Abstract
Aims
The relationship between dyspnoea, coronary artery disease (CAD), and major cardiovascular events (MACE) is poorly understood. This study evaluated (i) the association of dyspnoea with the severity of anatomical CAD by coronary computed tomography angiography (CCTA) and (ii) to which extent CAD explains MACE in patients with dyspnoea.
Methods and results
From the international COronary CT Angiography EvaluatioN for Clinical Outcomes: An InteRnational Multicenter (CONFIRM) registry, 4425 patients (750 with dyspnoea) with suspected but without known CAD were included and prospectively followed for ≥5 years. First, the association of dyspnoea with CAD severity was assessed using logistic regression analysis. Second, the prognostic value of dyspnoea for MACE (myocardial infarction and death), and specifically, the interaction between dyspnoea and CAD severity was investigated using Cox proportional-hazard analysis. Mean patient age was 60.3 ± 11.9 years, 63% of patients were male and 592 MACE events occurred during a median follow-up duration of 5.4 (IQR 5.1–6.0) years. On uni- and multivariable analysis (adjusting for age, sex, body mass index, chest pain typicality, and risk factors), dyspnoea was associated with two- and three-vessel/left main (LM) obstructive CAD. The presence of dyspnoea increased the risk for MACE [hazard ratio (HR) 1.57, 95% confidence interval (CI): 1.29–1.90], which was modified after adjusting for clinical predictors and CAD severity (HR 1.26, 95% CI: 1.02–1.55). Conversely, when stratified by CAD severity, dyspnoea did not provide incremental prognostic value in one-, two-, or three-vessel/LM obstructive CAD, but dyspnoea did provide incremental prognostic value in non-obstructive CAD.
Conclusion
In patients with suspected CAD, dyspnoea was independently associated with severe obstructive CAD on CCTA. The severity of obstructive CAD explained the elevated MACE rates in patients presenting with dyspnoea, but in patients with non-obstructive CAD, dyspnoea portended additional risk.
Keywords: coronary artery disease, coronary computed tomography angiography, dyspnoea, prognosis
Introduction
Patients presenting with dyspnoea comprise a diagnostic challenge, since symptoms can be caused by multiple cardiac and non-cardiac diseases, including coronary artery disease (CAD), heart failure, pulmonary disease (e.g. chronic obstructive pulmonary disease or lung fibrosis), neuro-muscular disease, or psychogenic disorders (e.g. hyperventilation syndrome). An extensive clinical evaluation may clarify the diagnosis, but often the symptoms remain unexplained and additional diagnostic testing is needed .
Dyspnoea can be secondary to CAD (also referred to as ‘angina-equivalent’) that causes myocardial ischaemia, ischaemic left ventricular (LV) dysfunction, and increased LV filling pressures.1 Some studies that evaluated patients with suspected CAD who presented with dyspnoea reported a significant association with myocardial ischaemia2,3 while other studies did not.4–6 However, all studies consistently showed that dyspnoea has a significant impact on prognosis.2–6 To which extent CAD (or its consequence, myocardial ischaemia) contributes to the association of dyspnoea and impaired outcome is therefore uncertain .
The presence and extent of myocardial ischaemia have traditionally been assessed with stress testing (stress echocardiography,2,6 single-photon emission computed tomography,3,4 or exercise stress testing5), which may underestimate CAD in the presence of balanced ischaemia.7 Coronary computed tomography angiography (CCTA) is more sensitive for the detection of CAD because it directly visualizes coronary atherosclerosis (as compared with an indirect assessment of CAD with stress testing).8 The COronary CT Angiography EvaluatioN For Clinical Outcomes: An InteRnational Multicenter (CONFIRM) is the largest registry including patients undergoing CCTA for suspected CAD and has systematically collected prospective, long-term follow-up data for major cardiovascular events (MACE).9,10 Accordingly, data from the CONFIRM registry were used to (i) evaluate the association of dyspnoea with the severity of anatomical CAD and vice versa and (ii) to which extent anatomical CAD severity contributes to the elevated rates of MACE among patients presenting with dyspnoea.
Methods
Patients
The CONFIRM registry is a dynamic, international, multicentre, observational cohort that prospectively collects clinical, procedural, and follow-up data from patients who underwent ≥64 slice CCTA for clinically suspected CAD, as previously described.9 The current analysis concerned patients from the long-term follow-up cohort of the CONFIRM registry, which included patients with ≥5 years follow-up for major events from 17 participating sites from nine countries between 2002 and 2009. We included patients without known CAD [defined as previous myocardial infarction (MI), coronary artery bypass grafting, or coronary revascularization] before CCTA acquisition. The institutional review board approval was obtained for all participating sites, with either informed consent or waiver of informed consent.11
Clinical data collection
Standardized data collection was performed at the participating centres. Each patient was evaluated by a physician or nurse before CCTA and demographical and clinical information was systematically obtained.9 Cardiovascular risk factors included diabetes, hypertension, hypercholesterolaemia, familial history of CAD, and current smoking as defined previously.12–14 Suspected cardiac symptoms, such as CCTA indication, were categorized as non-anginal, atypical and typical chest pain, and dyspnoea. Dyspnoea was defined as whether patients experienced shortness of breath and was binary classified. Chest pain symptoms could be present together with dyspnoea.
CCTA image acquisition and interpretation
CCTA acquisition protocols at each site were in adherence with the Society of Cardiovascular Computed Tomography guidelines.15 Specific details regarding the acquisition were previously described.9 Level III-trained experts uniformly interpreted the CT images using a 16-segment coronary artery tree model. For each coronary artery segment, the presence of atherosclerosis was reported with corresponding stenosis severity. The stenosis severity of coronary artery atherosclerosis was categorized as normal (0% stenosis), non-obstructive (1–49% stenosis), or obstructive CAD (≥50% stenosis) by visual assessment. Subsequently, the severity of CAD was categorized as no CAD, non-obstructive CAD, one coronary artery with ≥50% stenosis in any coronary segment, two coronary arteries with ≥50% stenosis, and three coronary arteries and/or left main (LM) with ≥50% stenosis.12 LV ejection fraction was calculated from CCTA as previously described.16
Outcomes
First, we evaluated the association of dyspnoea with CAD severity. Second, the relation of dyspnoea (and the influence of CAD) with MACE was assessed, defined as a composite endpoint of all-cause death and non-fatal MI. Each local institution systematically performed patient follow-up by a dedicated physician or nurse.9 The Social Security Index was reviewed for assessment of mortality within the United States or determined through email or telephone contact with the patients, family, or physician or review of medical records for the other countries. MI events were collected through a combination of direct interviewing of patients using scripted interview with confirmation of the event by screening patient’s medical files, in accordance with the universal definition of MI.9,17
Statistical analysis
Continuous variables were described as mean ± standard deviation (SD); categorical variables as frequencies with percentages. Continuous data were compared using the Student’s t-test and one-way analysis of variance (ANOVA) analysis for multiple groups. Categorical variables were compared with the χ2 test. The association of dyspnoea with CAD severity subgroups was evaluated using logistic regression analysis; odds ratios with 95% confidence intervals (CIs) were derived. Demographical and clinical variables, which were significantly associated with increasing CAD severity, were entered in multivariable models (multinominal logistic regression) to assess the independent association of dyspnoea with CAD. The Kaplan–Meier method was utilized to depict 5-year MACE-free survival rates for the presence vs. absence of dyspnoea and the CAD severity subgroups (normal, non-obstructive CAD, one-, two-, and three-vessel/LM CAD); comparisons were made with the log-rank test. Uni- and multivariable Cox-proportional hazard analyses were performed, and hazard ratios (HRs) were derived to assess the prognostic impact of dyspnoea on MACE, with significant univariate predictors (P < 0.10) entered in multivariable models. Adjusted MACE-free survival curves were provided for the presence or absence of dyspnoea among patients with no, non-obstructive, and obstructive CAD. Furthermore, whether the prognostic value of dyspnoea was modified by the severity of CAD was tested with an interaction term for dyspnoea and CAD severity, which was included in a multivariable model together with the two main effects. A two-sided P-value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS statistics (version 24, IBM, Armonk, NY, USA).
Results
Patients
Of the 4425 patients included in the current analysis, 750 (17%) had dyspnoea as presenting symptom. Patients with dyspnoea were significantly older (62.1 ± 12.3 vs. 59.9 ± 11.8 years, P < 0.001), less often male (56% vs. 64%, P < 0.001), had a higher body mass index (BMI 28.1 ± 6.0 vs. 26.9 ± 4.5 kg/m2), more frequently presented with concomitant typical angina (19.8% vs. 14.9%) and more often reported no chest pain (42.7% vs. 32.6%; overall P < 0.001) vs. absence of dyspnoea (Table 1). Patients with dyspnoea had a higher prevalence of two- and three-vessel/LM obstructive CAD than patients without dyspnoea: 12.0% vs. 9.0% and 10.4% vs. 7.4%, respectively, overall P < 0.001. Of interest, patients presenting with dyspnoea did not smoke more frequently (18.1% vs. 18.6%, P = 0.747).
Table 1.
Patient characteristics
| Dyspnoea (n = 750) | No dyspnoea (n = 3675) | P-value | |
|---|---|---|---|
| Age (years) | 62.1 ± 12.3 | 59.9 ± 11.8 | <0.001 |
| Male (gender) | 416 (56) | 2371 (64) | <0.001 |
| BMI (kg/m2) | 28.1 ± 6.0 | 26.9 ± 4.5 | <0.001 |
| Chest pain symptoms | <0.001 | ||
| No chest pain | 229 (32.6) | 1557 (42.7) | |
| Non-anginal | 108 (15.4) | 255 (7.0) | |
| Atypical | 226 (32.2) | 1287 (35.3) | |
| Typical | 139 (19.8) | 544 (14.9) | |
| Cardiovascular risk factors | |||
| Diabetes | 190 (25.5) | 572 (15.6) | <0.001 |
| Hypertension | 487 (65.4) | 1904 (52.0) | <0.001 |
| Hypercholesterolaemia | 422 (56.6) | 1976 (53.9) | 0.190 |
| Family history for CAD | 193 (26.4) | 1111 (30.6) | 0.023 |
| Current smoker | 134 (18.1) | 679 (18.6) | 0.747 |
| Medication use | |||
| Aspirin | 265 (43.3) | 954 (28.7) | <0.001 |
| Beta blocker | 210 (34.4) | 933 (28.1) | 0.002 |
| ACE-I | 125 (20.5) | 653 (19.7) | 0.655 |
| Statin | 270 (44.1) | 1207 (36.2) | <0.001 |
| Coronary CTA findings | <0.001 | ||
| Non-obstructive CAD | 214 (28.5) | 1055 (28.7) | |
| One-vessel obstructive CAD | 147 (19.6) | 700 (19.0) | |
| Two-vessel obstructive CAD | 90 (12.0) | 330 (9.0) | |
| Three-vessel/left main obstructive CAD | 78 (10.4) | 272 (7.4) |
Values are expressed as mean ± standard deviation or counts (%).
ACE-I, angiotensin converting enzyme-inhibiter; BMI, body mass index; CAD, coronary artery disease.
Association of dyspnoea with CAD
The clinical characteristics associated with CAD are shown in Table 2. Dyspnoea was more prevalent with increasing severity of CAD; 16.9% in patients with non-obstructive CAD vs. 22.3% in patients with three-vessel/LM obstructive CAD (P < 0.001). Other clinical variables significantly associated with increasing CAD severity were age, male gender, typical chest pain, diabetes, hypertension, hypercholesterolaemia, and current smoking (P < 0.001 for all). In multivariable analysis adjusting for age, sex, chest pain typicality, and cardiovascular risk factors, two- and three-vessel/LM obstructive CAD were independently associated with dyspnoea: odds ratio (OR) 1.43 (95% CI: 1.04–1.98) and 1.56 (95% CI: 1.11–2.21), respectively (Figure 1). With further adjustment for statin, aspirin, and beta blocker use, only three-vessel/LM obstructive CAD remained significant (OR 1.48, 95% CI: 1.02–2.15). No significant association was observed between dyspnoea and non-obstructive, or one-vessel obstructive CAD.
Table 2.
Associations between clinical characteristics and CAD
| Normal (n = 1539) | CAD ≤50% (n = 1269) | 1-VD ≥50% (n = 847) | 2-VD ≥50% (n = 420) | 3-VD/left main ≥50% (n = 350) | P-value | |
|---|---|---|---|---|---|---|
| Age (years) | 54.5 ± 12.6 | 61.9 ± 10.4 | 63.5 ± 10.0 | 64.9 ± 10.0 | 66.8 ± 9.6 | <0.001 |
| Male (gender) | 791 (51.4) | 811 (63.9) | 600 (70.8) | 101 (76.0) | 263 (75.1) | <0.001 |
| BMI (kg/m2) | 27.0 ± 4.9 | 27.6 ± 5.2 | 26.9 ± 4.7 | 27.3 ± 4.4 | 26.7 ± 3.7 | 0.003 |
| Cardiac symptoms | <0.001 | |||||
| No chest pain | 633 (41.8) | 589 (47.2) | 305 (36.8) | 155 (37.5) | 104 (30.5) | |
| Non-anginal | 147 (9.7) | 97 (7.8) | 59 (7.1) | 35 (8.5) | 25 (7.3) | |
| Atypical | 570 (37.6) | 451 (36.1) | 280 (33.8) | 111 (26.9) | 101 (29.6) | |
| Typical | 164 (10.8) | 111 (8.9) | 185 (22.3) | 112 (27.1) | 111 (32.6) | |
| Dyspnoea | 221 (14.4) | 214 (16.9) | 147 (17.4) | 90 (21.4) | 78 (22.3) | <0.001 |
| Cardiovascular risk factors | ||||||
| Diabetes | 193 (12.6) | 174 (13.7) | 174 (20.6) | 105 (25.3) | 116 (33.3) | <0.001 |
| Hypertension | 662 (43.1) | 703 (55.6) | 504 (59.8) | 277 (66.7) | 245 (70.4) | <0.001 |
| Hypercholesterolaemia | 653 (42.5) | 747 (59.0) | 527 (62.4) | 252 (60.4) | 219 (63.3) | <0.001 |
| Family history for CAD | 463 (30.5) | 375 (29.6) | 243 (29.2) | 117 (28.4) | 106 (31.1) | 0.878 |
| Current smoker | 248 (16.3) | 214 (16.9) | 177 (21.0) | 89 (21.4) | 85 (24.6) | <0.001 |
Values are expressed as mean ± standard deviation or counts (%).
BMI, body mass index; CAD, coronary artery disease; VD, vessel disease.
Figure 1.
Association of dyspnoea with severity of CAD. On uni- and multivariable analysis, dyspnoea was significantly associated with two- and three-vessel/left main obstructive CAD. CAD, coronary artery disease.
Prognostic value of dyspnoea for MACE
During a median follow-up duration of 5.4 years (25–75% interquartile range 5.1–6.0 years), a total of 592 MACE events occurred. Dyspnoea was associated with a 57% increased risk for MACE (HR 1.57, 95% CI: 1.29–1.90; P < 0.001; Table 3). Other important prognostic factors on univariable analysis were age, sex, typical chest pain, diabetes, hypertension, currently smoking and CAD severity. After adjusting for these variables, dyspnoea remained significantly associated with MACE, but the effect magnitude was modified (HR 1.26, 95% CI 1.02–1.55; P = 0.029).
Table 3.
Cox proportional hazard analysis for MACE
| Univariable HR (95% CI) | P-value | Multivariable HR (95% CI) | P-value | |
|---|---|---|---|---|
| Age (years) | 1.04 (1.04–1.05) | <0.001 | 1.03 (1.02–1.04) | <0.001 |
| Male (gender) | 0.87 (0.74–1.03) | 0.102 | 0.77 (0.64–0.93) | 0.005 |
| BMI (kg/m2) | 1.01 (1.00–1.03) | 0.106 | – | |
| Cardiac symptoms | 0.001 | 0.681 | ||
| No chest pain | Ref | Ref | ||
| Non-anginal | 1.41 (1.05–1.89) | 1.17 (0.87–1.57) | ||
| Atypical | 1.15 (0.95–1.41) | 1.11 (0.90–1.35) | ||
| Typical | 1.58 (1.25–1.29) | 1.05 (0.82–1.33) | ||
| Dyspnoea | 1.57 (1.29–1.90) | <0.001 | 1.26 (1.02–1.55) | 0.029 |
| Cardiovascular risk factors | ||||
| Diabetes | 1.75 (1.44–2.12) | <0.001 | 1.38 (1.13–1.68) | 0.001 |
| Hypertension | 1.75 (1.46–2.01) | <0.001 | 1.35 (1.12–1.63) | 0.002 |
| Hypercholesterolaemia | 0.87 (0.74–1.03) | 0.115 | – | |
| Family history for CAD | 0.95 (0.79–1.15) | 0.607 | – | |
| Current smoker | 1.44 (1.19–1.75) | <0.001 | 1.42 (1.16–1.74) | 0.001 |
| Coronary CTA findings | ||||
| Normal | Ref | Ref | Ref | |
| Non-obstructive CAD | 2.64 (2.01–3.46) | <0.001 | 2.21 (1.66–2.94) | <0.001 |
| One-vessel obstructive CAD | 4.69 (3.56–6.15) | <0.001 | 3.66 (2.73–4.92) | <0.001 |
| Two-vessel obstructive CAD | 5.39 (3.97–7.31) | <0.001 | 4.15 (2.98–5.78) | <0.001 |
| Three-vessel/left main obstructive CAD | 7.75 (5.74–10.4) | <0.001 | 5.28 (3.79–7.37) | <0.001 |
BMI, body mass index; CAD, coronary artery disease; MACE, major adverse cardiac events; Ref, reference category.
Interaction between dyspnoea and CAD severity
Among patients with one-, two-, or three-vessel/LM obstructive CAD, dyspnoea did not portend independent prognostic value (Table 4). In patients with non-obstructive CAD, dyspnoea was independently associated with elevated MACE rates (HR 1.59, 95 CI: 1.09–2.32; P = 0.017), after adjusting for age, sex, chest pain typicality, diabetes, hypertension, and current smoking. When separating the primary outcome, dyspnoea was only significantly associated with all-cause mortality in patients with non-obstructive CAD (HR 1.83, 95% CI: 1.14–2.93; P = 0.012), but not with MI (HR 1.51, 95% CI: 0.85–2.68; P = 0.161). When excluding patients with typical angina, results were similar and in non-obstructive CAD dyspnoea was associated with MACE (Appendix TableA1).
Table 4.
Prognostic value of dyspnoea according to CAD severity
| Adjusted HR for MACE (95% CI)a |
||||||
|---|---|---|---|---|---|---|
| Normal (n = 1539) events=77 | CAD ≤50% (n = 1269) events=161 | 1-VD ≥50% (n = 847) events=164 | 2-VD ≥50% (n = 420) events=90 | 3-VD/left main ≥50% (n = 350) events=100 | P-interactionb | |
| Dyspnoea | 1.73 (0.99–3.02) | 1.59 (1.09–2.32) | 1.02 (0.65–1.59) | 1.38 (0.83–2.31) | 0.77 (0.46–1.31) | 0.031 |
| P = 0.053 | P = 0.017 | P = 0.994 | P = 0.219 | P = 0.341 | ||
CAD, coronary artery disease; HR, hazard ratio; MACE, major adverse cardiac events; VD, vessel disease.
Adjusted for age, sex, chest pain typicality, diabetes, hypertension, and current smoking.
P-value for interaction term between CAD severity and dyspnoea when entered in the multivariable model with: age, sex, chest pain typicality, diabetes, hypertension, and current smoking, CAD severity, and dyspnoea.
The Kaplan–Meier curves showed significant differences in 5-year cumulative event-free survival for the CAD severity subgroups (95.6% for no CAD vs. 70.8% for three-vessel/LM obstructive CAD, P < 0.001) and for the absence vs. presence of dyspnoea (88.0% vs. 82.5%, P < 0.001; Figure 2). Adjusted MACE-free survival analyses visualize the lowest event rates for patients without CAD, and the worst outcomes among patients with obstructive CAD, regardless of dyspnoea (Figure 3). Among patients with non-obstructive CAD, the presence of dyspnoea was significantly associated with increased MACE risk (P = 0.017). Exploratory analyses demonstrated that in non-obstructive CAD, dyspnoeic patients were more likely to be female, have diabetes, hypertension, had an elevated BMI, but did not have a lower LVEF (59.5 ± 18.7% vs. 60.7 ± 11.7%, P = 0.539; Appendix TableA2).
Figure 2.
Kaplan–Meier MACE-free survival curves for patients according to CAD severity and dyspnoea. Five-year cumulative event-free survival curves show the graded increase of events with increasing CAD severity and higher MACE rates for dyspnoea vs. absence of dyspnoea (log-rank P<0.001). CAD, coronary artery disease; MACE, major adverse cardiac events.
Figure 3.
Interaction between dyspnoea and CAD severity on 5-year survival. Five-year cumulative event-free survival curves for the presence of dyspnoea among patients without CAD, with non-obstructive and obstructive CAD after adjusting for age, sex, chest pain typicality, diabetes, hypertension, and current smoking. Dyspnoea was associated with increased MACE risk in non-obstructive CAD only (P=0.017) and showed a trend in normal CCTA (P=0.053). The curves were overlapping in patients with obstructive CAD. CAD, coronary artery disease; MACE, major adverse cardiac events.
Discussion
Among patients without known CAD referred for CCTA, dyspnoea is independently associated with severely obstructive CAD. Secondly, dyspnoea was predictive for MACE, but this effect was significantly modified after adjusting for clinical variables and anatomical CAD severity. In patients with obstructive CAD, all-cause death and MI were determined by CAD severity and the presence of dyspnoea did not increase long-term risk for MACE. However, in patients with non-obstructive CAD, excess risk for MACE (especially all-cause mortality) existed for dyspnoea.
Dyspnoea and CAD
Dyspnoea as presenting symptom can be a sign of symptomatic CAD (angina equivalent); therefore, these patients frequently undergo diagnostic testing to detect or exclude CAD. Coronary atherosclerosis may lead to myocardial ischaemia which provokes diastolic dysfunction that subsequently leads to dyspnoea.1 Ilia et al.18 elegantly investigated the relationship between shortness of breath, LV filling pressures, and the severity of CAD assessed by invasive coronary angiography and observed that increasing severity of CAD was correlated with worsening grades of shortness of breath and increasing LV end-diastolic pressures. However, relatively little is known about the prevalence and severity of CAD among patients presenting with dyspnoea and unexpected results have been reported in comparison to patients presenting with chest pain. A meta-analysis including 24 491 patients with chest pain and 5753 patients with dyspnoea who underwent stress testing, observed myocardial ischaemia in 37.5% of patients with dyspnoea, which was similar to patients presenting with chest pain (36.5%).19 Among 17 991 patients with suspected CAD, Abidov et al.4 reported the prevalence of ischaemia in patients with dyspnoea vs. absence of symptoms, and observed only a minor increased incidence of ischaemia (19.4% vs. 16.7%) and no difference in percentage of ischaemia myocardium (2.4% vs. 2.2%). In contrast, the incidence of myocardial ischaemia and the percentage of ischaemia myocardium in the left ventricle were much higher in patients presenting with typical angina (30.1% and 4.4%). These data did not convincingly support that patients presenting with dyspnoea can have a large burden of myocardial ischaemia. Nor did they identify myocardial ischaemia as a mediator in the relationship between dyspnoea and MACE.4 Associations between dyspnoea and myocardial ischaemia have been largely investigated with relative perfusion imaging techniques (single-photon emission computed tomography), which may have reduced accuracy in the presence of balanced ischaemia, provoked by severe, multivessel disease.7,20 Accordingly, the prevalence of CAD in patients presenting with dyspnoea may have even been underestimated.
This study examined this specific question using CCTA, which is highly sensitive test to quantify the extent and severity of coronary atherosclerosis. We observed an independent association of dyspnoea with two- and three-vessel/LM obstructive CAD only, which supports this hypothesis of anatomically severe CAD, leading to myocardial ischaemia, diastolic dysfunction, and dyspnoea. These findings are in line with Nakanishi et al.,21 who demonstrated among 1443 patients without known CAD that dyspnoea is associated with ≥70% stenosis and proximal coronary plaque on CCTA. Given the strong prognostic value of CAD severity for MACE, obstructive CAD will likely interact in the association between dyspnoea and elevated rates of MACE.
Dyspnoea and prognosis
Studies have consistently shown that dyspnoea is associated with worse outcome among patients with suspected CAD.2–6 The previously mentioned meta-analysis19 reported a 2.57 times increased mortality risk for dyspnoea vs. chest pain among 30 244 patients. Also, the largest individual study which examined 17 991 patients using single-photon emission computed tomography demonstrated a 1.82 (95% CI: 1.30–2.55) times increased risk for cardiac death of dyspnoea vs. absence of cardiac symptoms. After extensive multivariable modelling, including the percentage of myocardium with fixed defects and ischaemia, the effect was only little modified to 1.76 (95% CI: 1.25–2.47), which was consistent with the only mildly increased burden of ischaemia in patients with dyspnoea.
We observed that dyspnoea was associated with severely obstructive CAD, and among patients with obstructive CAD, all-cause death, and MI were determined by CAD severity, without additional prognostic effect of dyspnoea. Among patients with non-obstructive CAD, dyspnoea was significantly associated with increased MACE risk. Given the absence of association between dyspnoea with the presence vs. absence of non-obstructive CAD, the elevated MACE rates among this subgroup of patients are likely caused by other factors. Extra-cardiac deaths such as pulmonary, psychiatric, or vascular diseases may have occurred. However, CONFIRM included patients undergoing CCTA because of suspected cardiac pathology and heart failure, or microvascular coronary dysfunction may explain MACE in non-obstructive CAD. Recent work from Taqueti et al.22 demonstrated that in symptomatic patients with normal LVEF and without overt CAD, the prevalence of coronary microvascular dysfunction [defined as coronary flow reserve <2 on positron emission tomography imaging (PET)] exceeded 50% and it was independently associated with diastolic dysfunction. Also, this study population consisted of predominantly women (65%), and the prevalence of diabetes and hypertension was high, similar to the patients with non-obstructive CAD in this study. As demonstrated by the same group, among 329 symptomatic patients referred for invasive coronary angiography after PET imaging, coronary flow reserve was associated with MACE, independently from the CAD prognostic index derived from angiography.23 Importantly, the correlation between decreasing flow reserve and increasing CAD severity was only modest.
Clinically, our findings suggest that patients presenting with dyspnoea and obstructive CAD should be treated according to the severity of atherosclerosis with medical therapy with or without revascularization according to guideline directed care for symptomatic CAD. It is apparent that dyspnoea in these patients is associated with hemodynamically significant obstructive stenosis. This supports the guidelines that have included dyspnoea as increasing factor for pre-test probability for CAD.24 However, when obstructive stenosis is excluded, the presence of dyspnoea increases the risk over CAD severity, which may be related to extra-cardiac causes or adverse coronary physiologically. Myocardial blood flow imaging in these patients may improve risk stratification.
Limitations
The current evaluation is an observational cohort study with all its inherent limitations including selection bias and unobserved confounders. Data regarding functional testing of CAD were unavailable, and therefore, whether the relationship between three-vessel/LM obstructive CAD and dyspnoea was mediated by myocardial ischaemia leading to subsequent LV dysfunction remains unresolved. Specific causes of death (coronary or cardiac) were not available for the current analyses. No further detailed information regarding the duration and severity of dyspnoea was available. The large group of asymptomatic individuals may not fully represent an asymptomatic cohort because of their clinical indication for CCTA, which poses a limitation to interpretability of the results and potentially reduces the strength of the observed associations. Also, it would have been preferred if several grades of dyspnoea severity were available. Possibly, more severe symptoms relate more strongly with obstructive CAD. Finally, LVEF and heart failure status were not available for all patients and it cannot be excluded that some patients presented with symptomatic heart failure.
Conclusion
Among patients without known CAD referred for CCTA, dyspnoea was independently associated with two-vessel or three-vessel/LM obstructive CAD. In patients with obstructive CAD, all-cause death and MI were determined by CAD severity and the presence of dyspnoea did not increase long-term risk for MACE. In patients without obstructive CAD, dyspnoea associated with MACE independently from the presence or absence of non-obstructive CAD. Additional cardiac or non-cardiac investigations may improve risk stratifications of this subgroup.
Funding
The research reported in this publication was funded, in part, by the National Institute of Health (Bethesda, MD, USA) under award number R01 HL115150. This research was also supported, in part, by the Dalio Institute of Cardiovascular Imaging (New York, NY, USA) and the Michael Wolk Foundation (New York, NY, USA) .
Conflict of interest: J.A.L. is a consultant to and has stock options in Circle CVI and HeartFlow. J.K.M. is employed by and owns equity interest in Cleerly, Inc. and has served on the Advisory Board at Arineta. All other authors have no conflict of interest to declare .
Appendix
Table A1.
Prognostic value of dyspnoea according to CAD severity excluding patients with typical angina
| Adjusted HR for MACE (95% CI)a | |||||
|---|---|---|---|---|---|
| Normal | CAD ≤50% | 1-VD ≥50% | 2-VD ≥50% | 3-VD/left main ≥50% | |
| Dyspnoea | 1.31 (0.65–2.65) P = 0.450 | 1.67 (1.03–2.71) P = 0.038 | 0.71 (0.33–1.48) P = 0.358 | 1.40 (0.66–2.98) P = 0.384 | 0.86 (0.40–1.85) P = 0.692 |
CAD, coronary artery disease; HR, hazard ratio; MACE, major adverse cardiac events; VD, vessel disease.
Adjusted for age, sex, chest pain typicality, diabetes, hypertension, and current smoking.
Table A2.
Sub-analysis in patients with non-obstructive CAD
| Dyspnoea (N = 214) | No dyspnoea (N = 1055) | P-value | |
|---|---|---|---|
| Women | 103 (48.1) | 355 (33.6) | <0.001 |
| Body mass index | 29.2 ± 7.4 | 27.3 ± 4.6 | 0.001 |
| Diabetes | 48 (22.5) | 126 (12.0) | <0.001 |
| Hypertension | 142 (67.0) | 561 (53.3) | <0.001 |
| LV ejection fraction (%) | 59.5 ± 18.7a | 60.7 ± 11.7b | 0.539 |
CAD, coronary artery disease; LV, left ventricular.
Available in 109 (50.9%) patients.
Available in 321 (30.4%) patients.
References
- 1. Pepine CJ, Wiener L.. Relationship of anginal symptoms to lung mechanics during myocardial ischemia. Circulation 1972;46:863–9. [DOI] [PubMed] [Google Scholar]
- 2. Bergeron S, Ommen SR, Bailey KR, Oh JK, McCully RB, Pellikka PA.. Exercise echocardiographic findings and outcome of patients referred for evaluation of dyspnea. J Am Coll Cardiol 2004;43:2242–6. [DOI] [PubMed] [Google Scholar]
- 3. Balaravi B, Miller TD, Hodge DO, Gibbons RJ.. The value of stress single photon emission computed tomography in patients without known coronary artery disease presenting with dyspnea. Am Heart J 2006;152:551–7. [DOI] [PubMed] [Google Scholar]
- 4. Abidov A, Rozanski A, Hachamovitch R, Hayes SW, Aboul-Enein F, Cohen I. et al. Prognostic significance of dyspnea in patients referred for cardiac stress testing. N Engl J Med 2005;353:1889–98. [DOI] [PubMed] [Google Scholar]
- 5. Christopher Jones R, Pothier CE, Blackstone EH, Lauer MS.. Prognostic importance of presenting symptoms in patients undergoing exercise testing for evaluation of known or suspected coronary disease. Am J Med 2004;117:380–9. [DOI] [PubMed] [Google Scholar]
- 6. Bernheim AM, Kittipovanonth M, Scott CG, McCully RB, Tsang TS, Pellikka PA.. Relation of dyspnea in patients unable to perform exercise stress testing to outcome and myocardial ischemia. Am J Cardiol 2009;104:265–9. [DOI] [PubMed] [Google Scholar]
- 7. Melikian N, De Bondt P, Tonino P, De Winter O, Wyffels E, Bartunek J. et al. Fractional flow reserve and myocardial perfusion imaging in patients with angiographic multivessel coronary artery disease. JACC Cardiovasc Interv 2010;3:307–14. [DOI] [PubMed] [Google Scholar]
- 8. Budoff MJ, Dowe D, Jollis JG, Gitter M, Sutherland J, Halamert E. et al. Diagnostic performance of 64-multidetector row coronary computed tomographic angiography for evaluation of coronary artery stenosis in individuals without known coronary artery disease: results from the prospective multicenter ACCURACY (Assessment by Coronary Computed Tomographic Angiography of Individuals Undergoing Invasive Coronary Angiography) trial. J Am Coll Cardiol 2008;52:1724–32. [DOI] [PubMed] [Google Scholar]
- 9. Min JK, Dunning A, Lin FY, Achenbach S, Al-Mallah MH, Berman DS. et al. Rationale and design of the CONFIRM (COronary CT Angiography EvaluatioN For Clinical Outcomes: an InteRnational Multicenter) Registry. J Cardiovasc Comput Tomogr 2011;5:84–92. [DOI] [PubMed] [Google Scholar]
- 10. Schulman-Marcus J, Hartaigh BO, Gransar H, Lin F, Valenti V, Cho I. et al. Sex-specific associations between coronary artery plaque extent and risk of major adverse cardiovascular events: the CONFIRM long-term registry. JACC Cardiovasc Imag 2016;9:364–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Leipsic J, Taylor CM, Gransar H, Shaw LJ, Ahmadi A, Thompson A. et al. Sex-based prognostic implications of nonobstructive coronary artery disease: results from the international multicenter CONFIRM study. Radiology 2014;273:393–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Cho I, Chang HJ, Sung JM, Pencina MJ, Lin FY, Dunning AM. et al. Coronary computed tomographic angiography and risk of all-cause mortality and nonfatal myocardial infarction in subjects without chest pain syndrome from the CONFIRM Registry (coronary CT angiography evaluation for clinical outcomes: an international multicenter registry). Circulation 2012;126:304–13. [DOI] [PubMed] [Google Scholar]
- 13. Min JK, Dunning A, Lin FY, Achenbach S, Al-Mallah M, Budoff MJ. et al. Age- and sex-related differences in all-cause mortality risk based on coronary computed tomography angiography findings results from the International Multicenter CONFIRM (Coronary CT Angiography Evaluation for Clinical Outcomes: an International Multicenter Registry) of 23,854 patients without known coronary artery disease. J Am Coll Cardiol 2011;58:849–60. [DOI] [PubMed] [Google Scholar]
- 14. Schulman-Marcus J, Lin FY, Gransar H, Berman D, Callister T, DeLago A. et al. Coronary revascularization vs. medical therapy following coronary-computed tomographic angiography in patients with low-, intermediate- and high-risk coronary artery disease: results from the CONFIRM long-term registry. Eur Heart J Cardiovasc Imag 2017;18:841–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Abbara S, Blanke P, Maroules CD, Cheezum M, Choi AD, Han BK. et al. SCCT guidelines for the performance and acquisition of coronary computed tomographic angiography: a report of the society of Cardiovascular Computed Tomography Guidelines Committee: endorsed by the North American Society for Cardiovascular Imaging (NASCI). J Cardiovasc Comput Tomogr 2016;10:435–49. [DOI] [PubMed] [Google Scholar]
- 16. Arsanjani R, Berman DS, Gransar H, Cheng VY, Dunning A, Lin FY. et al. ; For the CONFIRM Investigators. Left ventricular function and volume with coronary CT angiography improves risk stratification and identification of patients at risk for incident mortality: results from 7758 patients in the prospective multinational CONFIRM observational cohort study. Radiology 2014;273:70–7. [DOI] [PubMed] [Google Scholar]
- 17. Thygesen K, Alpert JS, Jaffe AS, Simoons ML, Chaitman BR, White HD. et al. ; the Writing Group on behalf of the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction. Third universal definition of myocardial infarction. Eur Heart J 2012;33:2551–67. [DOI] [PubMed] [Google Scholar]
- 18. Ilia R, Carmel S, Carlos C, Gueron M.. Relation between shortness of breath, left ventricular end diastolic pressure and severity of coronary artery disease. Int J Cardiol 1995;52:153–5. [DOI] [PubMed] [Google Scholar]
- 19. Argulian E, Agarwal V, Bangalore S, Chatterjee S, Makani H, Rozanski A. et al. Meta-analysis of prognostic implications of dyspnea versus chest pain in patients referred for stress testing. Am J Cardiol 2014;113:559–64. [DOI] [PubMed] [Google Scholar]
- 20. Ragosta M, Bishop AH, Lipson LC, Watson DD, Gimple LW, Sarembock IJ. et al. Comparison between angiography and fractional flow reserve versus single-photon emission computed tomographic myocardial perfusion imaging for determining lesion significance in patients with multivessel coronary disease. Am J Cardiol 2007;99:896–902. [DOI] [PubMed] [Google Scholar]
- 21. Nakanishi R, Rana JS, Rozanski A, Cheng VY, Gransar H, Thomson LE. et al. Relationship of dyspnea vs. typical angina to coronary artery disease severity, burden, composition and location on coronary CT angiography. Atherosclerosis 2013;230:61–6. [DOI] [PubMed] [Google Scholar]
- 22. Taqueti VR, Solomon SD, Shah AM, Desai AS, Groarke JD, Osborne MT. et al. Coronary microvascular dysfunction and future risk of heart failure with preserved ejection fraction. Eur Heart J 2018;39:840–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Taqueti VR, Hachamovitch R, Murthy VL, Naya M, Foster CR, Hainer J. et al. Global coronary flow reserve is associated with adverse cardiovascular events independently of luminal angiographic severity and modifies the effect of early revascularization. Circulation 2015;131:19–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Knuuti J, Wijns W, Saraste A, Capodanno D, Barbato E, Funck-Brentano C. et al. ; ESC Scientific Document Group. 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes. Eur Heart J 2020;41:407–77. [DOI] [PubMed] [Google Scholar]