test result, regardless of type of test, and that most have elevated CAC scores. After a positive noninvasive test result, the next step in evaluation frequently is invasive coronary angiography. To improve the selection of patients for an invasive evaluation of CAD in cohorts, such as ours, with low overall event rates, it is essential to understand the prognostic value associated with a positive test result.

In light of these considerations, the differences we observed in the prognostic information provided by a positive test result have the potential for broad applicability. For example, the 2018 ACC/AHA cholesterol guideline recommends the use of CAC score for further risk stratification in patients deemed to be of intermediate risk.17 Our study suggests that the age of the patient may also be important when evaluating the prognostic information gained from CAC score. Similarly, the different prognostic values of stress testing and coronary CTA that we observed among the different age groups suggest that the age of the patient may be an important consideration when choosing a test modality. In these ways, both the testing strategy and test interpretation can be tailored to the unique patient characteristics and used to guide patient-centered discussions surrounding treatment based on the individual risk for future CV events.

Limitations

Several limitations associated with our study should also be acknowledged. First, although this is a prespecified substudy of the PROMISE trial, it should be considered as hypothesis generating rather than offering definitive information about optimal testing strategies. Second, randomization for the PROMISE trial was not stratified based on age. However, similar numbers of patients were randomized to each test modality in each age group. While stratification into 3 age groups for our analysis allowed us to compare differences between the groups, there is inherent variability based on age within each strata that may not be fully assessed. Additionally, the cohort of patients 75 years and older was the smallest group enrolled in the trial. Unfortunately, this group of clinically important patients remains largely underrepresented in the literature; the small number enrolled in our cohort suggests that this may in part be owing to the age distribution of stable outpatients with symptoms concerning for CAD. Next, given that less than 10% of patients enrolled in the PROMISE study underwent coronary angiography, we were unable to assess the diagnostic accuracy of testing based on age. Further, the median duration of follow-up (25 months) with a relatively small number of events may limit the ability to assess long-term prognosis. Additionally, we did not assess the value of zero vs nonzero CAC score because a finding of zero CAC was too infrequent in the 75 years and older age group to perform the analysis.18 Given the nature of the PROMISE study, CAC score was calculated in patients both with and without active statin therapy, which may have affected these results.17 While we performed analyses with adjustment for clinically important variables, given the small number of clinical events, we could not adjust for many captured and uncaptured variables that could contribute to future CV risk. Finally, we did not adjust for patient atherosclerotic CV disease risk given that this calculation includes age and thus would confound the associations between age and clinical outcomes.

Conclusions

Among patients with stable symptoms undergoing evaluation for CAD, important differences exist in the patient presentation, test results, and prognostic capabilities of noninvasive testing based on the patient’s age. Coronary CTA provides additional risk stratification information for patients younger than 65 years whereas stress testing results are associated with risk in patients 65 years and older. These results suggest that age-specific approaches to the noninvasive evaluation of CAD should be further explored.

Back to top
Article Information

Corresponding Author: Angela Lowenstern, MD, Duke Clinical Research Institute, Duke University, 2301 Erwin Road, DUMC 3845, Durham, NC 27710 (angela.sandelin@duke.edu).

Accepted for Publication: October 29, 2019.

Published Online: November 18, 2019. doi:10.1001/jamacardio.2019.4973

Author Contributions: Drs Lowenstern and Douglas had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Lowenstern, Hill, Pellikka, Nanna, Mehta, Hoffmann, Douglas.

Acquisition, analysis, or interpretation of data: Lowenstern, Alexander, Hill, Alhanti, Nanna, Cooper, Bullock-Palmer, Hoffmann, Douglas.

Drafting of the manuscript: Lowenstern, Alexander, Hill, Nanna.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Lowenstern, Hill, Alhanti.

Obtained funding: Hoffmann, Douglas.

Administrative, technical, or material support: Cooper, Hoffmann, Douglas.

Supervision: Alexander, Alhanti, Pellikka, Nanna, Cooper, Hoffmann, Douglas.

Conflict of Interest Disclosures: Dr Lowenstern reported grants from the National Heart, Lung, and Blood Institute (NHLBI) and the National Institutes of Health (NIH) during the conduct of the study. Dr Nanna reported grants from the NIH during the conduct of the study. Dr Hoffmann reported grants from KOWA, MedImmune, HeartFlow, Duke University (Abbott), Oregon Health & Science University (AHA, 13FTF16450001), Columbia University (NIH, 5R01-HL109711), and the NIH/NHLBI and personal fees from Abbott, Duke University (NIH), and Recor Medical outside the submitted work. Dr Douglas reported grants from HeartFlow outside the submitted work and other support from GE HealthCare. No other disclosures were reported.

Funding/Support: This project was supported by grants R01HL098237, R01HL098236, R01HL98305, and R01HL098235 from the National Heart, Lung, and Blood Institute.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the paper and its final contents. This article does not necessarily represent the official views of the National Heart, Lung, and Blood Institute.

Meeting Presentation: This paper was presented at the American Heart Association’s Scientific Sessions 2019; November 18, 2019; Philadelphia, Pennsylvania.

References
1.
McClelland  RL, Chung  H, Detrano  R, Post  W, Kronmal  RA.  Distribution of coronary artery calcium by race, gender, and age: results from the Multi-Ethnic Study of Atherosclerosis (MESA).  Circulation. 2006;113(1):30-37. doi:10.1161/CIRCULATIONAHA.105.580696PubMedGoogle ScholarCrossref
2.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
3.
Douglas  PS, Hoffmann  U, Patel  MR,  et al; PROMISE Investigators.  Outcomes of anatomical versus functional testing for coronary artery disease.  N Engl J Med. 2015;372(14):1291-1300. doi:10.1056/NEJMoa1415516PubMedGoogle ScholarCrossref
4.
Douglas  PS, Hoffmann  U, Lee  KL,  et al; PROMISE investigators.  Prospective Multicenter Imaging Study for Evaluation of chest pain: rationale and design of the PROMISE trial.  Am Heart J. 2014;167(6):796-803.e1. doi:10.1016/j.ahj.2014.03.003PubMedGoogle ScholarCrossref
5.
Rich  MW, Chyun  DA, Skolnick  AH,  et al; American Heart Association Older Populations Committee of the Council on Clinical Cardiology, Council on Cardiovascular and Stroke Nursing, Council on Cardiovascular Surgery and Anesthesia, and Stroke Council; American College of Cardiology; and American Geriatrics Society.  Knowledge gaps in cardiovascular care of the older adult population: a scientific statement from the American Heart Association, American College of Cardiology, and American Geriatrics Society.  Circulation. 2016;133(21):2103-2122. doi:10.1161/CIR.0000000000000380PubMedGoogle ScholarCrossref
6.
Alexander  KP, Newby  LK, Cannon  CP,  et al; American Heart Association Council on Clinical Cardiology; Society of Geriatric Cardiology.  Acute coronary care in the elderly, part I: non-ST-segment-elevation acute coronary syndromes: a scientific statement for healthcare professionals from the American Heart Association Council on Clinical Cardiology: in collaboration with the Society of Geriatric Cardiology.  Circulation. 2007;115(19):2549-2569. doi:10.1161/CIRCULATIONAHA.107.182615PubMedGoogle ScholarCrossref
7.
Alexander  KP, Newby  LK, Armstrong  PW,  et al; American Heart Association Council on Clinical Cardiology; Society of Geriatric Cardiology.  Acute coronary care in the elderly, part II: ST-segment-elevation myocardial infarction: a scientific statement for healthcare professionals from the American Heart Association Council on Clinical Cardiology: in collaboration with the Society of Geriatric Cardiology.  Circulation. 2007;115(19):2570-2589. doi:10.1161/CIRCULATIONAHA.107.182616PubMedGoogle ScholarCrossref
8.
Newman  AB, Naydeck  BL, Sutton-Tyrrell  K, Feldman  A, Edmundowicz  D, Kuller  LH.  Coronary artery calcification in older adults to age 99: prevalence and risk factors.  Circulation. 2001;104(22):2679-2684. doi:10.1161/hc4601.099464PubMedGoogle ScholarCrossref
9.
Giallauria  F, Vigorito  C, Ferrara  N, Ferrucci  L.  Cardiovascular calcifications in old age: mechanisms and clinical implications.  Curr Transl Geriatr Exp Gerontol Rep. 2013;2(4):255-267. doi:10.1007/s13670-013-0063-4Google ScholarCrossref
10.
Rai  M, Baker  WL, Parker  MW, Heller  GV.  Meta-analysis of optimal risk stratification in patients >65 years of age.  Am J Cardiol. 2012;110(8):1092-1099. doi:10.1016/j.amjcard.2012.05.048PubMedGoogle ScholarCrossref
11.
Marwick  TH, Mehta  R, Arheart  K, Lauer  MS.  Use of exercise echocardiography for prognostic evaluation of patients with known or suspected coronary artery disease.  J Am Coll Cardiol. 1997;30(1):83-90. doi:10.1016/S0735-1097(97)00148-4PubMedGoogle ScholarCrossref
12.
Herrera  AP, Snipes  SA, King  DW, Torres-Vigil  I, Goldberg  DS, Weinberg  AD.  Disparate inclusion of older adults in clinical trials: priorities and opportunities for policy and practice change.  Am J Public Health. 2010;100(suppl 1):S105-S112. doi:10.2105/AJPH.2009.162982PubMedGoogle ScholarCrossref
13.
Dai  X, Busby-Whitehead  J, Alexander  KP.  Acute coronary syndrome in the older adults.  J Geriatr Cardiol. 2016;13(2):101-108.PubMedGoogle Scholar
14.
Fihn  SD, Gardin  JM, Abrams  J,  et al; American College of Cardiology Foundation; American Heart Association Task Force on Practice Guidelines; American College of Physicians; American Association for Thoracic Surgery; Preventive Cardiovascular Nurses Association; Society for Cardiovascular Angiography and Interventions; Society of Thoracic Surgeons.  2012 ACCF/AHA/ACP/AATS/PCNA/SCAI/STS Guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, and the American College of Physicians, American Association for Thoracic Surgery, Preventive Cardiovascular Nurses Association, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons.  J Am Coll Cardiol. 2012;60(24):e44-e164. doi:10.1016/j.jacc.2012.07.013PubMedGoogle ScholarCrossref
15.
Madhavan  MV, Gersh  BJ, Alexander  KP, Granger  CB, Stone  GW.  Coronary artery disease in patients ≥80 years of age.  J Am Coll Cardiol. 2018;71(18):2015-2040. doi:10.1016/j.jacc.2017.12.068PubMedGoogle ScholarCrossref
16.
Kuller  LH, Lopez  OL, Mackey  RH,  et al.  Subclinical Cardiovascular Disease and Death, Dementia, and Coronary Heart Disease in Patients 80+ Years.  J Am Coll Cardiol. 2016;67(9):1013-1022. doi:10.1016/j.jacc.2015.12.034PubMedGoogle ScholarCrossref
17.
Grundy  SM, Stone  NJ, Bailey  AL,  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(25):e1082-e1143.PubMedGoogle Scholar
18.
Budoff  MJ, Mayrhofer  T, Ferencik  M,  et al; PROMISE Investigators.  Prognostic value of coronary artery calcium in the PROMISE Study (Prospective Multicenter Imaging Study for Evaluation of Chest Pain).  Circulation. 2017;136(21):1993-2005. doi:10.1161/CIRCULATIONAHA.117.030578PubMedGoogle ScholarCrossref
×