Background
The etiology of reduced left ventricular (LV) ejection fraction after exercise, without obstructive coronary artery disease or other established causes, is unclear. The aims of this study were to determine whether patients undergoing treadmill stress echocardiography with this abnormal LV contractile response to exercise (LVCRE) without established causes have resting LV long-axis dysfunction or microvascular dysfunction and to determine associations with this abnormal LVCRE.
Methods
Of 5,275 consecutive patients undergoing treadmill stress echocardiography, 1,134 underwent cardiac computed tomography angiography or invasive angiography. Having excluded patients with obstructive coronary artery disease, hypertensive response, submaximal heart rate response, resting LV ejection fraction < 50%, and valvular disease, 110 with “abnormal LVCRE” and 212 with “normal LVCRE” were analyzed. Resting mitral annular velocities were measured to assess LV long-axis function. Myocardial blush grade and corrected Thrombolysis In Myocardial Infarction frame count were determined angiographically to assess microvascular function.
Results
Comparing normal LVCRE with abnormal LVCRE, age (mean, 59.7 ± 11.1 vs 61.4 ± 10.0 years), hypertension (53% vs 55%), diabetes (16% vs 20%), and body mass index (mean, 29.1 ± 5.4 vs 29.5 ± 6.4 kg/m 2 ) were similar ( P > .05). Abnormal LVCRE had reduced resting LV long-axis function with lower septal (mean, 6.1 ± 1.9 vs 7.7 ± 2.2 cm/sec) and lateral (mean, 8.1 ± 2.9 vs 10.4 ± 3.0 cm/sec) e′ velocities ( P < .001) and larger resting left atrial volumes (mean, 37.3 ± 10.1 vs 31.1 ± 7.2 mL/m 2 , P < .001). On multivariate analysis, female gender (odds ratio [OR], 1.21; 95% confidence interval [CI], 1.15–1.99; P < .001), exaggerated chronotropic response (OR, 1.49; 95% CI, 1.09–2.05; P < .001), resting left atrial volume (OR, 2.38; 95% CI, 1.63–3.47; P < .001), and resting lateral e′ velocity (OR, 1.70; 95% CI, 1.22–2.49; P = .003) were associated with abnormal LVCRE, but not myocardial blush grade or corrected Thrombolysis In Myocardial Infarction frame count.
Conclusions
An abnormal LVCRE in the absence of established causes is associated with resting LV long-axis dysfunction and is usually seen in women.
Treadmill stress echocardiography (TME) is a robust technique that is widely used for the noninvasive diagnosis of coronary artery disease (CAD). A normal left ventricular (LV) contractile response to exercise (LVCRE) involves a reduction in LV end-systolic volume and an increase in LV ejection fraction (LVEF), which excludes obstructive CAD with high diagnostic accuracy. In contrast, an abnormal LVCRE is defined as an increase in LV end-systolic volume and no change or reduction in LVEF after maximal exercise. As well as CAD, other conditions associated with an abnormal LVCRE are hypertensive response to exercise, failure to achieve ≥85% maximum age-predicted heart rate, reduced resting LVEF, hypertrophic cardiomyopathy, and severe resting valvular disease. Although current theories postulate that in the absence of these conditions, microvascular dysfunction or resting LV long-axis dysfunction may cause an abnormal LVCRE, this remains unclear. However, regardless of etiology, such patients with an abnormal LVCRE in the absence of a clear cause may not have a benign prognosis.
Resting LV long-axis dysfunction may be identified on echocardiography using pulsed-wave Doppler tissue imaging by measuring long-axis tissue velocities at the septal and lateral mitral annulus. Furthermore, myocardial blush grade (MBG) and corrected Thrombolysis In Myocardial Infarction frame count (CTFC) are sensitive measures that can be measured on invasive coronary angiography (ICA) to determine resting microvascular function. Therefore, we sought to determine whether patients with an abnormal LVCRE in the absence of CAD or other known causes have resting LV long-axis dysfunction or microvascular dysfunction and to determine associations with this response.
Methods
Study Population
We conducted a case-control study to evaluate 5,275 consecutive patients who underwent first-time TME at our urban tertiary hospital (Monash Medical Centre, Melbourne, Australia) between January 2009 and June 2012 for the investigation of suspected CAD. Of these, 1,134 underwent clinically indicated computed tomographic coronary angiography (CTCA) or ICA within 90 days of TME. Patients were excluded if they had known etiologies of abnormal LVCRE, namely, obstructive CAD (>50% luminal stenosis in any coronary artery) on CTCA or ICA (622 patients), a hypertensive response (defined as >210/105 mm Hg in men and >190/105 mm Hg in women) (68 patients), failure to achieve >85% of maximum age-predicted ([220 − age] × 0.85) heart rate (45 patients), moderate or greater severity valvular heart disease (18 patients), resting LVEF < 50% (15 patients), or hypertrophic cardiomyopathy (12 patients). The 354 remaining patients were divided into three groups on the basis of LVCRE: (1) 32 patients had LV regional hypokinesis with normal overall LVCRE (>5% increase in postexercise LVEF compared with rest), representing a likely false-positive result for obstructive CAD, and were therefore excluded from further analysis; (2) 110 patients had an abnormal global LVCRE, defined as either a fall or no significant (<5%) increase in postexercise LVEF compared with rest; and (3) 212 patients demonstrated a normal LVCRE, defined as a >5% increase in postexercise LVEF compared with rest ( Figure 1 ). Our institutional ethics committee approved the study, and all patients gave informed consent.
Baseline Rest and TME
All patients underwent symptom-limited treadmill exercise testing using the Bruce protocol. Symptoms and electrocardiography were monitored continuously during exercise. Blood pressure was measured noninvasively at rest and every 3 min during exercise. All patients underwent digital two-dimensional echocardiographic examinations in the left lateral decubitus position at rest and immediately after exercise using one of two commercially available systems: Vivid 7 (GE Healthcare, Australia) or iE33 (Philips Medical Systems, Australia). Postexercise imaging was completed within 60 sec of the termination of exercise in all patients. Images were acquired using a 3.5-MHz transducer in standard parasternal and apical two-, three-, and four-chamber views and viewed on a quad screen in cine-loop format. An echocardiologist, blinded to clinical and exercise data, traced LV endocardial borders on optimal nonforeshortened apical two- and four-chamber views and subsequently derived LV volumes and LVEF using the biplane method of disks on both rest and immediate postexercise images. At rest, a separate blinded analysis was performed to assess LV long-axis function by measuring mitral annular velocities with pulsed-wave Doppler tissue imaging according to American Society of Echocardiography recommendations in the apical four-chamber view at the septal and lateral mitral annulus, with measurements averaged over three consecutive cardiac cycles. The following were also measured at rest according to American Society of Echocardiography recommendations : LV end-diastolic dimension, interventricular septal and posterior wall thickness, left atrial volume, mitral inflow velocities and deceleration time, and right ventricular systolic pressure. Also, mitral leaflets were assessed for billowing or prolapse. Effective arterial elastance index, a measure of noninvasive arterial load imposed on the heart, was calculated as end-systolic pressure divided by stroke volume index. End-systolic pressure was noninvasively approximated by multiplying the systolic blood pressure by 0.9. Stroke volume was calculated by subtracting the end-systolic volume from the end-diastolic volume.
Coronary Angiography
CTCA
All studies were performed on a 320–detector row system (AquilionOne Dynamic Volume CT; Toshiba Medical, Tokyo, Japan). A bolus of 75 mL of iopromide (Ultravist 370; Bayer HealthCare) was injected into an antecubital vein at a flow rate of 5 to 6 mL/sec. An axial scanning technique was used, with slice collimation of 0.5 mm and gantry rotation time of 350 msec. Exposure parameters included an x-ray tube potential of 100 to 135 kVp and effective tube current of 400 to 580 mA, based on vendor specifications and determined by patient body mass index. Scans were performed with prospective electrocardiographic triggering using a 70% to 80% phase window. Computed tomographic coronary angiographic data sets were analyzed by a cardiologist, blinded to clinical and echocardiographic data, for the presence of coronary plaque and to verify absence of >50% luminal stenosis in any coronary segment.
ICA
ICA was performed with a 5- or 6-Fr catheter according to standard techniques via the femoral or radial approach. Two interventional cardiologists, blinded to clinical and echocardiographic data, visually evaluated images by consensus to verify absence of >50% luminal stenosis in any coronary segment and performed MBG and CTFC analyses using standard methodology to determine resting microvascular function. In brief, duration of cine filming exceeded three cardiac cycles in the washout phase to assess the washout of myocardial blush. MBG was identified for each coronary artery according to a dye density score (0 = no myocardial blush or contrast density, 1 = minimal blush, 2 = moderate blush, and 3 = normal blush). CTFC was determined according to the number of cine frames required for dye to reach standardized distal landmarks, as previously described. Because the left anterior descending coronary artery is longer, the frame counts were corrected by dividing by 1.7.
Reproducibility
To define the reproducibility of the resting mitral annular velocities as well as rest and poststress LVEF measurements, 20 patients across both groups were randomly selected and measured by a second, blinded echocardiologist. These data were used for intraobserver and interobserver variability measurements.
Statistical Analysis
Continuous data are expressed as mean ± SD. Categorical data are expressed as absolute count and percentage of cohort. Continuous variables were compared using two-tailed unpaired t tests or Wilcoxon rank sum tests. Categorical variables were compared using χ 2 tests unless the expected frequency was <5, in which case Fisher exact tests were used. The association of selected variables with LVCRE were assessed with univariate and single-step multivariate logistic regression modeling, with significance of .15 used as the cutoff for exclusion from the multivariate model. The following covariates were analyzed: gender, exaggerated chronotropic response, resting systolic blood pressure, peak systolic blood pressure, resting left bundle branch block, LV mass index, resting LVEF, resting left atrial volume index, and resting lateral early diastolic mitral annular velocity (e′). Odds ratios (ORs) with corresponding 95% confidence intervals (CIs) were estimated. To determine whether catheter size or access site approach influenced CTFC, simple linear regression was performed. To determine whether catheter size or access site approach influenced MBG, logistic regression was performed. The reproducibility of LVEF and mitral annular velocities was evaluated by calculating the intraobserver and interobserver variability using the intraclass correlation coefficient. A P value < .05 was considered statistically significant. Analyses were performed using SPSS version 20.0 (SPSS, Inc, Chicago, IL).
Results
Study Population
The final study population comprised 212 patients in the normal LVCRE group and 110 in the abnormal LVCRE group ( Figure 1 ). The median duration between TME and angiography was 41 days, with no cardiac events in any patient between tests. The clinical characteristics of patients in the normal LVCRE group and abnormal LVCRE group are outlined in Table 1 . No significant differences were present between groups for age, body mass index, history of hypertension, diabetes mellitus, hypercholesterolemia, smoking, family history of CAD, or antihypertensive medication use. Patients in the abnormal LVCRE group were more likely to be women.
| Characteristic | Normal LVCRE ( n = 212) | Abnormal LVCRE ( n = 110) | P |
|---|---|---|---|
| Age (y) | 59.7 ± 11.1 | 61.4 ± 10.0 | .16 |
| Women | 113 (53%) | 90 (82%) | <.001 |
| Body mass index (kg/m 2 ) | 29.1 ± 5.4 | 29.5 ± 6.4 | .54 |
| Hypertension ∗ | 113 (53%) | 60 (55%) | .83 |
| Diabetes mellitus | 34 (16%) | 22 (20%) | .37 |
| Hypercholesterolemia † | 98 (46%) | 49 (45%) | .72 |
| Family history coronary disease | 70 (33%) | 27 (25%) | .12 |
| Current smoker | 33 (16%) | 9 (8%) | .06 |
| Obesity ‡ | 82 (39%) | 46 (42%) | .59 |
| Atrial fibrillation § | 9 (4%) | 10 (9%) | .08 |
| Chronic renal impairment | 3 (1%) | 3 (3%) | .41 |
| Permanent pacemaker | 2 (1%) | 1 (1%) | >.99 |
| Antihypertensive medications | |||
| β-blockers | 41 (19%) | 22 (20%) | .89 |
| ACE inhibitors | 30 (14%) | 18 (16%) | .60 |
| Angiotensin receptor blockers | 35 (17%) | 21 (19%) | .56 |
| Calcium channel blockers | 29 (14%) | 18 (16%) | .52 |
| Diuretics | 15 (7%) | 15 (14%) | .06 |
| Other antihypertensive agents | 3 (1%) | 1 (1%) | >.99 |
| Indication for TME | |||
| Chest pain | 163 (77%) | 64 (58%) | <.001 |
| Dyspnea | 16 (8%) | 27 (25%) | <.001 |
| Chest pain and dyspnea | 28 (13%) | 15 (13%) | .96 |
| Presyncope/syncope | 5 (2%) | 3 (3%) | >.99 |
∗ Blood pressure >140/90 mm Hg or treatment for hypertension.
† Total cholesterol >180 mg/dL or treatment for hypercholesterolemia.
‡ Body mass index >30 kg/m 2 .
Exercise and Electrocardiographic Characteristics
Exercise and electrocardiographic characteristics are summarized in Table 2 . Patients in the abnormal LVCRE group exhibited shorter exercise duration, lower workload (metabolic equivalents), marginally higher resting systolic and diastolic blood pressure, marginally higher postexercise systolic blood pressure, higher heart rate at end of stage 1 exercise (126 ± 19 vs 106 ± 19 beats/min, P < .001), higher proportion reaching ≥85% maximum age-predicted heart rate before the end of stage 1 exercise compared with the normal LVCRE group (39 [35%] vs 17 [8%], P < .001), and delayed heart rate recovery at 1-min recovery compared with peak exercise (27 ± 12 vs 33 ± 12 beats/min, P < .001). Patients in the abnormal LVCRE group also had higher peak exercise effective arterial elastance index (6.1 ± 1.5 vs 5.2 ± 1.2 mm Hg · m 2 /mL, P = .002) but no difference in resting effective arterial elastance index (4.6 ± 1.3 vs 4.1 ± 1.1 mmHg · m 2 /mL, P = .49).
| Characteristic | Normal LVCRE ( n = 212) | Abnormal LVCRE ( n = 110) | P |
|---|---|---|---|
| Exercise duration (min) | 8.6 ± 2.9 | 7.1 ± 2.9 | <.001 |
| Workload (METs) | 10.5 ± 4.3 | 8.8 ± 3.0 | <.001 |
| Heart rate (beats/min) | |||
| Rest | 76 ± 14 | 79 ± 14 | .03 |
| End stage 1 | 106 ± 19 | 126 ± 19 | <.001 |
| Peak exercise | 160 ± 16 | 159 ± 16 | .66 |
| 1-min recovery | 127 ± 17 | 132 ± 15 | .02 |
| Fall at 1-min recovery | 33 ± 12 | 27 ± 12 | <.001 |
| Age-predicted heart rate (%) | |||
| End stage 1 | 64 ± 13 | 80 ± 13 | <.001 |
| Peak exercise | 97 ± 8 | 101 ± 9 | <.001 |
| Blood pressure (mm Hg) | |||
| Rest systolic | 127 ± 17 | 136 ± 18 | <.001 |
| Rest diastolic | 78 ± 9 | 81 ± 9 | .04 |
| Peak exercise systolic | 168 ± 17 | 173 ± 16 | .01 |
| Peak exercise diastolic | 80 ± 9 | 82 ± 11 | .06 |
| Effective arterial elastance index (mm Hg · m 2 /mL) | |||
| Rest | 4.1 ± 1.1 | 4.6 ± 1.3 | .49 |
| Peak exercise | 5.2 ± 1.2 | 6.1 ± 1.5 | .002 |
| Limiting symptom | |||
| Exhaustion | 128 (60%) | 41 (37%) | <.001 |
| Dyspnea | 51 (24%) | 55 (50%) | <.001 |
| Chest pain | 22 (11%) | 8 (7%) | .36 |
| Dizziness | 7 (3%) | 4 (4%) | .88 |
| Leg pain | 4 (2%) | 2 (2%) | >.99 |
| Resting ECG rhythm | |||
| Sinus rhythm | 207 (98%) | 98 (89%) | .001 |
| Sinus rhythm/LBBB | 3 (1%) | 10 (9%) | .002 |
| Atrial fibrillation | 1 (0.5%) | 1 (1%) | >.99 |
| Paced rhythm | 1 (0.5%) | 1 (1%) | >.99 |
| Postexercise arrhythmia | |||
| None | 186 (88%) | 83 (75%) | .005 |
| Ventricular bigeminy/frequent VPBs | 16 (8%) | 15 (13%) | .08 |
| Nonsustained ventricular tachycardia | 4 (2%) | 4 (4%) | .45 |
| Nonsustained atrial tachycardia | 3 (1%) | 5 (5%) | .13 |
| New-onset atrial fibrillation | 3 (1%) | 2 (2%) | >.99 |
| Sustained atrial tachycardia | — | 1 (1%) | .34 |
| Other postexercise ECG changes | |||
| None | 176 (83%) | 63 (57%) | <.001 |
| ≥1-mm ST-segment depression | 31 (15%) | 32 (29%) | .002 |
| New-onset LBBB | 1 (0.5%) | 4 (4%) | .04 |
| Not interpretable ∗ | 4 (2%) | 11 (10%) | .003 |
Patients in the abnormal LVCRE group were more likely to have an exercise-limiting symptom of dyspnea, whereas patients in the normal LVCRE group were more likely to have an exercise-limiting symptom of exhaustion, with no difference between groups in the rates of exercise-limiting chest pain, dizziness, or leg pain. Patients in the abnormal LVCRE group were more likely to have left bundle branch block on resting electrocardiography and had a higher rate of exercise-induced ≥1-mm ST-segment depression and exercise-induced left bundle branch block than patients in the normal LVCRE group.
Echocardiographic Findings
All echocardiographic images were of diagnostic quality. Resting and peak-exercise LV volumes and LVEF are shown in Table 3 . Resting LV volumes were similar, with lower resting LVEF in the abnormal LVCRE group (60 ± 6% vs 62 ± 6%, P = .005). As per study design, postexercise LV volumes were higher and postexercise LVEF was lower in the abnormal LVCRE group (54 ± 8% vs 75 ± 6%, P < .001). Postexercise regional wall motion abnormalities were identified in 34 patients (31%) in the abnormal LVCRE group, despite no obstructive CAD. Of patients in the abnormal LVCRE group, 11 (10%) had absolute reductions of LVEF of >15% postexercise compared with rest, 18 (16%) had 11% to 15% reductions, 23 (21%) had 6% to 10% reductions, 37 (34%) had 1% to 5% reductions, and 21 (19%) had either no change or <5% increases in LVEF after exercise.
| Variable | Normal LVCRE ( n = 212) | Abnormal LVCRE ( n = 110) | P |
|---|---|---|---|
| Rest | |||
| LVd4C volume (mL) | 88 ± 25 | 91 ± 29 | .40 |
| LVd2C volume (mL) | 88 ± 26 | 90 ± 29 | .47 |
| LVs4C volume (mL) | 33 ± 12 | 37 ± 16 | .009 |
| LVs2C volume (mL) | 33 ± 13 | 36 ± 14 | .06 |
| Calculated LVEF (%) | 62 ± 6 | 60 ± 6 | .005 |
| Peak exercise | |||
| LVd4C volume (mL) | 75 ± 24 | 92 ± 29 | <.001 |
| LVd2C volume (mL) | 79 ± 21 | 95 ± 27 | <.001 |
| LVs4C volume (mL) | 20 ± 10 | 44 ± 18 | <.001 |
| LVs2C volume (mL) | 20 ± 9 | 43 ± 17 | <.001 |
| Calculated LVEF (%) | 75 ± 6 | 54 ± 8 | <.001 |
| Change in LVEF (%) ∗ | +12 ± 5 | −6 ± 7 | <.001 |
Resting echocardiographic parameters are described in Table 4 . Importantly, compared with the normal LVCRE group, the abnormal LVCRE group demonstrated lower septal (6.1 ± 1.9 vs 7.7 ± 2.2 cm/sec, P < .001) and lower lateral (8.1 ± 2.9 vs 10.4 ± 3.0 cm/sec, P < .001) e′ velocities, as well as higher left atrial volume index (37.3 ± 10.1 vs 31.1 ± 7.2 mL/m 2 , P < .001). There was no difference in LV relative wall thickness between groups, but patients in the abnormal LVCRE group had higher LV end-diastolic dimension, higher LV mass index, lower early diastolic transmitral flow velocity (E), shorter mitral deceleration time, lower septal and lateral systolic mitral annular velocities (s′), and higher E/e′ ratio than patients in the normal LVCRE group. Patients in the abnormal LVCRE group were also more likely to have resting diastolic dysfunction and had a higher proportion of mitral leaflet billowing or prolapse.
| Variable | Normal LVCRE ( n = 212) | Abnormal LVCRE ( n = 110) | P |
|---|---|---|---|
| Two-dimensional measurements | |||
| LVDd (cm) | 4.8 ± 0.5 | 5.0 ± 0.5 | .02 |
| LVDd index (cm/m 2 ) | 2.5 ± 0.3 | 2.7 ± 0.3 | <.001 |
| IVSd (cm) | 0.9 ± 0.2 | 0.8 ± 0.2 | .02 |
| LVPWd (cm) | 0.8 ± 0.1 | 0.8 ± 0.1 | .50 |
| LV RWT (%) | 34.7 ± 6.0 | 34.0 ± 6.1 | .37 |
| LV mass (g) | 143.4 ± 43.4 | 146.7 ± 42.1 | .52 |
| LV mass index (g/m 2 ) | 73.4 ± 18.6 | 79.6 ± 20.5 | .006 |
| LA volume index (mL/m 2 ) | 31.1 ± 7.2 | 37.3 ± 10.1 | <.001 |
| Mitral inflow velocities | |||
| E (cm/sec) | 0.8 ± 0.2 | 0.7 ± 0.2 | .002 |
| A (cm/sec) | 0.7 ± 0.2 | 0.8 ± 0.2 | <.001 |
| E/A ratio | 1.2 ± 0.7 | 0.9 ± 0.3 | <.001 |
| DT (msec) | 216.8 ± 44.8 | 204.1 ± 38.6 | .01 |
| Doppler tissue imaging | |||
| Septal e′ (cm/sec) | 7.7 ± 2.1 | 6.1 ± 1.9 | <.001 |
| Septal a′ (cm/sec) | 9.3 ± 1.9 | 9.3 ± 2.4 | .83 |
| Septal e′/a′ ratio | 0.88 ± 0.36 | 0.71 ± 0.40 | <.001 |
| Septal s′ (cm/sec) | 7.6 ± 1.6 | 6.6 ± 1.6 | <.001 |
| Lateral e′ (cm/sec) | 10.4 ± 3.0 | 8.1 ± 2.9 | <.001 |
| Lateral a′ (cm/sec) | 10.0 ± 2.2 | 10.4 ± 2.7 | .22 |
| Lateral e′/a′ ratio | 1.1 ± 0.65 | 0.85 ± 0.45 | <.001 |
| Lateral s′ (cm/sec) | 8.6 ± 2.1 | 7.7 ± 2.2 | .001 |
| Diastolic dysfunction grade | |||
| None | 61 (29%) | 6(5%) | <.001 |
| Grade I | 115 (54%) | 77 (70%) | <.001 |
| Grade II | 35 (16%) | 26 (24%) | <.001 |
| Grade III | 1 (1%) | 1 (1%) | <.001 |
| LV diastolic filling pressure | |||
| Septal E/e′ ratio | 10.4 ± 3.6 | 11.8 ± 3.9 | .002 |
| Septal E/e′ ratio < 15 | 185 (87%) | 90 (82%) | .19 |
| Septal E/e′ ratio ≥ 15 | 27 (13%) | 20 (18%) | .19 |
| RVSP (mm Hg) ∗ | 27.4 ± 5.0 | 29.5 ± 5.5 | .004 |
| Mitral leaflet billowing/prolapse | 2 (1%) | 6 (5%) | .02 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
