In patients suspected of having coronary artery disease (CAD), we compared the diagnostic sensitivity and specificity of exercise testing using ST-segment changes alone and ST-segment changes, angina pectoris, and hemodynamic variables compared to coronary computed tomographic angiography (CTA). Quantitative invasive coronary angiography was the reference method (>50% coronary lumen reduction). A positive exercise test was defined as the development of significant ST-segment changes (≥1 mV measured 80 ms from the J-point), and the occurrence of one or more of the following criteria: ST-segment changes ≥1 mV measured 80 ms from the J-point, angina pectoris, ventricular arrhythmia (the occurrence of ≥3 premature ventricular beats), and ≥20 mm Hg decrease in systolic blood pressure during the test. Positive results on CTA were defined as a coronary lumen reduction of ≥50%. In 100 patients (61 ± 9 years old, 50% men, and 29% prevalence of significant CAD), the diagnostic sensitivity and specificity of exercise testing using ST-segment changes was 45% (95% confidence interval 53% to 87%) and 63% (95% confidence interval 61% to 84%), respectively. However, the inclusion of all test variables yielded a sensitivity of 72% (95% confidence interval 53% to 87%) and a specificity of 37% (95% confidence interval 26% to 49%). The diagnostic sensitivity of 97% (95% confidence interval 82% to 100%) and specificity of 80% (95% confidence interval 69% to 89%) for CTA, however, were superior to any of the exercise test analysis strategies. In conclusion, in patients suspected of having CAD, the diagnostic sensitivity of exercise testing significantly improves if all test variables are included compared to using ST-segment changes exclusively. Furthermore, the superior diagnostic performance of CTA for the detection and exclusion of significant CAD might favor CTA as the first-line diagnostic test in patients suspected of having CAD.
Exercise testing is widely available, safe, and easy to perform. Despite limited diagnostic performance, the exercise test is still recommended as a first-line diagnostic test in patients suspected of having coronary artery disease (CAD). Noninvasive coronary computed tomographic angiography (CTA) has emerged as a novel diagnostic tool. In numerous studies, CTA has demonstrated a high diagnostic accuracy for the detection and exclusion of obstructive CAD. CTA is currently considered appropriate for evaluating symptomatic patients with intermediate pretest likelihood of significant CAD. Different diagnostic algorithms for evaluating patients suspected of having CAD have been proposed, but few studies have explored the relation between exercise testing and CTA. In previous head-to-head comparisons in which CTA demonstrated its superiority to exercise testing in detecting significant CAD, a positive exercise test was defined by electrocardiographic ST-segment changes without taking symptoms or hemodynamic variables into account. Evaluating an exercise test by considering ST-segment changes exclusively might be considered an oversimplification. The purpose of the present study was to investigate the diagnostic performance of exercise testing using a diagnostic definition according to the ST-segment changes or the development of angina pectoris, ST-segment changes, and hemodynamic variables compared to CTA.
Methods
From August 2006 to November 2007, patients referred for invasive coronary angiography (CAG) because of suspicion of CAD were prospectively enrolled in a study evaluating the diagnostic performance of CTA. Patients underwent 64-slice CTA or dual-source CTA within 1 week after CAG and before any interventional treatment. The exclusion criteria for CTA included renal insufficiency, clinical instability (Canadian Cardiovascular Society class IV, New York Heart Association class IV, or systolic blood pressure <95 mm Hg), known allergy to iodinated contrast, known CAD, inadequate scanner capacity, and pregnancy. The patients scheduled for CTA with 64-slice computed tomography were excluded if they presented with atrial fibrillation, irregular heart rate, or baseline heart rates ≥65 and contraindications to β-blocking agents. Of 211 consecutive patients undergoing both CAG and CTA, 106 also underwent exercise testing as a part of the diagnostic workup before angiography. Of these 106 patients, complete exercise test reports, including symptoms, 12-lead electrocardiographic recordings, blood pressure, and heart rate response were obtained for the 100 patients included in the present study.
The patient risk factors were derived through a structured interview and medical records. Hypertension was defined as a documented history of blood pressure ≥140/90 mm Hg or antihypertensive treatment. Hypercholesterolemia was defined as the presence of total cholesterol ≥4.5 mmol/L or lipid-lowering treatment. A positive family history of premature CAD was defined as the presence of CAD in a first-line relative before age 55 years for men or 65 years for women. Smoking was defined as current or previous smoker. Patients with a fasting plasma glucose level ≥7.0 mmol/L or receiving antidiabetic treatment were classified as having diabetes. Using age, gender, and symptoms, the patients were stratified into a low (<13.4%), intermediate (≥13.4% but ≤87.2%), or high (>87.2%) pretest probability of CAD (≥50% coronary artery lumen reduction). Typical angina pectoris was defined as substernal discomfort or chest pain provoked by physical exercise or emotional stress and relieved by rest or nitroglycerin. The presence of 2 of these characteristics defined atypical angina, and the presence of 1 defined nonanginal chest pain. The local ethics committee approved the study, and all patients provided written informed consent.
A maximum symptom-limited bicycle exercise test was performed according to a standardized protocol. A 12-lead electrocardiogram was monitored continuously throughout the test, and the blood pressure was measured at every incremental exertion stage, immediately after discontinuation, and every 2 minutes until the blood pressure at rest values had been attained. The initial exercise load was 25 or 50 W, increasing by 25 W every 2 minutes until discontinuation. The exercise test was discontinued at maximum exertion, typical angina pectoris of ≥6 on the Borg scale, ≥1 mV ST-elevation in non–Q-wave QRS complexes, or ventricular arrhythmia (occurrence of ≥3 premature ventricular beats). The exercise test results were independently analyzed in a core laboratory by 2 experienced readers who were unaware of the results from CTA and CAG. In the case of disagreement between the 2 observers, consensus was obtained by a third observer. Positive exercise test results considering the ST-segment changes was defined by the presence of ≥0.1-mV horizontal/descending ST-depression or ST-elevation measured >80 ms from the J-point in a non–Q-wave lead during the exercise or recovery phase. Positive exercise test results considering all test variables were defined by the presence of at least one of the following criteria: ≥0.1-mV horizontal/descending ST-depression or ST-elevation measured >80 ms from the J-point in a non–Q-wave-lead during the exercise or recovery phase, the development of angina pectoris, ventricular arrhythmia (occurrence of ≥3 premature ventricular beats), or ≥20 mm Hg decrease in systolic blood pressure during the test. ST-segment changes were considered nonevaluable in patients presenting with left bundle branch block. The exercise test considering ST-segment changes only was considered inconclusive in the presence of left bundle branch block or the absence of ST-segment changes as described earlier in this paragraph without reaching 85% of the age-adjusted maximum heart rate (maximum heart rate = 220 beats/min − age in years). For the exercise test that included all test variables, an inconclusive test result was defined as the absence of the above-mentioned predictors without reaching 85% of the age-adjusted maximum heart rate.
Before 64-slice CTA, patients with a heart rate at rest of ≥65 beats/min received 50 mg of metoprolol orally, and, if necessary additionally intravenous metoprolol was given to lower the heart rate further. However, CTA was performed regardless of the achieved heart rate. β-Blocking agents were not routinely administered before CTA using dual-source CTA. All patients received 0.25 mg nitroglycerin sublingually 5 minutes before CTA. The first 51 patients were examined using a 64-slice CTA scanner (Somatom Sensation, Siemens Medical Solutions, Forcheim, Germany), and the last 49 examinations were performed using a dual-source CTA scanner (Siemens Definition, Siemens Medical Solutions). An initial nonenhanced scan was performed for calcium scoring. Coronary artery calcifications were quantified (Syngo Calcium Scoring, Siemens Medical Solutions) and measured using the Agatston score. For the enhanced scan, 70 to 100 ml of iodinated contrast (Iomeron 400, Bracco, Milan, Italy) was injected (5 ml/s) in the antecubital vein, followed by a 40-ml saline flush. The scan was initiated when the ascending aorta contrast attenuation reached a predefined threshold of 100 Hounsfield units (bolus tracking). The scan parameters for 64-slice CT were helical scan mode, detector collimation 32 × 0.6 mm with z-flying focal spot, gantry rotation time 330 ms, pitch 0.2, tube voltage 120 kV (both tubes) and tube current 131 to 850 mA. The scan parameters for dual-source CTA were helical scan mode, detector collimation 32 × 2 × 0.6 mm, with z-flying focal spot, gantry rotation time 330 ms, pitch 0.2 to 0.43 (adapted to heart rate), tube voltage 120 kV (both tubes) and tube current 131 to 850 mA. The effective temporal resolution was 165 ms for 64-slice CTA and 83 ms for dual-source CTA. Full tube current was applied at 30% to 80% of the RR interval if the heart rate was <65 beats/min (or <72 beats/min for dual-source CTA) or during the entire RR interval if the heart rate was faster. The effective radiation dose of CTA was estimated using the dose–length product × conversion factor (chest 0.014 m/Sv mGy/cm).
The images were reconstructed using a medium smooth kernel (B26f, Siemens Medical Solution), with a slice thickness of 0.75 and a 0.4-mm increment. In the presence of coronary calcification, additional reconstructions with a sharp convolution kernel (B46f) were generated. The scans were typically reconstructed during the mid-to-end-diastolic phase. Multiple data sets were evaluated per patient, if necessary. One observer, who was unaware of the patient history and CAG findings, performed the CTA using axial source images, multiplanar reconstructions and curved multiplanar reconstructions. All coronary segments >1.5 mm were evaluated according to a modified 16-segment classification model in which the intermediate artery represented segment 16. Significant CAD was defined as the presence of coronary lumen reduction of ≥50%. Nonsignificant CAD was defined as coronary lesions resulting in <50% coronary lumen reduction.
CAG was performed according to standard techniques. Standardized projections were acquired, and intracoronary nitroglycerin was administered if coronary lumen reduction was detected. The angiograms were evaluated using quantitative coronary angiography by 2 experienced observers who were unaware of the findings from CTA (Medis, Medical Imaging Systems, Leiden, The Netherlands). Consensus readings were performed in the case of discrepancies. The coronary segments were identified according to a modified 16-segment classification model, as previously described. A segment with a significant stenosis was defined by a lumen diameter reduction of ≥50%.
Diagnostic accuracy is presented as the sensitivity, specificity, and positive and negative predictive values and corresponding 95% confidence intervals (CIs). The pair-wise McNemar test was used to compare the diagnostic sensitivity and specificity of the exercise testing and CTA on a per-patient basis with reference to the CAG findings. In an intention-to-diagnose approach, inconclusive findings from either exercise testing or CTA were considered positive, if not otherwise stated. Interobserver reproducibility of the exercise test was evaluated by kappa statistics on a per-patient basis. Categorical data are expressed as counts (or proportions), and continuous data are expressed as the mean ± SD or median values with corresponding twenty-fifth and seventy-fifth percentiles, as appropriate. To evaluate differences in demographic and clinical characteristics, the Fischer exact test, Wilcoxon rank sum test, and Student t test were applied as appropriate. p Values <0.05 were considered statistically significant. Data analyses were performed using Stata/IC, version 10.0 (StataCorp, College Station, Texas).
Results
The baseline characteristics and findings from the exercise testing and CTA are presented in Tables 1, 2, and 3 , respectively. The pretest probability of significant CAD was low in 10% (n = 10), intermediate in 55% (n = 55), and high in 35% (n = 35) of the patients. Significant CAD was detected in 29% of the patients (22 men and 7 women), 1-vessel disease in 18%, 2-vessel disease in 5%, and 3-vessel disease in 6% of the patients.
| Characteristic | Value |
|---|---|
| Age (years) | 61 ± 9 |
| Men | 50 (50%) |
| Hypertension ⁎ | 50 (50%) |
| Hypercholesterolemia † | 69 (69%) |
| Family history of premature coronary artery disease ‡ | 53 (53%) |
| Smoker § | 52 (52%) |
| Diabetes mellitus ¶ | 3 (3%) |
| Body mass index (kg/m 2 ) | 27 ± 4 |
| Nonangina pectoris | 35 (35%) |
| Atypical angina pectoris | 26 (26%) |
| Typical angina pectoris | 39 (39%) |
⁎ Antihypertensive treatment or documented history of blood pressure ≥140/90 mm Hg.
† Lipid-lowering treatment or total cholesterol ≥4.5 mmol/L.
‡ Coronary artery disease in first-line relative before 55 years old for men or 65 years old for women.
¶ Antidiabetic treatment or fasting plasma glucose levels ≥7.0 mmol/L.
| Characteristic | Value |
|---|---|
| Median interval before invasive coronary angiography ⁎ (wk) | 9 (6–15) |
| Exercise time (min) | 10.2 ± 3.3 |
| Maximum systolic blood pressure (mm Hg) | 203 ± 26 |
| Maximum heart rate (beats/min) | 151 ± 21 |
| Maximum systolic blood pressure × maximum heart rate | 31,067 ± 6,500 |
| Age-adjusted maximum heart rate (%) | 96 ± 12 |
| Exercise test | |
| ST-segment changes | |
| Negative | 61 (61%) |
| Positive | 27 (27%) |
| Inconclusive | 12 (12%) |
| All variables † | |
| Negative | 34 (34%) |
| Positive | 63 (63%) |
| Inconclusive | 3 (3%) |
⁎ Data in parentheses are interquartile range.
| Characteristic | Value |
|---|---|
| Heart rate (beats/min) | 61 ± 10 |
| Contrast media (ml) | 83 ± 20 |
| Effective radiation dose (mSv) | 9.4 ± 3.4 |
| Coronary artery calcium score, median (IQR) | 23 (0–189) |
| Coronary artery calcium score = 0 | 26 (26%) |
| Test result | |
| Negative | 58 (58%) |
| Positive | 37 (37%) |
| Inconclusive | 5 (5%) |
The pretest electrocardiogram was without pathologic findings for 74 patients. One patient presented with left bundle branch block, 6 patients with right bundle branch block, and 19 patients with pathologic Q waves. The test was positive in 63 patients, with angina pectoris in 28, ST-segment depression in 12, ST-segment depression and angina pectoris in 14, a decrease in systolic blood pressure in 4, ventricular arrhythmia in 3, ventricular arrhythmia combined with ST-segment depression in 1, and a decrease in systolic blood pressure in 1. The exercise test diagnostic values are listed in Table 4 . No significant difference in the diagnostic sensitivity or specificity between men and women was detected: 73% (95% CI 50% to 89%) versus 71% (95% CI 29% to 96%; p = NS) and 46% (95% CI 28% to 66%) versus 30% (95% CI 17% to 46%; p = NS), respectively. The core laboratory readings warranted consensus readings for 6 patients, reaching a kappa of 0.87 (95% CI 0.76 to 0.97).
| Analysis | Patients (n) | True Positive | False Negative | False Positive | True Negative | Sensitivity | Specificity | PPV | NPV |
|---|---|---|---|---|---|---|---|---|---|
| Reference ≥ 50% quantitative invasive coronary angiography stenosis, intention-to-diagnose | |||||||||
| Exercise test | |||||||||
| ST-segment changes | 100 | 13 | 16 | 26 | 45 | 45% (26–64%) ⁎ | 63% (61–84%) † | 33% (19–50%) | 74% (61–84%) |
| All variables ‡ | 100 | 21 | 8 | 45 | 26 | 72% (53–87%) § | 37% (26–49%) ⁎ | 32% (21–44%) | 77% (59–89%) |
| Coronary CTA | 100 | 28 | 1 | 14 | 57 | 97% (82–100%) | 80% (69–89) | 67% (51–80%) | 98% (91–100%) |
| Reference ≥50% quantitative invasive CAG stenosis, inconclusive tests excluded | |||||||||
| Exercise test | |||||||||
| ST-segment changes | 88 | 7 | 16 | 20 | 45 | 30% (13–53%) ⁎ | 69% (57–80%) ¶ | 26% (11–46%) | 74% (61–84%) |
| All variables ‡ | 97 | 20 | 8 | 43 | 26 | 71% (51–87%) § | 38% (26–50%) ⁎ | 32% (21–45%) | 77% (59–89%) |
| Coronary CTA | 95 | 26 | 1 | 11 | 57 | 96% (81–100%) | 84% (73–92%) | 70% (53–84%) | 98% (91–100%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
