Abstract
Background:
Low values of high-sensitivity cardiac troponin (hs-cTn) and coronary artery calcium (CAC) scores of zero are associated with a low risk for atherosclerotic cardiovascular disease (ASCVD).
Objective:
Evaluate baseline hs-cTnT and CAC in relation to ASCVD
Methods:
Baseline hs-cTnT (limit of detection, LoD 3 ng/L) and CAC measurements were analyzed across participants aged 45–84 years without clinical cardiovascular disease from the prospective, Multi-Ethnic Study of Atherosclerosis (MESA) in relationship to incident ASCVD.
Results:
Among 6749 participants, 1002 ASCVD events occurred during a median follow up of 15 years. Participants with detectable CAC (20.1 versus 5.0 events per 1000 person-years; adjusted hazard ratio [HR]: 2.35, 95% CI: 2.0-2.76, p<0.001) and detectable hs-cTnT (15.4 versus 5.2 per 1000 person-years; adjusted HR: 1.47, 95% CI: 1.21-1.77, p<0.001) had higher rates of ASCVD than those with undetectable results. Individuals with undetectable hs-cTnT (32%) had similar risk for ASCVD as did those with a CAC of zero (50%) (5.2 vs. 5.0 per 1000 person-years). Together, hs-cTnT and CAC (discordance 38%) resulted in the following ASCVD event rates: hs-cTnT <LoD/CAC=0: 2.8 per 1000 person-years (reference), hs-cTnT≥LoD/CAC=0: 6.8 per 1000 person-years (HR 1.59, 95% CI: 1.17-2.16, p=0.003), hs-cTnT<LoD/CAC≥0: 11.1 per 1000 person-years (HR 2.74, 95% CI: 1.96-3.83, p<0.00001), and hs-cTnT≥LoD/CAC>0: 22.6 per 1000 person-years (HR 3.50, 95% CI: 2.60-4.70, p<0.00001).
Conclusions:
An undetectable hs-cTnT identifies patients with a similar, low risk for ASCVD as those with a CAC score of zero. The increased risk among those with discordant results supports their conjoint use for risk prediction.
Keywords: Coronary artery calcium, High-sensitivity cardiac troponin, Atherosclerotic cardiovascular disease, Risk-stratification, Primary prevention
CONDENSED ABSTRACT:
High-sensitivity cardiac troponin (hs-cTn) and coronary artery calcium (CAC) scores are markers of subclinical atherosclerosis and myocardial injury that are associated with a low risk for atherosclerotic cardiovascular disease (ASCVD) when undetectable. In the Multi-Ethnic Study of Atherosclerosis (MESA) we identified that an undetectable hs-cTnT identifies patients with a similar, low risk for ASCVD as those with a CAC score of zero. The frequent discordance between CAC and hs-cTnT and the increased risk among those with discordant results indicate that their prognostic information is complementary, favoring their conjoint use for risk prediction.
Tweet:
The conjoint use of high-sensitivity cardiac troponin T and CAC scoring for ASCVD risk-prediction: The Multi-Ethnic Study of Atherosclerosis (MESA).
Cardiovascular disease remains the leading cause of death in the United States (US). It is estimated that by 2035 almost half of the population will have cardiovascular disease, with projected costs of over 1 trillion dollars (1). Simple cost-effective approaches that accurately estimate the risk for atherosclerotic cardiovascular disease (ASCVD) are a major public health interest.
Coronary artery calcium (CAC) scoring facilitates risk-stratification for ASCVD (2–8). Primary prevention and cholesterol guidelines endorse its use to guide shared decision making about preventive therapies (9–11). The presence of CAC identifies patients at higher risk for ASCVD in whom evidence-based therapies should be considered. Its absence identifies a lower risk subset of patients in whom statin therapy may be withheld or postponed, with a CAC score of zero recognized as a potent negative risk factor (5–6, 9–10).
High-sensitivity (hs) cardiac troponin (cTn) I and T assays quantify cTn in most healthy men and women and facilitate risk stratification for cardiovascular disease in both acute and outpatient settings (12–22). In patients with suspected acute coronary syndrome, those with undetectable (below the limit of detection, LoD) hs-cTn concentrations are at low short and long term risk for adverse events (13–16).
Whether undetectable hs-cTn concentrations can be used in the primary prevention setting to identify the same individuals at low risk for ASCVD as a CAC of zero does is uncertain. We hypothesized that individuals with an undetectable hs-cTnT have a similar risk for ASCVD as those with undetectable CAC. To address this issue, we evaluated baseline hs-cTnT and CAC in relation to long-term incident ASCVD in the Multi-Ethnic Study of Atherosclerosis (MESA) cohort.
Methods
MESA is a longitudinal, community-based, multi-ethnic, cohort study funded by the National Heart, Lung, and Blood Institute (NHBLI). The study was approved by the Institutional Review Boards at all Field Centers and all participants gave written informed consent. Study design and methods have been reported (23) and are available at the MESA website (www.mesa-nhlbi.org). From 2000 to 2002, MESA enrolled 6814 participants aged 45 to 84 years who were free of clinical cardiovascular disease at baseline. Four racial/ethnic groups were enrolled from six US communities including Baltimore, MD; Chicago, IL; Forsyth County, NC; Los Angeles, CA; New York, NY; and St. Paul, MN. The present analysis focuses on 6749 participants with hs-cTnT and CAC measured at baseline who had follow up for incident ASCVD.
Coronary artery calcium scoring
The methods for cardiac computed tomography (CT) and CAC measurement have been described (23–24). Chest CT was performed with a cardiac-gated electron-beam CT scanner (Chicago, Los Angeles, and New York) or an electrocardiogram-triggered, multidetector CT system (Baltimore, Forsyth County, and St. Paul). All participants were scanned twice over phantoms of known physical calcium concentrations and scans were read centrally at the Harbor-UCLA Research and Education Institute (Torrance, California, USA). CAC was quantified using an averaged Agatston score. Kappa for intra- and inter-observer were 0.93 (95% confidence interval [CI]: 0.83-0.99) and 0.90 (95% CI: 0.81-0.99) (24).
High-sensitivity cardiac troponin T
The methods used for hs-cTnT measurement have been reported (25). Hs-cTnT was measured at examination 1 from ethylenediaminetetraacetic acid (EDTA) plasma samples. Hs-cTnT was measured using the Cobas e601 analyzer (Roche Diagnostics, Switzerland). For this instrument, the limit of detection of the assay (LOD) is 3 ng/L. With this analyzer, total imprecision of 10% occurs at 4.3 ng/L and 20% at 2.5 ng/L (25). Sex-specific 99th percentile upper-reference limits (URLs) of 10 ng/L for women and 15 ng/L for men were used (26–27).
Outcomes
Incident ASCVD was the primary composite outcome. It was defined prospectively from baseline exam as time to first instance of myocardial infarction, resuscitated cardiac arrest, fatal coronary heart disease, definite or probable angina (if followed by revascularization), fatal and non-fatal stroke, other atherosclerotic death, or other cardiovascular death. Each potential event was classified using standardized criteria as described in the MESA manual of operations (www.mesa-nhlbi.org) (23).
Each participant was contacted every 9-12 months. For classification of events occurring during follow-up in MESA, information was collected from death certificates, medical records, autopsy reports, interviews with participants, and in case of out-of-hospital deaths, interviews with physicians, relatives, or friends (23).
Statistical analysis
Participants were evaluated based on whether CAC and hs-cTnT were detectable (CAC score >0 and hs-cTnT ≥3 ng/L) or undetectable (CAC score of 0 and hs-cTnT <3 ng/L). Baseline characteristics were compared with Kruskal-Wallis or Chi-square tests as appropriate. Concordance between undetectable/detectable hs-cTnT and CAC was assessed using Cohen’s kappa coefficient. Cumulative incidence of ASCVD, including 10-year proportion with ASCVD, were estimated using the Kaplan-Meier method. Cox proportional hazards regression adjusting for age, sex, race/ethnicity, smoking status, body mass index, education, systolic blood pressure, hypertension medication use, diabetes, total and HDL cholesterol was used to evaluate the associations of CAC and hs-cTnT with ASCVD. As a sensitivity analysis, a competing risk analysis was run with non-cardiovascular death as a competing risk of ASCVD for CAC, hs-cTnT, and CAC/hs-cTnT groups. The relationship between CAC and hs-cTnT with ASCVD risk was assessed using smoothing splines. All comparisons were made overall and separately for each race/ethnicity and sex. A p-value of <0.05 was considered statistically significant.
Results
The study cohort consisted of 6749 participants with hs-cTnT and CAC measured at baseline in who follow up data for incident ASCVD was available. Women comprised 53% of the cohort. Overall, 39% of the cohort was non-Hispanic white, 12% Chinese, 28% Black, and 22% Hispanic-American. CAC was undetectable in 3379 (50%) of participants (men 39%, women 60%), Agatston 1 to 100 in 1784 (26%) (men 29%, women 24%), and >100 in 1586 (24%) (men 32%, women 16%). Hs-cTnT was detectable (≥3 ng/L) in 68% of participants, including 83% of men and 55% of women. At baseline, hs-cTnT was undetectable (<LoD) in 2150 (32%) participants (men 18%, women 45%), measurable (≥LoD to sex-specific <99th percentile URL) in 4026 (60%) (men 73%, women 48%), and increased above sex-specific 99th percentiles in 573 (9%) (men 9%, women 8%) (Supplemental Figure 1).
Baseline characteristics for the entire cohort and in relationship to detectable and undetectable hs-cTnT and/or CAC results are shown in Table 1. As compared to patients with detectable results, patients with undetectable hs-cTnT and/or CAC were younger, more often women, and less likely to have comorbidities. Similar findings were observed for hs-cTnT when data were stratified by sex and race/ethnicity (Supplemental Tables 1–6). Among those with discordant CAC/hs-cTnT results (38%), except for a higher proportion of women in those with undetectable hs-cTnT and detectable CAC, comorbidities were similar. The relationship between hs-cTnT and CAC is shown in Supplemental Figure 2. Concordance between undetectable/detectable hs-cTnT and CAC demonstrated an agreement rate of 62% (Cohen’s k: 0.24, 95% CI 0.22-0.26), which varied slightly by race/ethnicity (Supplemental Table 7).
Table 1.
Baseline characteristics for the entire cohort and in relationship to detectable and undetectable hs-cTnT and CAC.
| Total n=6749 | CAC | Hs-cTnT | CAC and hs-cTnT | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CAC=0 n=3379 (50%)) | CAC>0 N=3370 (50%) | P value | <LoD n=2150 (32%) | ≥LoD n=4599 (68%) | P value | Hs-cTnT <LoD and CAC=0 n=1480 (22%) | Hs-cTnT ≥LoD and CAC=0 n=1899 (28%) | Hs-cTnT <LoD and CAC >0 n=670 (10%) | Hs-cTnT ≥LoD and CAC >0 n=2700 (40%) | P value | ||
| Age, mean(SD) | 62 (10) | 58 (9) | 66 (10) | <0.0001 | 57 (9) | 65 (10) | <0.0001 | 55 (8) | 60 (10) | 61 (9) | 68 (9) | <0.0001 |
| Women, n (%) | 3563 (53) | 2141 (63) | 1422 (42) | <0.0001 | 1591 (74) | 1972 (43) | <0.0001 | 1152 (78) | 989 (52) | 439 (66) | 983 (36) | <0.0001 |
| Race/ethnicity | <0.0001 | <0.0001 | <0.0001 | |||||||||
| White, n (%) | 2602 (3939) | 1117 (3333) | 1485 (44) | - | 741 (3535) | 1861 (4141) | - | 488 (3333) | 629 (3333) | 253 (3838) | 1232 (46) | - |
| Chinese, n (%) | 798 (1212) | 396 (1212) | 402 (1212) | - | 330 (1515) | 468 (1010) | - | 204 (1414) | 192 (1010) | 126 (1919) | 276 (1010) | - |
| Black, n (%) | 1864 (2828) | 1053 (3131) | 811 (2424) | - | 565 (2626) | 1299 (2828) | - | 408 (2828) | 645 (3434) | 157 (2323) | 654 (2424) | - |
| Hispanic, (%) | 1485 (2222) | 813 (2424) | 672 (2020) | - | 514 (2424) | 971 (2121) | - | 380 (2626) | 433 (2323) | 134 (2020) | 538 (2020) | - |
| Never smoked, n (%) | 3388 (50) | 1886 (56) | 1502 (45) | <0.0001 | 1196 (56) | 2192 (48) | <0.0001 | 880 (60) | 1006 (53) | 316 (47) | 1186 (44) | <0.0001 |
| Body mass index, mean (SD) | 28.3 (5.5) | 28 (66) | 28 (55) | 0.17 | 27.9 (5.7) | 28.5 (5.4) | <0.0001 | 28 (66) | 29 (66) | 28 (66) | 29 (55) | <0.0001 |
| Physical activity (MET-min/week) | 5754 (5894) | 6012 (5808) | 5496 (5968) | <0.0001 | 5827 (5380) | 5720 (6119) | 0.001 | 5907 (5337) | 6093 (6150) | 5650 (5475) | 5457 (6085) | <0.0001 |
| Diabetes, n (%) | 844 (1313) | 311 (99) | 533 (1616) | <0.0001 | 155 (7.2) | 689 (1515) | <0.0001 | 86 (66) | 225 (1212) | 69 (1010) | 464 (1717) | <0.0001 |
| Hypertension, n (%) | 3028 (45) | 1185 (3535) | 1843 (55) | <0.0001 | 649 (3030) | 2379 (52) | <0.0001 | 379 (2626) | 806 (42) | 270 (4040) | 1573 (58) | <0.0001 |
| Systolic blood pressure (mmHg), mean (SD) | 126.6 (2222) | 122 (2020) | 131 (2222) | <0.0001 | 120.1 (2020) | 129.6 (2222) | <0.0001 | 118 (1919) | 126 (2121) | 125 (2020) | 132 (2222) | <0.0001 |
| Diastolic blood pressure (mmHg), mean (SD) | 71.9 (1010) | 71 (1010) | 73 (1010) | <0.0001 | 70.0 (9.9) | 72.8 (1010) | <0.0001 | 70 (1010) | 73 (1010) | 71 (99) | 73 (1010) | <0.0001 |
| Hypertension medication use, n (%) | 2509 (3737) | 971 (2929) | 1538 (46) | <0.0001 | 537 (2525) | 1972 (43) | <0.0001 | 317 (2121) | 654 (3434) | 220 (3232) | 1318 (49) | <0.0001 |
| Total cholesterol (mg/dL), mean (SD) | 194 (3636) | 194 (3535) | 195 (3636) | 0.58 | 196 (3535) | 193 (3636) | 0.0021 | 195 (3535) | 193 (3535) | 198 (3535) | 194 (3737) | 0.003 |
| LDL cholesterol (mg/dL), mean (SD) | 117 (3131) | 115.9 (3131) | 118.4 (3232) | 0.008 | 118 (3232) | 117 (3131) | 0.27 | 117 (3131) | 115 (3131) | 121 (3333) | 118 (3232) | 0.007 |
| HDL cholesterol (md/dL), mean (SD) | 51 (1515) | 52.5 (1515) | 49.4 (1515) | <0.0001 | 53 (1515) | 50 (1515) | <0.0001 | 54 (1515) | 51 (1515) | 51 (1414) | 49 (1515) | <0.0001 |
| Triglycerides (mg/dL), mean (SD) | 132 (88) | 127 (85) | 136 (92) | <0.0001 | 126 (80) | 134 (93) | 0.0002 | 122 (73) | 131 (94) | 134 (93) | 137 (92) | <0.0001 |
| Lipid-lowering medication use, n (%) |
1098 (1616) | 360 (1111) | 738 (2222) | <0.0001 | 258 (1212) | 840 (1818) | <0.0001 | 128 (99) | 232 (1212) | 130 (1919) | 608 (2323) | <0.0001 |
| Any coronary artery calcium present, n (%) | 3370 (50) | <0.0001 | 670 (3131) | 2700 (59) | <0.0001 | 0 (0) | 0 (0) | 146 (292) | 327 (594) | |||
| Coronary artery calcium, mean (SD) | 145 (416) | 0 (0) | 291 (552) | <0.0001 | 45 (176) | 192 (483) | <0.0001 | <0.0001 | ||||
| Coronary artery calcium score | - | <0.0001 | 1480(100) | 1899(100) | 0 (0) | 0 (0) | - | |||||
| Coronary artery calcium score, 0, n (%) | 3379 (50) | 3379 (100 ) | 0 (0.0) | 1480 (69) | 1899 (4141) | 0 (0) | 0 (0) | 453 (68) | 1331 (49) | |||
| Coronary artery calcium score, 1-99, n (%) | 1784 (2626) | 0 (0.0) | 1784 (53) | 453 (2121) | 1331 (2929) | 0 (0) | 0 (0) | 217 (3232) | 1369 (51) | |||
| Coronary artery calcium score >100, n (%) | 1586 (2424) | 0 (0.0) | 1586 (47) | 217 (1010) | 1369 (3030) | |||||||
Outcomes
A total of 1002 ASCVD events occurred during a median follow up of 15 years. Participants with detectable CAC had a higher incidence rate of ASCVD than those with undetectable CAC (Table 2, Supplemental Table 8). In multivariable Cox regression analyses, detectable CAC was an independent risk factor for ASCVD (20.1 versus 5.0 events per 1000 person-years; adjusted hazard ratio [HR] 2.35, 95% CI: 2.0-2.76, p<0.001). Competing risk analysis with non-cardiovascular death showed similar results Similar findings were observed for men (22.7 versus 6.0 events per 1000 person-years; adjusted HR: 2.55, 95% CI: 2.03-3.21, p<0.001) and women (16.7 versus 4.4 per 1000 person-years; adjusted HR: 2.14, 95% CI: 1.69-2.70, p<0.001). Cumulative incidence curves for the entire cohort and by sex based on the presence or absence of CAC are shown in Figure 1A–C. The relationship between baseline CAC scores and the risk for ASCVD is shown in Figure 2A.
Table 2.
ASCVD event rates in relationship to detectable and undetectable hs-cTnT and CAC for the entire cohort and by sex.
| CAC | Hs-cTnT | CAC and hs-cTnT | ||||||
|---|---|---|---|---|---|---|---|---|
| CAC=0 | CAC>0 | <LoD | ≥LoD | Hs-cTnT <LoD and CAC=0 | Hs-cTnT ≥LoD and CAC=0 | Hs-cTnT <LoD and CAC >0 | Hs-cTnT ≥LoD and CAC >0 | |
| Overall cohort | ||||||||
| Total (n) | 3379 | 3370 | 2150 | 4599 | 1480 | 1899 | 670 | 2700 |
| ASCVD events (n) | 228 | 774 | 150 | 852 | 57 | 171 | 93 | 681 |
| Total-person-years | 45721 | 38491 | 28791 | 55422 | 20418 | 25304 | 8373 | 30118 |
| Crude CVD rate per 1,000 person-years | 5.0 | 20.1 | 5.2 | 15.4 | 2.8 | 6.8 | 11.1 | 22.6 |
| 10-year proportion with ASCVD (%) (95% CI) | 3.6 (3.0-4.3) | 17.5 (16.1-18.8) | 4.7 (3.7-5.6) | 13.2 (12.2-14.2) | 2.4 (1.6-3.2) | 4.6 (3.6-5.5) | 9.8 (7.5-12.2) | 19.4 (17.8-21.0) |
| Male | ||||||||
| Total (n) | 1238 | 1948 | 559 | 2627 | 328 | 910 | 231 | 1717 |
| ASCVD events (n) | 100 | 496 | 54 | 542 | 15 | 85 | 39 | 457 |
| Total-person-years | 16539 | 21846 | 7373 | 31011 | 4520 | 12019 | 2854 | 18992 |
| Crude CVD rate per 1,000 person-years | 6.0 | 22.7 | 7.3 | 17.5 | 3.3 | 7.1 | 13.7 | 24.1 |
| 10-year proportion with ASCVD (%) (95% CI) | 4.1 (3.0-5.3) | 19.5 (17.6-21.3) | 6.0 (4.0-8.0) | 15.0 (13.6-16.4) | 2.0 (0.4-3.5) | 4.9 (3.4-.6.3) | 11.9 (7.5-16.1) | 20.6 (18.5-22.6) |
| Female | ||||||||
| Total (n) | 2141 | 1422 | 1591 | 1972 | 1152 | 989 | 439 | 983 |
| ASCVD events (n) | 128 | 278 | 96 | 310 | 42 | 86 | 54 | 224 |
| T otal-person-years | 29183 | 16646 | 21417 | 24411 | 15898 | 13284 | 5519 | 11127 |
| Crude CVD rate per 1,000 person-years | 4.4 | 16.7 | 4.5 | 12.7 | 2.6 | 6.5 | 9.8 | 20.1 |
| 10-year proportion with ASCVD (%) (95% CI) | 3.3 (2.5-4.1) | 14.7 (12.8-16.7) | 4.5 (3.2-5.2) | 10.7 (9.3-12.1) | 2.5 (1.6-3.4) | 4.3 (3.0-5.6) | 8.8 (6.011.5) | 17.4 (14.9-19.9) |
Figure 1. Cumulative incidence of ASCVD for undetectable/detectable CAC.
Panels A-C show the cumulative incidence of ASCVD for a CAC score of zero (CAC=0) (black line) and for detectable CAC (CAC>0) (red line) for the overall cohort (panel A), men (panel B), and women (panel C).
Figure 2. Cumulative incidence of ASCVD for undetectable/detectable hs-cTnT.
Panels A-C show the cumulative incidence of ASCVD for undetectable hs-cTnT (<LoD) (black line) and detectable hs-cTnT (>=LoD) (red line) for the overall cohort (A), men (B), and women (C).
Participants with detectable hs-cTnT (≥3 ng/L) had a higher risk of ASCVD than those with undetectable hs-cTnT (<3 ng/L) (Table 2 and Online Table 8). In multivariable Cox regression analyses, detectable hs-cTnT was an independent risk factor for incident ASCVD (15.4 versus 5.2 events per 1000 person-years; adjusted HR: 1.47, 95% CI: 1.21-1.77, p<0.001). Competing risk analysis with non-cardiovascular death showed similar results Similar findings were observed for men (17.5 versus 7.3 events per 1000 person-years; adjusted HR: 1.31, 95% CI: 0.98-1.76, p=0.071) and women (12.7 versus 4.5 events per 1000 person-years; adjusted HR: 1.52, 95% CI: 1.18-1.95, p=0.001). Cumulative incidence curves for the entire cohort and by sex based on detectable versus undetectable baseline hs-cTnT concentrations are shown in Figure 3A–C. The relationship between baseline hs-cTnT and the risk for ASCVD is shown Figure 2B.
Figure 3. Relationship between CAC and hs-cTnT and the risk for ASCVD.
Panel A shows the relationship between CAC and the hazard ratio for ASCVD. Panel B shows the relationship between hs-cTnT concentrations and the hazard ratio for ASCVD.
Long-term ASCVD outcomes, including 10-year proportion with ASCVD, stratified by detectable/undetectable hs-cTnT and CAC are shown in Table 2. Individuals with undetectable hs-cTnT (32% of participants) had a similar risk for incident ASCVD as those with a CAC of zero (50% of participants) (5.2 vs. 5.0 events per 1000 person-years). Together, hs-cTnT and CAC (discordance 38%) resulted in the following crude ASCVD rates (Central Illustration, Table 2, Figure 4): hs-cTnT<LoD/CAC=0 (n=1480): 2.8 events per 1000 person-years (reference), hs-cTnT≥LoD/CAC=0 (n=1899): 6.8 events per 1000 person-years (adjusted HR 1.59, 95% CI: 1.17-2.16, p=0.003), hs-cTnT<LoD/CAC>0 (n=670): 11.1 events per 1000 person-years (adjusted HR 2.74, 95% CI: 1.96-3.83, p<0.0001), and hs-cTnT≥LoD/CAC>0 (n=2700): 22.6 events per 1000 person-years (adjusted HR 3.50, 95% CI: 2.60-4.70, p<0.0001). Adjusted analyses showed that both hs-cTnT and CAC, as binary variables, were independent predictors of ASCVD (hs-cTnT: adjusted HR=1.38, 95% CI: 1.14 – 1.67, p=0.0009; and CAC: adjusted HR = 2.31, 95% CI: 1.96 – 2.72, p<0.0001). Competing risk analysis with non-cardiovascular death showed similar results. Sensitivity analyses evaluating the association of detectable/undetectable hs-cTnT and/or CAC with coronary heart disease events are shown in Supplemental Table 9. Similar results were observed.
Central Illustration. Independent and conjoint use of CAC and hs-cTnT for ASCVD risk prediction.
Section A: both a CAC score of zero and undetectable hs-cTnT result in similar ASCVD risk prediction. Section B: Conjoint use of CAC and hs-cTnT improves risk-stratification, particularly in those with discordant results (38%).
Figure 4. Cumulative incidence of ASCVD for undetectable/detectable CAC/hs-cTnT.
Cumulative incidence of ASCVD for CAC=0/hs-cTnT<LoD (dark line), CAC=0/hs-cTnT>=LoD (red line), CAC>0/hs-cTnT<LoD (green line), and CAC>0/hs-cTnT=>LoD (blue line).
Among participants with a CAC of zero (n=3379), 228 ASCVD events occurred. Those with a CAC of zero and detectable hs-cTnT (56%) had a higher risk for ASCVD than those with a CAC of zero and undetectable hs-cTnT (6.8 versus 2.8 events per 1000 person-years; adjusted HR: 1.66, 95% CI: 1.20-2.29, p=0.002) (Table 2, Figure 5). Similar findings were observed for men (hs-cTnT≥LoD: 7.1 versus hs-cTnT<LoD: 3.3 per 1000 person-years) and women (hs-cTnT≥LoD: 6.5 versus hs-cTnT<LoD: 2.6 per 1000 person-years) with a CAC of zero. Adjusted analyses showed a significant association between detectable hs-cTnT and ASCVD in women (HR: 1.7, 95% CI: 1.15-2.52, p=0.008), but not in men (HR: 1.49, 95% CI: 0.83-2.65). Using hs-cTnT as a continuous variable (log transformed), adjusted analyses showed hs-cTnT as an independent predictor of ASCVD among participants with a CAC of zero (HR 1.60, 95% CI: 1.27-2.01, p<0.0001); similar findings were observed for men (HR 1.84, 95% CI: 1.33-2.56, p=0.0003) and women (HR 1.48, 95% CI 1.04-2.09, p=0.029).
Figure 5. Cumulative incidence of ASCVD for undetectable/detectable hs-cTnT among patients with a CAC score of zero.
Cumulative incidence of ASCVD according to undetectable (<LoD) vs. detectable (>=LoD) hs-cTnT for all patients with a CAC score of zero (A), men (B), and women (C).
Among participants with undetectable hs-cTnT (n=2150), 150 ASCVD events occurred. Those with undetectable hs-cTnT and detectable CAC (31%) had a higher risk for ASCVD than those with undetectable hs-cTnT and a CAC of zero (11.1 versus 2.8 events per 1000 person-years; adjusted HR 2.91; 95% CI: 2.03-4.16, p<0.0001). Similar findings were observed for men (CAC>0: 13.7 versus CAC=0: 3.3 per 1000 person-years) and women (CAC>0: 9.8 versus CAC=0: 2.6 per 1000 person years) with undetectable hs-cTnT. Adjusted analyses showed a significant association between detectable CAC and ASCVD in women (HR 2.5; 95% CI: 1.61-3.88, p<0.0001) and men (HR 3.7; 95%: CI: 1.93-7.1, p<0.0001). Using CAC as a continuous variable (log transformed), adjusted analyses showed CAC as an independent predictor of ASCVD among participants with undetectable hs-cTnT (HR 1.26; 95% CI: 1.17-1.35, p<0.0001); similar findings were observed for men (HR 1.32; 95% CI: 1.17-1.49, p<0.0001) and women (HR 1.22; 95% CI: 1.11-1.35, p<0.0001).
Discussion
The present data suggests a potential novel approach to primary prevention risk stratification. They demonstrate that in a large, diverse, multiethnic population of individuals free of clinical cardiovascular disease, both undetectable hs-cTnT and a CAC score of zero have similar associations with long-term outcomes. Most importantly, they do not always detect the same individuals. The frequent discordance observed between CAC and hs-cTnT and the increased risk for ASCVD among those with discordant results indicate that their prognostic information is complementary, favoring their conjoint use for risk prediction. As compared to present risk stratification approaches that attempt to predict risk based on risk factors associated with ASCVD, the present approach uses established, objective, quantifiable measures of subclinical atherosclerosis and myocardial injury.
These results extend the value of hs-cTn from acute coronary syndromes, where such values are robustly prognostic for short and long term outcomes (13–16), to the primary prevention setting. Our data confirm prior investigations that report that hs-cTn assays provide a robust, continuous marker of risk in primary prevention cohorts (18). Our analyses demonstrate that individuals with undetectable hs-cTnT are less likely to develop ASCVD compared to those with detectable hs-cTnT, and the higher the concentration, the higher the risk for ASCVD.
The association with outcomes is clearly more precise when both tests are used together. Concordance analyses suggest that hs-cTnT and CAC results identify different patients with different risk profiles. As such, not only can they not be substituted for each other, but the significantly higher risk for ASCVD among those with discordant test results favors their conjoint use.
Conjoint testing allows for the identification of almost two-thirds of individuals as either being at low- or high-risk when both tests are concordant. Those with undetectable CAC/hs-cTnT have less than a 3% ten-year risk for ASCVD, as compared to those with both detectable CAC/hs-cTnT in whom the risk is almost 20%. Although a CAC of zero identifies patients unlikely to have a higher than intermediate risk (<7.5%), discordant CAC/hs-cTnT results, which were observed in 38% of participants, provide complementary prognostic information. They conjointly facilitate the detection of patients with a significantly greater risk for ASCVD when either test has a detectable result that provides additional prognostic information beyond what an isolated CAC score of zero or an undetectable hs-cTnT provides, and therefore permits more precise estimates of risk. Critically, among those with a CAC of zero, a detectable hs-cTnT concentration identifies a subset of individuals (28% of the cohort or 56% of those with a CAC of zero) at a significantly higher risk for ASCVD. For this latter group, particular differences in risk are observed in certain patient subsets such as blacks, for whom 10-year ASCVD risk was 6.3% when hs-cTnT was detectable, even with a CAC score of zero. Likewise, individuals with an undetectable hs-cTnT and a detectable CAC, a smaller subset of about 10% of participants, have 10-year ASCVD rates of almost 10%. Our data supports the recognition of hs-cTnT a “risk enhancer” when concentrations are detectable, and as a “negative risk factor” when concentrations are undetectable.
CAC scoring is an established primary prevention tool recommended across international guidelines. A CAC score of zero is recognized as a protective factor that allows downgrading risk with higher confidence than the 10-year pooled cohort equation that can either under- or over-estimate risk (10). Clinically, CAC scoring is advocated for individuals with intermediate (>=7.5% to <20%) or borderline (5% to <7.5%) 10-year ASCVD risk and in those reluctant/concerned about statin therapy (9–10). Recent studies indicate it is possible to measure CAC from chest CTs performed for other clinical indications (28–29). This may attenuate concerns related to cost and radiation (7,10), which have been impediments to the implementation of this technique. While we advocate the use of CAC and hs-cTnT together, when CAC scoring cannot be done, our data support the concept that measurement of hs-cTnT, a simple, low cost blood test, facilitates decisions about primary prevention of ASCVD.
Our study has multiple strengths. First, while studies have demonstrated the prognostic value of hs-cTnI/hs-cTnT assays (17,30–31) or CAC scoring (8,32–33) alone in individuals without cardiovascular disease, no data exist comparing their prognostic performance in a large cohort with long-term follow-up. Second, most investigations addressing the prognostic role of hs-cTnT/hs-cTnI assays in primary prevention have addressed mostly White and/or male populations, have shorter follow-up, and lack concomitant CAC scoring (17,19,34). Our data appear to apply across both sexes and the races/ethnicities tested. Third, given the fact that the intended clinical use of these strategies is as a decision aid to identify individuals who may or may not benefit from preventive therapies (7), the present study evaluates incident ASCVD, including stroke. Finally, prior primary prevention biomarker studies have often emphasized detectable hs-cTn concentrations as an enhancing risk marker (18), rather than undetectable hs-cTn concentrations as a negative risk factor.
Limitations exist. First, in MESA, CAC scores were reported to participants and their physicians, which might have influenced outcomes. Second, our findings are specific to the hs-cTnT assay and the specific analyzer used (27). Hs-cTn assays have varying analytical characteristics and performance and thus each hs-cTn assay needs to be individually validated for ASCVD risk prediction. Third, both hs-cTnT and CAC results can change on repeat measurements (35–37). Hs-cTnT results can be influenced by exercise, circadian rhythm, and biotin use (38–40). Anticipating an increasing use of hs-cTn measurements in the outpatient setting, best practice measurement guidelines will need to be followed (41). Data from MESA indicates that most individuals with undetectable CAC and hs-cTnT results continue to have undetectable results over the following few years, but results can change as time progresses (35–36). Recent MESA analyses indicate that for those with a CAC of zero, CAC develops in 11% at 2-years and 50% at 10-years and suggest the potential need for repeat scanning in 3-5 year time frame (35). Likewise, MESA analyses evaluating serial hs-cTnT at exam 1 (2000-2002) and exam 3 (2004-2005) indicate that most individuals with undetectable results remain undetectable (77%) on follow-up (36). Further data is needed to inform whether repeat measurements are warranted over time, as well as the prognostic implications of changes over time. The Atherosclerosis Risk in Communities Study (ARIC) reported that changes in hs-cTnT over time (6 years apart) are associated with a significant increase in the risk for coronary heart disease and death in those that shifted from a baseline hs-cTnT <5 ng/L to a more detectable value during follow-up (37). In a smaller MESA analysis, this was not shown to be the case (36). Last, it is difficult to know exactly where cutoffs for risk should be placed in the primary prevention setting, so this risk stratification approach requires additional validation and calibration.
Conclusion
In conclusion, in a large, diverse, multiethnic community-based cohort, our data demonstrate that individuals with undetectable hs-cTnT have a similar, low risk for long term incident ASCVD as those with a CAC score of zero. Most importantly, the increased risk among those with discordant results supports the conjoint use of CAC and hs-cTnT for primary prevention risk prediction.
Supplementary Material
Perspectives.
Competency in Medical Knowledge:
Similar to the absence of coronary artery calcification (CAC score of zero), an undetectable level of high-sensitivity cardiac troponin T (hs-cTn) indicates a low risk of atherosclerotic cardiovascular disease (ASCVD), whereas detectable levels are associated with elevated risk.
Translational Outlook:
Prospective studies are needed to evaluate the clinical utility and cost-effectiveness of measuring hs-cTn, alone or in combination with CAC to assess ASCVD risk in variously defined populations.
Acknowledgments
The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the U.S. Department of Health and Human Services.
The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.
Funding
This research was supported by contracts HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168 and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences (NCATS).
Disclosures
Dr. Sandoval has served on an advisory board/speaker for Abbott Diagnostics, and an advisory board for Roche Diagnostics; all without personal financial compensation.
Dr. Daniels has served on an Advisory Board for Quidel and Roche, and has served on clinical endpoints adjudication committees for Siemens and Abbott.
Dr. DeFilippis has grant support from the National Institutes of Health, AstraZeneca, Ionis, and has received consulting income from Radiometer.
Dr. Jaffe has consulted or is presently consulting for most of the major diagnostic companies, including Beckman, Abbott, Siemens, ET Healthcare, Roche, Quidel, Brava and Sphingotec. He also consults for Blade and Novartis.
Dr. deFilippi has grants/contracts to him through his institution from Abbott Diagnostics, FujiRebio, Ortho Diagnostics and Roche Diagnostics. He receives consulting income from Fuji Rebio, Ortho Diagnostics, Roche Diagnostics, and Siemens Healthineers. He receives royalty income from UpToDate.
Drs. Blaha, Bielinski, Larson, Michos, Szklo and Paul Decker, M.S. have nothing to disclose.
ABBREVIATIONS
- ASCVD
atherosclerotic cardiovascular disease
- CAC
coronary artery calcium
- cTn
cardiac troponin
- hs-cTn
high-sensitivity cardiac troponin
- LoD
limit of detection
- MESA
Multi-Ethnic Study of Atherosclerosis
- CT
computed tomography
- URL
upper-reference limit
- NHLBI
National Heart, Lung, and Blood Institute
- HR
hazard ratio
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Benjamin EJ, Muntner P, Alonso A et al. Heart disease and stroke statistics – 2019 update: a report from the American Heart Association. Circulation 2019; 139: e56–e528. [DOI] [PubMed] [Google Scholar]
- 2.Greenland P, Blaha MJ, Budoff MJ, Erbel R, Watson KE. Coronary artery calcium score and cardiovascular risk. J Am Coll Cardiol 2018; 72: 434–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Blaha MJ, Blumenthal RS, Budoff RS, Nasir K. Understanding the utility of zero coronary calcium as a prognostic test: a Bayesian approach. Circ Cardiovasc Qual Outcomes 2011; 4: 253–6. [DOI] [PubMed] [Google Scholar]
- 4.Budoff MJ, Young R, Burke G et al. Ten-year association of coronary artery calcium with atherosclerotic cardiovascular disease (ASCVD) events: the multi-ethnic study of atherosclerosis (MESA). Eur Heart J 2018; 39: 2401–2408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Blaha MJ, Cainzos-Achirica M, Greenland P et al. Role of coronary artery calcium score of zero and other negative risk markers for cardiovascular disease: The Multi-Ethnic Study of Atherosclerosis (MESA). Circulation 2016; 133: 849–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Blaha MJ, Blankstein R, Nasir K. Coronary artery calcium scores of zero and establishing the concept of negative risk factors. J Am Coll Cardiol 2019; 74: 12–14. [DOI] [PubMed] [Google Scholar]
- 7.Cardoso R, Nasir K, Blumenthal RS, Blaha MJ. Selective use of coronary artery calcium testing for shared decision making: guideline endorsed and ready for prime time. Ann Intern Med 2019; 170: 262–263. [DOI] [PubMed] [Google Scholar]
- 8.Detrano R, Guerci AD, Carr JJ et al. Coronary calcium as a predictor of coronary events in four racial or ethnic groups. N Engl J Med 2008; 358: 1336–45. [DOI] [PubMed] [Google Scholar]
- 9.Arnett DK, Blumenthal RS, Albert MA et al. 2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2019; 74: e177–e232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.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. J Am Coll Cardiol 2019; 73: e285–e350. [DOI] [PubMed] [Google Scholar]
- 11.Mach F, Baigent C, Catapano AL et al. 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J 2019; 00: 1–78. [DOI] [PubMed] [Google Scholar]
- 12.Apple FS, Sandoval Y, Jaffe AS, Ordonez-Llanos J; IFCC Task Force on Clinical Applications of Cardiac Bio-Markers. Cardiac troponin assays: guide to understanding analytical characteristics and their impact on clinical care. Clin Chem 2017; 63: 73–81. [DOI] [PubMed] [Google Scholar]
- 13.Chapman AR, Lee KK, McAllister DA et al. Association of high-sensitivity cardiac troponin I concentrations with cardiac outcomes in patients with suspected acute coronary syndrome. JAMA 2017; 318: 1913–1924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pickering JW, Than MP, Cullen L et al. Rapid-rule of acute myocardial infarction with a single high-sensitivity cardiac troponin T measurement below the limit of detection: a collaborative meta-analysis. Ann Intern Med 2017; 166: 715–724. [DOI] [PubMed] [Google Scholar]
- 15.Sandoval Y, Nowak R, deFilippi CR et al. Myocardial infarction risk stratification with a single measurement of high-sensitivity troponin I. J Am Coll Cardiol 2019; 74: 271–282. [DOI] [PubMed] [Google Scholar]
- 16.Than MP, Aldous SJ, Troughton RW et al. Detectable high-sensitivity cardiac troponin within the population reference interval conveys high 5-year cardiovascular risk: an observational study. Clin Chem 2018; 64: 1044–1053. [DOI] [PubMed] [Google Scholar]
- 17.Shah AS, Anand A, Sandoval Y et al. High-sensitivity cardiac troponin I at presentation in patients with suspected acute coronary syndrome: a cohort study. Lancet 2015; 386: 2481–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Willeit P, Welsh P, Evans JDW et al. High-sensitivity cardiac troponin concentration and risk of first-ever cardiovascular outcomes in 154,052 participants. J Am Coll Cardiol 2017; 70: 558–568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Blankenberg S, Salomaa V, Makarova N et al. Troponin I and cardiovascular risk prediction in the general population: the BiomarCaRE consortium. Eur Heart J 2016; 37: 2428–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.De Lemos JAD, 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Everett BM, Zeller T, Glynn RJ, Ridker PM, Blankenberg S. High-sensitivity cardiac troponin I and B-type natriuretic peptide as predictors of vascular events in primary prevention: impact of statin therapy. Circulation 2015; 131: 1851–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Januzzi JL Jr, Suchindran S, Coles A et al. High-sensitivity troponin I and coronary computed tomography in symptomatic outpatients with suspected CAD: insights from the PROMISE trial. JACC Cardiovasc Imaging 2019; 12: 1047–1055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bild DE, Bluemke DA, Burke GL et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol 2002; 156: 871–81. [DOI] [PubMed] [Google Scholar]
- 24.Carr JJ, Nelson JC, Wong ND et al. Calcified coronary artery plaque measurement with cardiac CT in population-based studies: standardized protocol of Multi-Ethnic Study of Atherosclerosis (MESA) and Coronary Artery Risk Development in Young Adults (CARDIA) study. Radiology 2005; 234: 35–43. [DOI] [PubMed] [Google Scholar]
- 25.Seliger SL, Hong SN, Christenson RH et al. High-sensitive cardiac troponin T as an early biochemical signature for clinical and subclinical heart failure: MESA (Multi-Ethnic Study of Atherosclerosis). Circulation 2017; 135: 1494–1505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Giannitsis E, Kurz K, Hallermayer K, Jarausch J, Jaffe AS, Katus HA. Analytical validation of a high-sensitivity cardiac troponin T assay. Clin Chem 2010; 56: 254–61. [DOI] [PubMed] [Google Scholar]
- 27.Sandoval Y, Jaffe AS. Using high-sensitivity cardiac troponin T for acute cardiac care. Am J Med 2017; 130: 1358–65. [DOI] [PubMed] [Google Scholar]
- 28.Hughes-Austin JM, Dominguez A 3rd, Allison MA et al. Relationship of coronary calcium on standard chest CT scans with mortality. JACC Cardiovasc Imaging 2016; 9: 152–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Azour L, Kadoch MA, Ward TJ, Eber CD, Jacobi AH. Estimation of cardiovascular risk on routine chest CT: ordinal coronary artery calcium scoring as an accurate predictor of Agatson score ranges. J Cardiovasc Comput Tomogr 2017; 11: 8–15. [DOI] [PubMed] [Google Scholar]
- 30.Parikh RH, Seliger SL, de Lemos J et al. Prognostic significance of high-sensitivity cardiac troponin T concentrations between the limit of blank and limit of detection in community-dwelling adults: a metaanalysis. Clin Chem 2015; 61: 1524–31. [DOI] [PubMed] [Google Scholar]
- 31.Ford I, Shah AS, Zhang R et al. High-sensitivity cardiac troponin, statin therapy, and risk of coronary heart disease. J Am Coll Cardiol 2016; 68: 2719–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sarwar A, Shaw LJ, Shapiro MD et al. Diagnostic and prognostic value of absence of coronary artery calcification. JACC Cardiovasc Imaging 2009; 2: 675–88. [DOI] [PubMed] [Google Scholar]
- 33.Blaha M, Budoff MJ, Shaw LJ et al. Absence of coronary artery calcification and all-cause mortality. JACC Cardiovasc Imaging 2009; 2: 692–700. [DOI] [PubMed] [Google Scholar]
- 34.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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Dzaye O, Dardari ZA, Cainzos-Achirica M et al. Incidence of new coronary calcification: time to conversion from CAC = 0. J Am Coll Cardiol 2020; 75: 1610–1613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Qian F, Allen N, McCabe M et al. Changes in high-sensitivity troponin and the risk of incident cardiovascular events: The Multi-Ethnic Study of Atherosclerosis (MESA). J Am Coll Cardiol 2020; 75 (11 Supplement 1) 1878 (abstr)32327097 [Google Scholar]
- 37.McEvoy JW, Chen Y, Ndumele CE et al. Six-year change in high-sensitivity cardiac troponin T and risk of subsequent coronary heart disease, heart failure, and death. JAMA Cardiol 2016; 1: 519–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Saenger AK, Jaffe AS, Body R et al. Cardiac troponin and natriuretic peptide analytical interferences from hemolysis and biotin: educational aids from the IFCC Committee on Cardiac Biomarkers (IFCC C-CB). Clin Chem Lab Med 2019; 57: 633–640 [DOI] [PubMed] [Google Scholar]
- 39.Klinkenberg LJ, van Dijk JW, Tan FE, van Loon LJ, van Dieijen-Visser MP, Meex SJ. Circulating cardiac troponin T exhibits a diurnal rhythm. J Am Coll Cardiol 2014; 63: 1788–95. [DOI] [PubMed] [Google Scholar]
- 40.Lee G, Twerenbold R, Tanglay Y et al. Clinical benefit of high sensitivity cardiac troponin I in the detection of exercise-induced myocardial ischemia. Am Heart J 2016; 173: 8–17. [DOI] [PubMed] [Google Scholar]
- 41.Jaffe AS, Jaffe HA. Use of high-sensitivity cardiac troponin in patients with chronic comorbidities: moving from theory to practice. J Am Coll Cardiol 2018; 72: 1138–1140 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.