The

n e w e ng l a n d j o u r na l

of

m e dic i n e

Original Article

Troponin and Cardiac Events in Stable

Ischemic Heart Disease and Diabetes
BrendanM. Everett, M.D., M.P.H., MariaMori Brooks, Ph.D.,
HelenE.A. Vlachos, M.S., BernardR. Chaitman, M.D., RobertL. Frye, M.D.,
and DeepakL. Bhatt, M.D., M.P.H., for the BARI 2D Study Group*

A BS T R AC T
BACKGROUND
From the Divisions of Cardiovascular
Medicine (B.M.E., D.L.B.) and Preventive
Medicine (B.M.E.), Brigham and Womens
Hospital and Harvard Medical School,
Boston; the Department of Epidemiology
and Graduate School of Public Health, University of Pittsburgh, Pittsburgh (M.M.B.,
H.E.A.V.); Center for Cardiovascular Care,
Saint Louis University School of Medicine, St. Louis (B.R.C.); and the Department of Medicine, Division of Cardiovascular Diseases, Mayo Clinic, Rochester,
MN (R.L.F.). Address reprint requests to
Dr. Everett at Brigham and Womens Hospital, 75 Francis St., Boston, MA 02115,
or at beverett@partners.org.
* A complete list of investigators in the
Bypass Angioplasty Revascularization
Investigation in Type 2 Diabetes (BARI
2D) Study Group is provided in the
Supplementary Appendix, available at
NEJM.org.
N Engl J Med 2015;373:610-20.
DOI: 10.1056/NEJMoa1415921
Copyright 2015 Massachusetts Medical Society.

Cardiac troponin concentrations are used to identify patients who would benefit
from urgent revascularization for acute coronary syndromes. We hypothesized that
they might be used in patients with stable ischemic heart disease to identify those
at high risk for cardiovascular events who might also benefit from prompt coronary revascularization.
METHODS

We measured the cardiac troponin T concentration at baseline with a high-sensitivity assay in 2285 patients who had both type 2 diabetes and stable ischemic
heart disease and were enrolled in the Bypass Angioplasty Revascularization Investigation in Type 2 Diabetes trial. We tested for an association between the
troponin T concentration and a composite end point of death from cardiovascular
causes, myocardial infarction, or stroke; we then evaluated whether random assignment to prompt revascularization reduced the rate of the composite end point
in patients with an abnormal troponin T concentration (14 ng per liter) as compared with those with a normal troponin T concentration (<14 ng per liter).
RESULTS

Of the 2285 patients, 2277 (99.6%) had detectable (3 ng per liter) troponin T concentrations and 897 (39.3%) had abnormal troponin T concentrations at baseline.
The 5-year rate of the composite end point was 27.1% among the patients who had
had abnormal troponin T concentrations at baseline, as compared with 12.9%
among those who had had normal baseline troponin T concentrations. In models
that were adjusted for cardiovascular risk factors, severity of diabetes, electrocardiographic abnormalities, and coronary anatomy, the hazard ratio for the composite end point among patients with abnormal troponin T concentrations was 1.85
(95% confidence interval [CI], 1.48 to 2.32; P<0.001). Among patients with abnormal troponin T concentrations, random assignment to prompt revascularization,
as compared with medical therapy alone, did not result in a significant reduction
in the rate of the composite end point (hazard ratio, 0.96; 95% CI, 0.74 to 1.25).
CONCLUSIONS

The cardiac troponin T concentration was an independent predictor of death from

cardiovascular causes, myocardial infarction, or stroke in patients who had both
type 2 diabetes and stable ischemic heart disease. An abnormal troponin T value
of 14 ng per liter or higher did not identify a subgroup of patients who benefited
from random assignment to prompt coronary revascularization. (Funded by the
National Institutes of Health and Roche Diagnostics; BARI 2D ClinicalTrials.gov
number, NCT00006305.)
610

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

Troponin in Stable Heart Disease and Diabetes

ardiac troponin concentration is

the preferred marker of myocardial necrosis.1 Elevated concentrations of cardiac troponin have a strong association with an
adverse prognosis in patients with acute coronary syndromes and are used to identify patients who are likely to benefit from an early
invasive management strategy.2-4 High-sensitivity assays that allow the measurement of very
low cardiac troponin levels in patients with
stable heart disease are now available for clinical and research use. These low, previously undetectable troponin concentrations have shown
strong associations with myocardial infarction,
stroke, and death in a variety of primary and
secondary prevention populations, including in
patients with stable ischemic heart disease.5-10
We hypothesized that a high-sensitivity cardiac
troponin assay might be used to identify a subgroup of patients with both stable ischemic
heart disease and diabetes who are at high risk
for cardiac events and, furthermore, might predict who would benefit from prompt coronary
revascularization. We also hypothesized that
coronary revascularization would lower subsequent measurements of circulating troponin
concentrations.
To test these hypotheses, we used a high-sensitivity assay to measure cardiac troponin T concentrations at baseline and at the 1-year followup in 2368 patients with both type 2 diabetes
and stable ischemic heart disease who underwent
randomization to either prompt coronary revascularization plus intensive medical therapy or
intensive medical therapy alone in the Bypass
Angioplasty Revascularization Investigation in
Type 2 Diabetes (BARI 2D) trial.

rent study is an ancillary study of the BARI 2D

trial. All enrolled BARI 2D participants provided
written informed consent.
Study Oversight

This study was approved by the Partners Human

Research Committee at Brigham and Womens
Hospital, Boston. The study was designed by the
first, second, and last authors. All the authors
collected the data. The initial independent data
analysis was conducted by the second and third
author. Subsequent interpretation and analyses
were conducted by all authors. All authors take
responsibility for the accuracy and completeness
of the analyses. The first author wrote the first
draft of the manuscript; no one who is not an
author contributed to the writing of the manuscript. All the authors made the decision to submit the manuscript for publication. The highsensitivity electrochemiluminescence assays used
to measure cardiac troponin T concentration
were donated by Roche Diagnostics, which had
no other role in the study.
Study End Points

The primary end point of this ancillary study of

the BARI 2D trial was a composite of death from
cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke. A core laboratory
adjudicated myocardial infarction, and an independent mortality and morbidity committee
classified the cause of death. The criteria that
were used to adjudicate myocardial infarction
and stroke are described elsewhere.11 Secondary
end points of this study were the rates of death
from any cause, the individual components of
the primary composite outcome, heart failure,
and the composite of death from any cause,
myocardial infarction, stroke, or heart failure.
Me thods
Details of the ascertainment of heart failure are
Study Population and Design
provided in the Methods section in the SuppleThe inclusion and exclusion criteria for the BARI mentary Appendix, available with the full text of
2D trial are described in detail elsewhere.11 In this article at NEJM.org.
brief, starting on January 1, 2001, a total of 2368
patients who had both type 2 diabetes and stable Laboratory Analysis
ischemic heart disease with mild or no symp- The plasma samples that were obtained at basetoms of angina were selected as candidates for line and at the 1-year follow-up were stored at
either percutaneous coronary intervention or 80C in the BARI 2D Fibrinolysis Core Laboracoronary-artery bypass grafting and were then tory at the University of Vermont. Cardiac troporandomly assigned to either prompt coronary nin T concentration was measured with the use
revascularization plus intensive medical therapy of high-sensitivity electrochemiluminescence as(revascularization group) or intensive medical says. According to the manufacturer, the limit of
therapy alone (medical-therapy group). The cur- detection of the troponin T assay is 3 ng per liter,
n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

611

The

n e w e ng l a n d j o u r na l

and the 99th percentile (upper reference limit) in

a healthy population of volunteers is 14 ng per
liter. The 10% coefficient of variation is reported
to be less than this value. The methods for measuring N-terminal probrain natriuretic peptide
(NT-proBNP) levels are described in the Methods
section in the Supplementary Appendix.
Statistical Analysis

High-sensitivity cardiac troponin T concentrations

were categorized as normal (<14 ng per liter) or
abnormal (14 ng per liter) according to the
manufacturers established upper reference limit.
We compared the baseline characteristics of the
group with normal troponin T concentrations
with those of the group with abnormal troponin
T concentrations using the Wilcoxon rank-sum
test for continuous variables and the chi-square
test for categorical variables. KaplanMeier analysis and log-rank statistics were used for betweengroup comparisons of the unadjusted rates of
each of the study end points. We constructed
Cox proportional-hazards models to estimate the
adjusted association between troponin T concentration and the risk of the primary composite
end point of death from cardiovascular causes,
nonfatal myocardial infarction, or nonfatal stroke,
as well as the risks of the secondary end points
of death from any cause, heart failure, and the
composite of death from any cause, myocardial
infarction, stroke, or heart failure. A two-sided
P value of less than 0.05 was considered to indicate statistical significance. The approach that
was taken to adjust for possible confounding
variables is described in the Methods section in
the Supplementary Appendix. We tested for evidence of heterogeneity of the effect of random
assignment to prompt revascularization among
patients with normal baseline troponin T concentrations and those with abnormal baseline troponin T concentrations in a model that included
a multiplicative interaction term along with the
main-effect terms for revascularization and biomarker category. Sensitivity analyses and the
methods for assessing change in troponin T concentrations, along with a landmark analysis involving the 1984 participants with baseline and
1-year follow-up troponin T values, are described
in the Methods section in the Supplementary
Appendix.

612

of

m e dic i n e

R e sult s
Patients

Of the 2368 patients enrolled in the BARI 2D

trial, 2314 (97.7%) had baseline plasma samples
available, and samples from 2285 of those patients (98.7%) were successfully analyzed for
troponin T concentration. Of these 2285 patients,
1984 (86.8%) also had 1-year follow-up samples
that were available and successfully analyzed
(Fig. S1 in the Supplementary Appendix). The
median follow-up time for the 2285 patients
with baseline plasma samples that could be analyzed for troponin T concentration was 5 years
(interquartile range, 4.1 to 6.0). Of these patients,
2277 (99.6%) had troponin T concentrations
above the limit of detection (3 ng per liter) and
897 (39.3%) had concentrations in the abnormal
range (14 ng per liter). Median troponin T concentrations were similar among patients in the
revascularization group (11.7 ng per liter; interquartile range, 8.0 to 19.9) and those in the
medical-therapy group (11.6 ng per liter; interquartile range, 7.8 to 19.3) (P=0.41). The baseline characteristics of the study population are
shown in Table1.
Baseline Troponin T Concentration
and Cardiovascular Events

In unadjusted KaplanMeier analyses, we observed

a significantly higher 5-year rate of the composite
end point of death from cardiovascular causes,
myocardial infarction, or stroke (Fig.1A) among
patients who had had abnormal baseline troponin T concentrations than among those who had
had normal baseline concentrations. The 5-year
incidence of the primary composite end point
was 27.1% among patients with abnormal baseline troponin T concentrations, versus 12.9%
among those with normal baseline troponin T
concentrations (P<0.001). There were similar
large between-group differences in the rates of
the individual components of the primary end
point (i.e., death from cardiovascular causes,
myocardial infarction, and stroke; Fig. S2, S3,
and S4, respectively, in the Supplementary Appendix), as well as in the rate of death from any
cause (Fig.1B), the rate of heart failure (Fig. S5
in the Supplementary Appendix), and the rate
of the secondary composite end point of death

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

Troponin in Stable Heart Disease and Diabetes

from any cause, myocardial infarction, stroke, or

heart failure (Fig. S6 in the Supplementary Appendix).
After adjustment for traditional cardiovascular risk factors, history of myocardial infarction,
history of heart failure, factors related to the
severity of type 2 diabetes, glomerular filtration
rate, electrocardiographic abnormalities, myocardial jeopardy index (the proportion [0 to 100%]
of the myocardium supplied by diseased coronary arteries),12 the number of coronary lesions,
and abnormal ejection fraction, the adjusted
hazard ratio for the composite end point among
patients with abnormal baseline troponin T concentrations remained robust and significant
(1.85; 95% confidence interval [CI], 1.48 to 2.32;
P<0.001) (Table2). The association of abnormal
troponin T concentrations with death from cardiovascular causes and myocardial infarction, as
well as with death from any cause, heart failure, and the composite of death from any
cause, myocardial infarction, stroke, or heart
failure remained significant after adjustment
(Table S1); additional adjustment for log-transformed NT-proBNP concentrations led to attenuation of the point estimates. However, in the
case of each of these end points, the risk estimates remained significant (Table S1, Model 4, in
the Supplementary Appendix).
In secondary analyses that used baseline
quintiles of troponin T concentration, the unadjusted rate of the primary end point at 5 years
differed significantly across the five quintile
groups; event rates were similar in the three lowest strata of troponin T concentration (<7.3 ng per
liter, 7.3 to <10.0 ng per liter, and 10.0 to
<14.0 ng per liter) and were substantially higher
in the fourth quintile (14.0 to <23.0 ng per liter)
and the fifth quintile (23.0 ng per liter) (Fig. S7
in the Supplementary Appendix). Adjusted models confirmed the observation that cardiovascular risk was elevated only in patients with abnormal baseline troponin T concentrations (Table
S2 in the Supplementary Appendix). In exploratory analyses, we observed that patients with
abnormal baseline troponin T concentrations
and NT-proBNP concentrations of 300 ng per
liter or higher were at increased risk for adverse
outcomes (Table S3 in the Supplementary Appendix).

Revascularization and Cardiovascular Events

Among the patients included in this BARI 2D

ancillary study, prompt revascularization was
not associated with a significantly lower risk of
the primary composite end point than that with
medical therapy alone (hazard ratio, 0.98; 95%
CI, 0.80 to 1.19; P=0.83), a finding that is similar to the overall results of the BARI 2D trial11;
this lack of significant benefit was seen both
among patients with normal troponin T concentrations and among those with abnormal troponin T concentrations (Fig.2). The effect of revascularization according to the two strata of
troponin T concentration appeared to be consistent for each of the components of the primary
composite end point, with no significant evidence of heterogeneity. Similar results were observed with respect to death from any cause,
heart failure, and the composite end point of
death from any cause, myocardial infarction,
stroke, or heart failure (Fig.2). In a series of
sensitivity analyses, we again observed no benefit of revascularization across increasing quintiles of troponin T concentration (Table S4),
when follow-up was truncated at 1 year (Table
S5) or when patients with high concentrations of
both troponin T and NT-proBNP were evaluated
(Table S6 in the Supplementary Appendix). Finally, no evidence of heterogeneity was observed
in analyses in which we used natural log-transformed troponin T concentration instead of
categorical troponin T concentration (data not
shown). We repeated these analyses separately in
strata of patients who were eligible for revascularization with coronary-artery bypass grafting
and those who were eligible for revascularization with percutaneous coronary intervention,
and there was no differential benefit of revascularization among those with abnormal troponin
T concentrations (Fig. S8 and S9 in the Supplementary Appendix).
Revascularization and Follow-up Troponin T
Concentrations

The 1984 patients with both a baseline and

1-year follow-up troponin T concentration available tended to be healthier than those for whom
only one value was available (Table S7 in the
Supplementary Appendix). There was a numerically small but statistically significant increase

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

613

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

Table 1. Baseline Characteristics of the Study Population, According to High-Sensitivity Cardiac Troponin T Concentration.*
Troponin T <14 ng/liter
(N=1388)

Troponin T 14 ng/liter
(N=897)

P Value

61 (5568)

64 (5871)

<0.001

903/1388 (65.1)

698/897 (77.8)

<0.001

White non-Hispanic

926/1388 (66.7)

571/897 (63.7)

Black non-Hispanic

218/1388 (15.7)

170/897 (19.0)

Hispanic

164/1388 (11.8)

127/897 (14.2)

Asian non-Hispanic

71/1388 (5.1)

24/897 (2.7)

Other non-Hispanic

9/1388 (0.6)

5/897 (0.6)

Characteristic
Median age (IQR) yr
Male sex no./total no. (%)
Race or ethnic group no./total no. (%)

Median body-mass index (IQR)

0.006

31.0 (27.734.7)

30.9 (27.335.0)

0.60

Systolic

129 (118141)

130 (119145)

0.07

Diastolic

74 (6881)

73 (6781)

0.18

Hypertension requiring treatment no./total no. (%)

1107/1371 (80.7)

757/888 (85.2)

0.006

Hypercholesterolemia requiring treatment

no./total no. (%)

1126/1367 (82.4)

721/889 (81.1)

0.45

History of cigarette smoking no./total no. (%)

919/1388 (66.2)

608/897 (67.8)

0.44

History of myocardial infarction no./total no. (%)

389/1369 (28.4)

334/881 (37.9)

<0.001

55/1377 (4.0)

93/893 (10.4)

<0.001

History of stroke or TIA no./total no. (%)

127/1388 (9.1)

94/897 (10.5)

0.29

Median blood pressure (IQR) mm Hg

History of congestive heart failure requiring treatment

no./total no. (%)
Median duration of diabetes mellitus (IQR) yr

7.2 (3.313.8)

10.4 (4.917.9)

<0.001

Chronic renal dysfunction no./total no. (%)

13/1385 (0.9)

57/887 (6.4)

<0.001

Median glycated hemoglobin (IQR) %

7.3 (6.48.4)

7.4 (6.58.7)

0.03

Total cholesterol

165 (142190)

163 (139194)

0.68

HDL cholesterol

37 (3243)

36 (3142)

0.003
0.30

Median lipids (IQR) mg/dl

LDL cholesterol

93 (73114)

90 (72115)

146.0 (100.0215.0)

151.5 (109.0223.0)

0.02

80.3 (66.794.9)

69.6 (55.183.3)

<0.001

93.0 (46.4185.8)

280.2 (104.2719.8)

<0.001

Insulin

322/1388 (23.2)

314/897 (35.0)

<0.001

Aspirin

1234/1381 (89.4)

771/895 (86.1)

0.02

Statin

1036/1385 (74.8)

666/895 (74.4)

0.84

41.0 (24.060.0)

46.0 (28.065.0)

0.001

Triglycerides
Median estimated GFR (IQR) ml/min/1.73 m2
Median NT-proBNP (IQR) ng/liter
Current medications no./total no. (%)

Median myocardial jeopardy index score (IQR)

Number of lesions with 50% stenosis
no./total no. (%)
01

614

<0.001
470/1388 (33.9)

216/897 (24.1)

23

581/1388 (41.9)

366/897 (40.8)

>3

337/1388 (24.3)

315/897 (35.1)

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

Troponin in Stable Heart Disease and Diabetes

Table 1. (Continued.)
Troponin T <14 ng/liter
(N=1388)

Characteristic

Troponin T 14 ng/liter
(N=897)

Number of vessels with lesions with 50% stenosis

no./total no. (%)
1

P Value
<0.001

574/1386 (41.4)

284/896 (31.7)

491/1386 (35.4)

326/896 (36.4)

321/1386 (23.2)

286/896 (31.9)

Proximal LAD stenosis of 50% no./total no. (%)

174/1388 (12.5)

130/897 (14.5)

0.18

LVEF <50% no./total no. (%)

163/1388 (11.7)

219/897 (24.4)

<0.001

Any ECG abnormality no./total no. (%)

656/1349 (48.6)

575/869 (66.2)

<0.001

CABG revascularization stratum no./total no. (%)

429/1388 (30.9)

322/897 (35.9)

0.01

Assignment to prompt revascularization

no./total no. (%)

675/1388 (48.6)

453/897 (50.5)

0.38

* To convert the values for cholesterol to millimoles per liter, multiply by 0.02586. To convert the values for triglycerides
to millimoles per liter, multiply by 0.01129. CABG denotes coronary-artery bypass grafting, ECG electrocardiogram, GFR
glomerular filtration rate, HDL high-density lipoprotein, IQR interquartile range, LAD left anterior descending coronary
artery, LDL low-density lipoprotein, LVEF left ventricular ejection fraction, NT-proBNP N-terminal probrain natriuretic
peptide, and TIA transient ischemic attack.
Race or ethnic group was self-reported.
The body-mass index is the weight in kilograms divided by the square of the height in meters.
The myocardial jeopardy index is the proportion (0 to 100%) of the myocardium supplied by significantly diseased
coronary arteries. Thus, a higher score represents more severe coronary artery disease and is associated with worse
prognosis.
In the Bypass Angioplasty Revascularization Investigation in Type 2 Diabetes trial, before randomization, patients were
stratified according to the most appropriate method of coronary revascularization (percutaneous coronary intervention
vs. CABG).

in troponin T concentration between baseline

and 1 year among all the patients in the BARI
2D trial who had measures available at both
time points (median, 3.4%; interquartile range,
4.8 to 11.1; P<0.001). This increase was similar
among patients in the revascularization group
(median, 3.7%; interquartile range, 5.1 to 11.5)
and those in the medical-therapy group (median,
3.1%; interquartile range, 4.7 to 10.6) (P=0.40)
(Table S8 in the Supplementary Appendix). Most
patients (86.4%) had troponin T concentrations
at the 1-year follow-up that changed (increased
or decreased) by 25% or less from the baseline
value. The proportion of patients who had an
increase of more than 25% in troponin T concentration was 6.7%, and the proportion who
had a decrease of more than 25% in troponin T
concentration was 6.9%. These proportions were
similar in the revascularization and medicaltherapy groups (Table S8 in the Supplementary
Appendix).

Changes in Troponin T Concentration and

Cardiovascular Risk

The baseline characteristics of the patients who

had a decrease of more than 25% in troponin T
concentration at 1 year, an increase of more than
25%, or a change (increase or decrease) of 25%
or less are shown in Table S9 in the Supplementary Appendix. A baseline troponin T concentration of 14 ng per liter or higher, a longer duration of diabetes, and a higher baseline glycated
hemoglobin level were independent predictors of
an increase of more than 25% in troponin T
concentration from baseline to 1 year (Table S10
in the Supplementary Appendix). In unadjusted
landmark analyses, among the 1984 patients
with both baseline and follow-up troponin T
values, the 4-year rate of the primary composite
end point of death from cardiovascular causes,
myocardial infarction, or stroke (Fig.3A) was
nonsignificantly higher and the rate of the secondary composite end point of death from any

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

615

The

n e w e ng l a n d j o u r na l

A 5-Yr Rate of Primary Composite End Point

100

Event Rate (%)

80
P<0.001
60
40
Troponin T 14 ng/liter 27.1
20
0

12.9
0

12

24

36

48

60

No. at Risk
897

737

684

620

455

255

1388

1281

1229

1124

892

529

B 5-Yr Rate of Death from Any Cause

100

Event Rate (%)

80
P<0.001
60
40

Troponin T <14 ng/liter

Troponin T 14 ng/liter 19.6

20

7.1
0

12

24

36

48

60

Months since Randomization

No. at Risk
Troponin T
14 ng/liter
Troponin T
<14 ng/liter

897

847

813

787

665

406

1388

1370

1355

1334

1160

739

Figure 1. Unadjusted KaplanMeier Estimates of Adverse Outcomes

over 5 Years.
The primary end point (Panel A) was a composite of death from cardiovascular causes, myocardial infarction, or stroke. Death from any cause (Panel B)
was a secondary end point. Data are shown for patients with an abnormal
high-sensitivity cardiac troponin T level (14 ng per liter) at baseline, as
compared with those with a normal level (<14 ng per liter) at baseline, who
were enrolled in the Bypass Angioplasty Revascularization Investigation in
Type 2 Diabetes (BARI 2D) trial.

cause, myocardial infarction, stroke, or heart

failure (Fig.3B) was significantly higher in those
who had an increase of more than 25% in troponin T concentration from baseline to 1 year,
as compared with those who had an increase or

616

m e dic i n e

decrease of 25% or less and with those who had

a decrease of more than 25%. In adjusted models, in which an increase or decrease in troponin
T concentration of 25% or less was considered
the reference category, both baseline troponin T
concentration and an increase of more than 25%
in troponin T concentration were independent
predictors of the primary composite end point
and the secondary composite end point (Table
S11 in the Supplementary Appendix).

Troponin T <14 ng/liter

Months since Randomization

Troponin T
14 ng/liter
Troponin T
<14 ng/liter

of

Discussion
In this study involving patients with both type 2
diabetes and stable ischemic heart disease, baseline cardiac troponin T concentrations above the
upper limit of normal were associated with approximately a doubling of the risks of myocardial infarction, stroke, heart failure, death from
cardiovascular causes, and death from any cause.
Nearly 40% of the patients in our study had a
high-sensitivity cardiac troponin T concentration at baseline that was above the upper reference limit used to define myocardial injury. The
incidence of the primary composite end point of
death from cardiovascular causes, myocardial
infarction, or stroke at 5 years in this group was
27%, which was double the rate in the group
with normal baseline troponin T concentrations.
Similar results were seen with respect to other
important outcomes, such as the secondary
composite outcome of death from any cause,
myocardial infarction, stroke, or heart failure.
The addition of prompt coronary revascularization to intensive medical therapy did not improve the outcome in these patients. Despite
aggressive medical therapy for type 2 diabetes
and stable ischemic heart disease, the median
troponin T concentration increased over 1 year
of follow-up, and no significant reductions in
troponin T concentrations were observed in patients who underwent coronary revascularization. Finally, the patients in the BARI 2D trial
who had an increase of more than 25% in troponin T concentration from baseline to the
1-year follow-up had a worse outcome than
those who had a stable or decreasing troponin T
concentration.
The strength of the relationship between troponin T concentration and the subsequent risk

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

Troponin in Stable Heart Disease and Diabetes

of myocardial infarction, stroke, heart failure,

death from cardiovascular causes, and death
from any cause suggests that high-sensitivity
cardiac troponin T concentration is a powerful
prognostic marker in patients who have both
type 2 diabetes and stable ischemic heart disease. The relationship that we observed in our
study was independent of the possible confounding variables of traditional risk factors, measures of diabetes severity, the extent of coronary
artery disease, electrocardiographic abnormalities, left ventricular ejection fraction, and renal
function. All the patients enrolled in the BARI
2D trial received intensive medical therapy, with
substantial and clinically significant reductions
in blood pressure and in levels of low-density
lipoprotein cholesterol, triglycerides, and glycated
hemoglobin during the course of the trial.13 Despite this intensive medical therapy, patients
with high-sensitivity troponin T concentrations
of 14 ng per liter or higher at baseline were at
substantial residual risk for death from cardiovascular causes, myocardial infarction, or stroke,
with an absolute risk of more than 27% at
5years.
Prompt revascularization did not appear to
alter the risk of adverse outcomes, including
death from any cause and heart failure, in patients with abnormal concentrations of troponin
T. These results differ from those of another
randomized trial of an early invasive strategy for
patients with acute coronary syndromes, in
which the benefit of that strategy was seen only
among patients with elevated troponin I concentrations.2 The use of high-sensitivity troponin
assays facilitates the diagnosis of myocardial
infarction,14,15 but the effects of this improved
accuracy on patient outcomes remain uncertain.
In one recent study involving patients with acute
coronary syndromes, troponin T concentration
was a predictor of adverse outcomes only among
those who received noninvasive treatment, and
the use of this measure identified patients who
were more likely to benefit from treatment with
ticagrelor than from treatment with clopidogrel.16 In our study involving patients with both
type 2 diabetes and stable ischemic heart disease, troponin T concentration was a predictor
of adverse cardiovascular outcomes regardless of
revascularization strategy, and prompt revascu-

Table 2. Hazard Ratio for Major Cardiovascular Events in Adjusted Models,

According to Baseline Cardiac Troponin T Concentration.*
Model

Troponin T <14 ng/liter

(N=1388)

Troponin T 14 ng/liter
(N=897)

P Value

hazard ratio (95% CI)

Model 1

Reference

2.35 (1.912.89)

<0.001

Model 2

Reference

2.00 (1.612.49)

<0.001

Model 3

Reference

1.85 (1.482.32)

<0.001

Model 4

Reference

1.56 (1.221.99)

<0.001

* In the subgroup of patients who had baseline high-sensitivity cardiac troponin
T concentrations of less than 14 ng per liter, 158 events of the primary composite end point (death from cardiovascular causes, myocardial infarction, or
stroke) occurred, representing an event rate of 12.9%; in the subgroup with
baseline troponin T concentrations of 14 ng per liter or more, 210 events of
the primary composite end point occurred, representing an event rate of
27.1% (P<0.001 for the comparison between the two subgroups).
Model 1 was adjusted for age, race, sex, and assignment to revascularization.
Model 2 was adjusted for age, race, sex, assignment to revascularization, systolic blood pressure, history of smoking, history of stroke or transient ischemic attack, history of myocardial infarction, history of heart failure, total and
high-density lipoprotein cholesterol, glycated hemoglobin, baseline insulin
use, body-mass index, and estimated glomerular filtration rate.
Model 3 was adjusted for the factors in model 2 plus the myocardial jeopardy
index, the number of coronary-artery lesions, left ventricular ejection fraction
less than 50%, and major electrocardiographic abnormalities.
Model 4 was adjusted for the factors in model 3 plus natural log-transformed
N-terminal probrain natriuretic peptide concentration.

larization did not lead to reductions in the composite outcome in patients with abnormal troponin T concentrations at baseline. The circulating
concentrations of cardiac troponin T observed in
the BARI 2D trial, however, were much lower
than those that are seen in patients with acute
coronary syndromes.2,16 Indeed, conventional
troponin T assays would have been unable to
detect any circulating cardiac troponin in most
of the patients in the BARI 2D trial. Our results
suggest that the factors leading to troponin release in patients with stable ischemic heart disease, including chronic injury from small-vessel
ischemia, hypertension, metabolic abnormalities, and renal dysfunction, may be less responsive to epicardial coronary revascularization
than the ischemic injury that leads to troponin
release in patients with acute coronary syndromes.7-9,16-18
The association of an abnormal troponin T
concentration with adverse outcome bears special emphasis because of the 40% prevalence of
abnormal troponin T concentration in the BARI
2D population. In this study, patients with

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

617

The

Revascularization
Group

Subgroup

n e w e ng l a n d j o u r na l

Medical-Therapy
Group

of

m e dic i n e

P Value for
Heterogeneity

Hazard Ratio (95% CI)

5-yr event rate (%)

Death from cardiovascular causes,
myocardial infarction, or stroke
Troponin T <14 ng/liter
Troponin T 14 ng/liter
Death from cardiovascular causes
Troponin T <14 ng/liter
Troponin T 14 ng/liter
Myocardial infarction
Troponin T <14 ng/liter
Troponin T 14 ng/liter
Stroke
Troponin T <14 ng/liter
Troponin T 14 ng/liter
Death from any cause
Troponin T <14 ng/liter
Troponin T 14 ng/liter
Heart failure
Troponin T <14 ng/liter
Troponin T 14 ng/liter
Death from any cause, myocardial
infarction, stroke, or heart failure
Troponin T <14 ng/liter
Troponin T 14 ng/liter

0.99
11.8
26.5

14.0
27.6

0.96 (0.711.30)
0.96 (0.741.25)

2.6
11.7

4.4
10.1

0.63 (0.341.14)
1.27 (0.851.91)

8.2
15.6

10.1
21.7

0.94 (0.661.35)
0.72 (0.521.00)

2.1
4.6

2.5
4.2

0.87 (0.431.77)
1.03 (0.512.05)

6.4
19.4

7.7
19.8

0.77 (0.531.14)
1.06 (0.801.39)

11.5
23.8

10.7
27.6

1.07 (0.781.47)
0.87 (0.671.13)

0.05

0.28

0.75

0.20

0.32

0.24
22.3
42.2

1.03 (0.831.29)
0.87 (0.711.05)

22.2
47.8
0.25

0.50

1.00

Revascularization
Better

2.00

4.00

Medical Therapy
Better

Figure 2. Hazard Ratios for the Primary Composite End Point and Selected Secondary End Points.
Data are shown for the patients who were randomly assigned to either prompt revascularization plus intensive medical therapy (revascularization group) or intensive medical therapy alone (medical-therapy group), stratified according to a normal (<14 ng per liter) or an abnormal (14 ng per liter) high-sensitivity cardiac troponin T concentration at baseline. The size of the boxes is proportional to the size of
the subgroups.

known coronary anatomy, stable coronary heart

disease, diabetes, and abnormal troponin T concentration did not benefit from percutaneous
coronary intervention, and the potential benefit
of coronary-artery bypass grafting was the same
regardless of baseline troponin T concentration.
These results, when viewed in the context of
other trials involving patients with chronic stable angina, emphasize the importance of taking
a detailed history in distinguishing stable from
unstable coronary heart disease and support
initial medical therapy as a reasonable first approach in patients who do not have unstable
coronary artery disease.11,19,20 The high prevalence of abnormal troponin T concentration, at
least according to the established threshold of
14 ng per liter, raises the possibility that the

618

widespread use of this assay in populations

similar to the population enrolled in the BARI
2D trial could lead to an increase in the use of
revascularization procedures that, on the basis
of our findings, appear to offer little benefit
with respect to the reduction of key outcomes,
including a composite of death from any cause,
myocardial infarction, stroke, or heart failure.
However, increasing the threshold at which
troponin T concentration would be considered
abnormal ignores the risk of adverse cardiovascular outcomes in patients with troponin T
concentrations between 14 and 23 ng per liter,
for whom there was also no evident benefit of
revascularization.
Revascularization did not appear to reduce
the proportion of patients who had an abnormal

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

Troponin in Stable Heart Disease and Diabetes

Event Rate (%)

Event Rate (%)

troponin T concentration at follow-up. An in>25% Decrease

25% Change
>25% Increase
crease of more than 25% in troponin T concentration at follow-up was uncommon among the A 4-Yr Rates of Death from Cardiovascular Causes, Myocardial Infarction,
or Stroke
patients in our study (approximately 7%), but we
P=0.12
P=0.04
P=0.08
found that the patients with this percentage in80
crease, as compared with those with a stable
70
troponin T concentration, were at an increased
60
risk for death from cardiovascular causes, myo50
cardial infarction, or stroke, even after the
40
analysis was adjusted for a wide array of possible
confounding variables. This observation is con30
sistent with the findings from previous studies
20
of the association between change in troponin
10
concentration and outcome in a population of
21
0
patients with hypercholesterolemia and in a
All Patients
Baseline Troponin T Baseline Troponin T
9
<14 ng/liter
14 ng/liter
population of elderly patients. However, troponin concentrations in our study appeared to be
No. of Events
17 162 19
2
73 13
15
89
6
136 1715 133
10 1123 112
126 592
21
No. at Risk
more stable over time than they were in those
populations. These observations raise the posB 4-Yr Rates of Death from Any Cause, Myocardial Infarction, Stroke,
sibility that serial measurements of troponin
or Heart Failure
concentration may improve its prognostic value
P=0.04
P=0.002
P<0.001
80
and that persistently elevated and increasing
70
troponin concentrations may be the best predictors of adverse outcomes.
60
The strengths of our study include the ran50
dom assignment to prompt revascularization
40
plus intensive medical therapy or to intensive
30
medical therapy alone and the relatively large
20
number of patients in the subgroup with abnor10
mal troponin T concentrations. The limitations
include the lack of power to detect a treatment
0
All Patients
Baseline Troponin T Baseline Troponin T
effect of revascularization in subgroups such as
<14 ng/liter
14 ng/liter
those defined according to baseline troponin
No. of Events
26 359 39
3
158 28
23 201
11
concentration and the fact that 13% of the pa136 1715 133
10 1123 112
126 592
21
No. at Risk
tients who had troponin T concentrations availFigure 3. Unadjusted KaplanMeier Estimates.
able at baseline did not have follow-up values
Unadjusted KaplanMeier estimates are shown for the rate of the primary
available. Although 14 ng per liter is the current
composite end point of death from cardiovascular causes, myocardial inestablished upper reference limit for troponin T
farction, or stroke (Panel A) and for the rate of the secondary composite
in both men and women, there has been interest
end point of death from any cause, myocardial infarction, stroke, or heart
in developing more age- and sex-specific referfailure (Panel B), according to the change in high-sensitivity cardiac tropo1,22,23
ence limits.
nin T concentration from baseline to 1-year follow-up. The 4-year event
rates are derived from a landmark analysis starting at the time of plasma
In conclusion, the results of this study showed
sample collection at 1 year and continuing to the close of the study 4 years
a strong, consistent association between baselater. I bars indicate 95% confidence intervals. Data on the numbers of
line concentrations of circulating cardiac tropoevents and numbers at risk are for the 1984 patients in the BARI 2D trial
nin T and the risk of death from any cause,
who had both baseline and 1-year follow-up troponin T concentrations
myocardial infarction, stroke, and heart failure
available.
in patients with both type 2 diabetes and stable
ischemic heart disease. In this population, 40%
of patients had troponin T concentrations above no differential benefit of prompt coronary revasthe upper reference limit of 14 ng per liter, and cularization was observed.

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.

619

Troponin in Stable Heart Disease and Diabetes

Supported by grants (U01HL061744, U01HL061746,

U01HL061748, U01HL063804, and R21HL121495) from the
National Heart, Lung, and Blood Institute and the National Institute of Diabetes and Digestive and Kidney Diseases of the
National Institutes of Health and by an investigator-initiated
research grant from Roche Diagnostics (to Dr. Everett).
Dr. Everett reports receiving grant support from Roche and
Novartis and fees for serving on a clinical events committee
from Genzyme. Dr. Brooks reports receiving grant support from
Gilead Sciences. Dr. Chaitman reports receiving fees for serving
on clinical events committees from Merck, Eli Lilly, Novo Nordisk,
and Roche and fees for serving on data safety monitoring
boards from Pfizer and Sanofi-Aventis. Dr. Bhatt reports serving

References
1. Thygesen K, Alpert JS, Jaffe AS, et al.
Third universal definition of myocardial
infarction. J Am Coll Cardiol 2012;
60:
1581-98.
2. Morrow DA, Cannon CP, Rifai N, et al.
Ability of minor elevations of troponins I
and T to predict benefit from an early invasive strategy in patients with unstable
angina and non-ST elevation myocardial
infarction: results from a randomized
trial. JAMA 2001;286:2405-12.
3. Amsterdam EA, Wenger NK, Brindis
RG, et al. 2014 AHA/ACC Guideline for
the Management of Patients with Non-STElevation Acute Coronary Syndromes: a
report of the American College of Cardiology/American Heart Association Task
Force on Practice Guidelines. J Am Coll
Cardiol 2014;64(24):e139-e228.
4. Bhatt DL, Roe MT, Peterson ED, et al.
Utilization of early invasive management
strategies for high-risk patients with nonST-segment elevation acute coronary syndromes: results from the CRUSADE Quality Improvement Initiative. JAMA 2004;
292:2096-104.
5. Omland T, de Lemos JA, Sabatine MS,
et al. A sensitive cardiac troponin T assay
in stable coronary artery disease. N Engl J
Med 2009;361:2538-47.
6. Omland T, Pfeffer MA, Solomon SD,
et al. Prognostic value of cardiac troponin
I measured with a highly sensitive assay in
patients with stable coronary artery disease. J Am Coll Cardiol 2013;61:1240-9.
7. Everett BM, Cook NR, Magnone MC,
et al. Sensitive cardiac troponin T assay
and the risk of incident cardiovascular
disease in women with and without diabetes mellitus: the Womens Health Study.
Circulation 2011;123:2811-8.
8. de Lemos JA, Drazner MH, Omland T,
et al. Association of troponin T detected
with a highly sensitive assay and cardiac
structure and mortality risk in the general
population. JAMA 2010;304:2503-12.
9. deFilippi CR, de Lemos JA, Christen-

620

on advisory boards for Cardax, Medscape Cardiology, and Regado Biosciences, receiving fees for serving on a steering committee from WebMD, receiving research funding through his
institution from Amarin, AstraZeneca, Bristol-Myers Squibb,
Eisai, Ethicon, Forest Laboratories, Ischemix, Medtronic, Pfizer,
Roche, Sanofi-Aventis, and the Medicines Company; participating in unfunded research collaborations with FlowCo, PLx
Pharma, and Takeda; and serving as an investigator in studies
funded by Biotronik and St. Jude Medical through his institution. No other potential conflict of interest relevant to this article was reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.

son RH, et al. Association of serial measures of cardiac troponin T using a sensitive assay with incident heart failure and
cardiovascular mortality in older adults.
JAMA 2010;304:2494-502.
10. Saunders JT, Nambi V, de Lemos JA,
et al. Cardiac troponin T measured by a
highly sensitive assay predicts coronary
heart disease, heart failure, and mortality
in the Atherosclerosis Risk in Communities Study. Circulation 2011;123:1367-76.
11. Frye RL, August P, Brooks MM, et al.
A randomized trial of therapies for type 2
diabetes and coronary artery disease.
N Engl J Med 2009;360:2503-15.
12. Bypass Angioplasty Revascularization
Investigation 2 Diabetes Study Group.
Baseline characteristics of patients with
diabetes and coronary artery disease enrolled in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI
2D) trial. Am Heart J 2008;156:528-36.
13. Chaitman BR, Hardison RM, Adler D,
et al. The Bypass Angioplasty Revascularization Investigation 2 Diabetes randomized trial of different treatment strategies
in type 2 diabetes mellitus with stable
ischemic heart disease: impact of treatment strategy on cardiac mortality and
myocardial infarction. Circulation 2009;
120:2529-40.
14. Reichlin T, Hochholzer W, Bassetti S,
et al. Early diagnosis of myocardial infarction with sensitive cardiac troponin
assays. N Engl J Med 2009;361:858-67.
15. Keller T, Zeller T, Peetz D, et al. Sensitive troponin I assay in early diagnosis of
acute myocardial infarction. N Engl J Med
2009;361:868-77.
16. Wallentin L, Lindholm D, Siegbahn A,
et al. Biomarkers in relation to the effects
of ticagrelor in comparison with clopidogrel in non-ST-elevation acute coronary
syndrome patients managed with or without in-hospital revascularization: a substudy from the Prospective Randomized
Platelet Inhibition and Patient Outcomes

(PLATO) trial. Circulation 2014;129:293303.

17. White HD. Pathobiology of troponin
elevations: do elevations occur with myocardial ischemia as well as necrosis? J Am
Coll Cardiol 2011;57:2406-8.
18. Bergmann O, Bhardwaj RD, Bernard S,
et al. Evidence for cardiomyocyte renewal
in humans. Science 2009;324:98-102.
19. Boden WE, ORourke RA, Teo KK,
et al. Optimal medical therapy with or
without PCI for stable coronary disease.
N Engl J Med 2007;356:1503-16.
20. Fihn SD, Gardin JM, Abrams J, et al.
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.
Circulation 2012;126(25):e354-e471.
21. White HD, Tonkin A, Simes J, et al.
Association of contemporary sensitive
troponin I levels at baseline and change at
1 year with long-term coronary events following myocardial infarction or unstable
angina: results from the LIPID Study
(Long-Term Intervention With Pravastatin
in Ischaemic Disease). J Am Coll Cardiol
2014;63:345-54.
22. Gore MO, Seliger SL, Defilippi CR, et
al. Age- and sex-dependent upper reference limits for the high-sensitivity cardiac
troponin T assay. J Am Coll Cardiol 2014;
63:1441-8.
23. Apple FS, Jaffe AS. Men are different
than women: its true for cardiac troponin
too. Clin Biochem 2014;47:867-8.
Copyright 2015 Massachusetts Medical Society.

n engl j med 373;7nejm.org August 13, 2015

The New England Journal of Medicine

Downloaded from nejm.org on September 10, 2015. For personal use only. No other uses without permission.
Copyright 2015 Massachusetts Medical Society. All rights reserved.