Utility of E/e′ Ratio During Low-Level Exercise to Diagnose Heart Failure With Preserved Ejection Fraction
Original Research
Central Illustration
Abstract
Background
E/e′ ratio during exercise is the key parameter in identifying elevated pulmonary capillary wedge pressure (PCWP), and thus heart failure with preserved ejection fraction (HFpEF). However, its diagnostic value is limited when mitral inflow or tissue velocities are fused during elevated heart rate.
Objectives
The authors hypothesized that E/e‘ ratio during low-level (20 W) exercise (E/e′20W) can help diagnose HFpEF.
Methods
Ergometric exercise stress echocardiography was performed in 215 dyspneic patients with an EF ≥50%. The authors determined the feasibility of E/e′ ratio at each stage (frequency of patients who had measurable E/e′ without E-A fusion among 215 participants) and examined whether E/e′20W could predict normal E/e′ ratio during peak exercise (E/e′peak ≤15). The authors also evaluated whether E/e′20W could predict normal PCWP during exercise (PCWP <25 mm Hg) in a subset of participants (n = 45) who underwent exercise right heart catheterization.
Results
The feasibility of the E/e′ ratio decreased from 100% at rest to 96.3% during 20-W exercise and 74.9% during peak exercise caused by E-A fusion. In patients with E/e′peak >15, there was an increase in E/e′ ratio from rest to 20-W exercise (11.2 ± 2.1 to 16.3 ± 3.5; P < 0.0001), but it did not change significantly from 20-W exercise to peak exercise (P = 0.12). E/e′20W predicted E/e′peak ≤15 (AUC: 0.91; P < 0.0001) with the cutoff value of ≤12.4 showing high specificity (94%) and positive predictive value (98%). During 20-W exercise, 93% of the HFpEF patients developed PCWP ≥25 mm Hg. E/e′20W predicted normal PCWP during exercise (AUC: 0.77; P = 0.01) with the cutoff value of ≤12.4 showing high specificity (83%).
Conclusions
E/e′ ratio during low-level exercise is highly feasible and predicts normal E/e′ ratio or PCWP during peak exercise with high specificity. These data suggest that E/e′20W could be used as an alternative to the peak exercise value to rule out HFpEF in patients with dyspnea.
Introduction
Approximately one-half of patients with heart failure (HF) have heart failure with preserved ejection fraction (HFpEF).1,2 HFpEF represents one of the greatest health care problems worldwide, given its increasing prevalence, limited therapeutic options, and high morbidity and mortality rates.1,3 The diagnosis of HFpEF among people with no overt congestion remains a challenge and relies on the demonstration of objective evidence of elevated left ventricular (LV) filling pressure.1,2,4-6 This difficulty could partly be related to the fact that LV filling pressure is often normal at rest but increases abnormally only during physiologic stress activities, such as exercise.4-7 To address this diagnostic difficulty, recent consensus recommendations have proposed exercise stress echocardiography (ie, diastolic stress test) to unmask the diastolic abnormalities that develop only during exercise.8,9
The ratio of early diastolic mitral inflow velocity to early diastolic mitral annular tissue velocity (E/e′) is the key parameter in estimating the LV filling pressure in exercise stress echocardiography. Normal E/e′ ratio during peak exercise (<15) may be useful to rule out elevated LV filling pressure or HFpEF among patients with dyspnea.5,10-16 However, the diagnostic utility is limited if the transmitral flows are fused when the heart rate is elevated. Although some guidelines have proposed the acquisition of E/e′ ratio during a submaximal stage (heart rate: 100-110 beats/min) or after exercise, the data supporting this are limited.9,17-19 Studies using invasive hemodynamics have consistently shown that an abnormal increase in the LV filling pressure occurs early during low-level (20 W) exercise in patients with HFpEF.4,5,7,20,21 During this exercise stage, the acquisition of images is technically easy and diastolic Doppler flows are rarely fused. However, it remains unknown whether normal E/e′ ratio during 20-W exercise (E/e′20W) could be useful to rule out HFpEF.
Based on this hemodynamic background, we hypothesized that E/e′20W is a preferred alternative to the E/e′ ratio during peak exercise (E/e′peak). Given the high negative predictive value of E/e′peak for ruling out HFpEF, we assessed the diagnostic ability of E/e′20W to predict normal E/e′ ratio during peak exercise (E/e′peak ≤15) in patients with unexplained dyspnea without evidence of elevated LV filling pressure at rest. We further examined whether E/e′20W could predict normal pulmonary capillary wedge pressure (PCWP) during exercise (<25 mm Hg) based on exercise right heart catheterization (RHC).
Methods
This was a retrospective study conducted at the Gunma University Hospital (Maebashi, Japan). Some participant data from this study have been previously published,22-24 but not in relation to the diagnostic evaluation of HFpEF. This study was approved by our institutional review board, and the requirement for patient consent was waived.
Protocol 1
Participants
A total of 528 consecutive patients were screened who were referred to the echocardiography laboratory of the Gunma University Hospital for supine bicycle stress echocardiography to evaluate exertional dyspnea from September 2019 to July 2022. The clinical indication of exercise echocardiography was determined by the referring physicians. Patients aged <20 years, those with reduced ejection fraction (<50%), significant left-side valvular heart disease (more than moderate regurgitation and more than mild stenosis), hypertrophic cardiomyopathy, or insufficient exercise performance (<30 seconds), and those subjected to repeated studies were excluded. We also excluded patients who could not reach the 40 W exercise stage and those with elevated resting E/e′ ratio (E/e′rest >15) or natriuretic peptide levels (B-type natriuretic peptide [BNP] >80 pg/mL or N-terminal pro–B-type natriuretic peptide [NT-proBNP] >220 pg/mL in sinus rhythm; BNP >240 pg/mL or NT-proBNP >660 pg/mL in atrial fibrillation [AF]).18 Finally, 215 patients were included in protocol 1 (Figure 1). The probability of HFpEF was evaluated according to the HFA-PEFF (Heart Failure Association pre-test assessment, echocardiography and natriuretic peptide, functional testing, and final etiology) algorithm in step 2.18
Study Flowchart
Of the 528 dyspneic patients who were referred to exercise echocardiography, 215 patients were included in protocol 1. Forty-five patients who underwent exercise right heart catheterization were included in protocol 2. EF = ejection fraction; HFpEF = heart failure with preserved ejection fraction; NP = natriuretic peptide; PCWP = pulmonary capillary wedge pressure.
Exercise stress echocardiography
Transthoracic echocardiography at rest and during exercise was performed by experienced sonographers, and all Doppler measurements represented the mean measurements from at least 3 heartbeats. Using mitral E velocity recorded at the mitral leaflet tips and mitral annular e′ velocity measured at the septal annulus, the septal E/e′ ratio was determined. Tricuspid regurgitation (TR) severity at rest and during exercise was visually categorized as none/trivial, mild, moderate, or severe. Left atrial (LA) reservoir strain at rest and during peak exercise was measured from apical 4-chamber views with the use of commercially available software (EchoPAC, GE).
A symptom-limited, supine, bicycle exercise stress echocardiography was performed according to the contemporary guidelines using cycle ergometry (Load). The first stage of exercise (20 W) was performed for 5 minutes, followed by graded 20-W increments in the workload (3-minute stages) up to patient-reported exhaustion. Echocardiographic data were obtained during all stages of the exercise and recovery phase (pedaling with no load). Mitral inflow velocity and tissue Doppler were acquired 30 seconds after the initiation of each exercise stage. Peak oxygen consumption was measured simultaneously in a subset of the patients (n = 160).
Protocol 2
Participants
Among all of the participants, we identified those who underwent exercise RHC after exercise stress echocardiography (n = 68). Exercise RHC was performed as confirmatory testing in patients with equivocal noninvasive data, and the indication was determined by the referring physicians. Patients with significant left-side heart disease, ejection fraction <50%, high resting LV filling pressure (PCWP >15 mm Hg), and those subjected to repeated studies were excluded (Figure 1). Finally, 45 patients were included in protocol 2.
Exercise cardiac catheterization protocol
RHC was performed using a 9-F sheath inserted through the right internal jugular vein. Details are provided in the Supplemental Methods. Intracardiac pressures were measured at rest, during all stages of exercise, and during 1-minute recovery (pedaling with no load). The patients underwent the same supine cycle ergometry protocol as that in exercise stress echocardiography. The diagnosis of HFpEF was adjudicated by PCWP >15 mm Hg at rest and/or ≥25 mm Hg during ergometry exercise. Patients who did not meet the criteria were included as controls.
Assessment of diagnostic performance of E/e′ ratio during the 20-W exercise in each protocol
In protocol 1, we investigated the diagnostic ability of E/e′20W to predict normal E/e′ ratio during peak exercise (E/e′peak ≤15). In protocol 2, we examined the diagnostic performance of E/e′20W in predicting normal PCWP during exercise (<25 mm Hg) based on the aforementioned PCWP criteria.
Statistical analysis
Data are reported as mean ± SD, median (IQR), or n (%). Between-group differences were compared by unpaired Student's t-test or Wilcoxon rank sum test. Within-group differences were compared by 1-way repeated analysis of variance. The diagnostic accuracy of the E/e′ ratio at each stage was determined with the use of receiver-operating characteristic curves in participants with the obtainable data. The optimal cutoff was obtained according to the Youden index. Comparison of areas under the receiver-operating characteristic curves (AUCs) was performed using the method of DeLong.25 All tests were 2-sided, with values of P < 0.05 considered to be statistically significant. All analyses were performed using JMP 14.0.0 (SAS Institute).
Results
Protocol 1
Patient characteristics and echocardiographic findings at rest
Overall, the participants were elderly and had hypertension, an intermediate HFA-PEFF, and normal natriuretic peptide levels (Table 1). The patients displayed normal LV size, mass, and ejection fraction. The E/e′ ratio, LA size and reservoir strain, TR velocity, and right ventricular function were within normal ranges.
| Age, y | 68 ± 12 |
| Female | 111 (52) |
| Body mass index, kg/m2 | 24 ± 6 |
| HFA-PEFF score | 3 (2-4) |
| Comorbidities | |
| Hypertension | 148 (69) |
| Coronary artery disease | 13 (6) |
| Atrial fibrillation | 51 (24) |
| Diabetes mellitus | 31 (14) |
| Significant mitral annular calcification | 2 (1) |
| Medications | |
| ACEI or ARB | 69 (32) |
| Beta-blocker | 38 (18) |
| Calcium channel blocker | 70 (33) |
| Diuretic | 21 (10) |
| Laboratory tests | |
| Hemoglobin, g/dL | 13.4 ± 1.5 |
| Creatinine, mg/dL | 0.7 (0.6-0.9) |
| BNP, pg/mL (n = 122) | 32 (15-55) |
| NT-proBNP, pg/mL (n = 124) | 95 (56-156) |
| Vital signs | |
| Heart rate, beats/min | 73 ± 13 |
| Systolic blood pressure, mm Hg | 129 ± 21 |
| Echocardiography | |
| LV end-diastolic volume, mL | 72 ± 24 |
| LV mass index, g/m2 | 81 ± 19 |
| LV ejection fraction, % | 64 ± 6 |
| Mitral E, cm/s | 64 ± 16 |
| Mitral e′, cm/s | 6.9 ± 1.8 |
| Mitral s′, cm/s | 7.9 ± 1.7 |
| E/e′ ratio | 9.6 ± 2.3 |
| Left atrial volume index, mL/m2 | 28 ± 12 |
| Left atrial reservoir strain, % | 33 ± 14 |
| Tricuspid regurgitant velocity, m/s | 2.1 ± 0.4 |
| TAPSE, mm | 20 ± 4 |
| Right ventricular s′, cm/s | 12 ± 3 |
E/e′ ratio at rest and during exercise
By definition, all participants performed exercise of ≥40 W workload, and the median exercise workload achieved was 60 W (IQR: 50-80 W). E/e′rest and E/e′ during passive leg raise (E/e′PLR) could be obtained in all participants (feasibility 100% for both). Overall, the heart rate increased to 92 beats/min during the 20-W exercise and to 115 beats/min at peak exercise. The feasibility of E/e′20W remained high (96.3%; n = 207) but decreased to 74.9% (n = 161) at peak exercise because of E-A fusion. E/e′recovery could be measured in 192 patients (89.3%). Overall, the prevalence of significant TR (TR ≥ moderate) was slightly increased from rest to peak exercise (0% to 4%).
Changes in septal E/e′ ratio throughout the exercise were compared between the 2 groups divided based on peak E/e′ ratio: peak E/e′ ratio ≤15 or >15. For this analysis, we used the latest E/e′ ratio that could be measured without fusion between 20-W and peak exercise in 54 patients without obtainable E/e′ ratio during peak exercise. Patients with E/e′peak >15 (n = 38) had a higher E/e′rest than those with E/e′peak ≤15 (n = 177) (11.2 ± 2.1 vs 9.3 ± 2.2; P < 0.0001) (Figure 2). Although the E/e′ ratio increased significantly in both groups during exercise, the increase was more prominent in patients with E/e′peak >15 than in those with E/e′peak ≤15. Notably, the mean value of E/e′20W was >15 in patients with E/e′peak >15 (mean: 16.3 ± 3.5). By case definition, E/e′peak was >15 in the patients, but the change in the E/e′ ratio from 20-W exercise to peak exercise was not significant (16.3 ± 3.5 and 17.7 ± 2.8; P = 0.12). The E/e′ ratio decreased in the early recovery phase (30 seconds after exercise) in patients with E/e′peak >15. Among the patients with E/e′peak >15, E/e′recovery was ≤15 in 17 patients (44.7%), whereas it could not be obtained in 3 patients (7.9%) because of E-A fusion. Compared with patients with E/e′peak >15, the E/e′ ratio remained lower in those with E/e′peak ≤15 in each exercise stage (all P < 0.0001). Sensitivity analysis performed using a cutoff value of E/e′peak >11 among 16 patients with AF rhythm at echocardiographic examination showed similar results to the primary analyses (Supplemental Figure 1).9 In patients with low HFA-PEFF score (1 point), none developed E/e′peak >15 (Supplemental Figure 2).
Feasibility and Changes in the E/e′ Ratio Throughout Exercise According to the Peak E/e′ Ratio
The feasibility of E/e′ during the 20-W exercise (E/e′20W) was high (96.3%), but that of E/e′ ratio during peak exercise (E/e′peak) decreased to 74.9%. Compared with patients with E/e′peak ≤15 (n = 177; green), the resting E/e′ ratio was higher in patients with E/e′peak >15 (n = 38; red) (9.3 ± 2.2 vs 11.2 ± 2.1; P < 0.0001). An abnormal increase in the E/e′ ratio occurred during the 20-W exercise, but the E/e′ ratio remained unchanged from 20-W exercise to peak exercise in the patients with E/e′peak >15 (16.3 ± 3.5 and 17.7 ± 2.8; P = 0.12). There was a significant reduction in the E/e′ ratio from peak exercise to recovery period (15.8 ± 3.5; P = 0.01). Values are presented as means and 95% CIs. ∗P < 0.001 vs baseline. PLR = passive leg raise; Rec = recovery.
Diagnostic performance of E/e′ ratio during the 20-W exercise
We then determined whether E/e′20W can predict normal E/e′ ratio during peak exercise (E/e′peak ≤15). In patients with E-A fusion during peak exercise, the latest E/e′ ratio that could be obtained between 40-W and peak exercise was used. We excluded 25 patients whose E/e′ ratio could not be obtained because of E-A fusion during the ≥40-W exercise. The heart rate during peak exercise was higher in patients with E-A fusion than in those with obtainable E/e′ ratio (130 ± 19 beats/min vs 113 ± 20 beats/min; P = 0.0001). Of the 190 patients eligible for this analysis, 35 (18%) had E/e′peak >15. E/e′rest could identify patients with E/e′peak ≤15 (AUC: 0.76; P < 0.0001) with the optimal cutoff value of ≤10.4 showing modest sensitivity (75%) and specificity (74%) (Figure 3A). E/e′20W demonstrated excellent diagnostic performance in identifying E/e′peak ≤15 (AUC: 0.91; P < 0.0001) with the optimal cutoff value of ≤12.4 showing excellent specificity (94%) and positive predictive value (PPV) (98%) but modest sensitivity (77%) and negative predictive value (NPV; 49%) (Figure 3B). A similarly high AUC was obtained for E/e′recovery (AUC: 0.91; P < 0.0001) (Figure 3C). Diagnostic performance using other cutoff values of septal E/e′ are listed in Supplemental Table 1. Actual numbers of patients according to E/e′peak are provided in Supplemental Table 2. After excluding patients with AF, the results remained similar to the primary analyses (Supplemental Figure 3). LA reservoir strain at rest and during exercise could identify E/e′peak ≤15, but their diagnostic abilities were inferior to that of E/e′20W (Supplemental Table 3). E/e′20W >12.4 was associated with poorer exercise capacity, but E/e′peak >15 was not (Supplemental Table 4).
Diagnostic Ability of E/e′ Ratio in Predicting the Peak E/e′ Ratio to Be ≤15
(A) Resting E/e′ ratio helped identify patients with E/e′peak ≤15, with the cutoff value of ≤10.4 showing modest sensitivity and specificity. (B) E/e′20W demonstrated excellent diagnostic ability to identify E/e′peak ≤15, with the cutoff value of ≤12.4 providing high specificity. (C) The diagnostic ability of recovery E/e′ ratio was similar to that of E/e′20W. AUC = area under the receiver-operating characteristic curve.
Protocol 2
Patient characteristics
Exercise RHC was performed for a median of 41 days (21-63 days) after the exercise stress echocardiography. Of the 45 patients, there were 27 patients with HFpEF and 18 control subjects. The overall patient characteristics in protocol 2 were similar to those in protocol 1, except for a higher prevalence of women (Table 2).
| Age, y | 71 ± 10 |
| Female | 34 (76) |
| Body mass index, kg/m2 | 23 ± 4 |
| HFA-PEFF score | 4 (3-5) |
| Comorbidities | |
| Hypertension | 37 (82) |
| Coronary artery disease | 5 (11) |
| Atrial fibrillation | 3 (7) |
| Diabetes mellitus | 8 (18) |
| Significant mitral annular calcification | 3 (7) |
| Medications | |
| ACEI or ARB | 11 (24) |
| Beta-blocker | 6 (13) |
| Calcium channel blocker | 20 (44) |
| Diuretic | 5 (11) |
| Laboratory tests | |
| Hemoglobin, g/dL | 12.5 ± 1.5 |
| Creatinine, mg/dL | 0.8 (0.6-0.9) |
| BNP, pg/mL (n = 44) | 46 (17-105) |
| NT-proBNP, pg/mL (n = 41) | 146 (61-453) |
| Vital signs | |
| Heart rate, beats/min | 76 ± 11 |
| Systolic blood pressure, mm Hg | 128 ± 22 |
| Echocardiography | |
| LV end-diastolic volume, mL | 67 ± 19 |
| LV mass index, g/m2 | 87 ± 24 |
| LV ejection fraction, % | 65 ± 7 |
| Mitral E, cm/s | 68 ± 24 |
| Mitral e′, cm/s | 6.2 ± 2.2 |
| Mitral s′, cm/s | 7.6 ± 1.8 |
| E/e′ ratio | 12.1 ± 5.0 |
| Left atrial volume index, mL/m2 | 31 ± 11 |
| Left atrial reservoir strain, % | 30 ± 13 |
| Tricuspid regurgitant velocity, m/s | 2.4 ± 0.4 |
| TAPSE, mm | 21 ± 4 |
| Right ventricular s′, cm/s | 13 ± 3 |
Changes in PCWP during exercise
Based on the exercise RHC, 27 patients (60%) met the invasive criteria for HFpEF. The changes in the PCWP in HFpEF patients and control subjects throughout the exercise are shown in Figure 4. Patients with HFpEF had a higher PCWP at rest compared with the control subjects (11.7 ± 2.3 mm Hg vs 8.8 ± 3.7 mm Hg; P = 0.002). The increase in the PCWP was greater in patients with HFpEF than in the control subjects at peak exercise (32.0 ± 7.4 mm Hg vs 20.1 ± 3.9 mm Hg; P < 0.0001). In patients with HFpEF, most of the increase in the PCWP occurred during the 20-W exercise (accounting for 96% of peak exercise), and 25 patients (93%) developed PCWP ≥25 mm Hg in this low-level workload exercise. In the immediate recovery phase (1-min after exercise), PCWP decreased in both groups (20.9 ± 7.2 mm Hg in HFpEF vs 14.6 ± 4.0 mm Hg in control subjects; P = 0.002).
Changes in the PCWP Throughout the Exercise
At rest, PCWP was higher in patients with HFpEF than in those without heart failure. An abnormal increase in the PCWP occurred during the 20-W exercise in patients with HFpEF. The increase in the PCWP from the 20-W exercise to peak exercise was modest (P = 0.24), even in patients with HFpEF. PCWP decreased immediately in the recovery phase in both groups. Values are presented as means and 95% CIs. ∗P < 0.01 vs baseline. Abbreviations as in Figures 1 and 2.
Diagnostic performance of E/e′ ratio in predicting normal PCWP during exercise
We examined whether the E/e′ ratio during exercise stress echocardiography performed before catheterization could predict normal PCWP during exercise (<25 mm Hg). The E/e′rest could be obtained in all patients (feasibility 100%), but it did not distinguish HFpEF from control subjects (AUC: 0.71; P = 0.07) (Figure 5A). The diagnostic ability was improved during PLR (AUC: 0.72; P = 0.02), with the optimal cutoff value of ≤10.8 showing high specificity (82%) but low sensitivity (48%). E/e′20W predicted normal PCWP during exercise (AUC: 0.77; P = 0.01) (Figure 5B), with reasonably high feasibility (86.7%). Whereas E/e′20W ≤15 showed high sensitivity (81%) but poor specificity (61%), E/e′20W ≤12.4 showed high specificity (83%) and modest sensitivity (75%) to predict normal PCWP during exercise. Although E/e′peak showed the largest AUC (AUC: 0.81; P = 0.003) (Figure 5C), the feasibility of E/e′peak was 66.7%. E/e′recovery showed reasonable feasibility (86.7%) and a modest diagnostic ability in ruling out HFpEF (AUC: 0.71; P = 0.04). Actual numbers of patients according to PCWP during peak exercise are provided in Supplemental Table 5. The LA reservoir strain during peak exercise demonstrated a similar diagnostic ability to that of E/e′20W (Supplemental Table 6).
Diagnostic Ability of E/e′ Ratio at Each Stage of Exercise to Predict Normal PCWP During Exercise
The presence of HFpEF was defined by PCWP during supine ergometry exercise ≥25 mm Hg. (A) The E/e′ ratio could be obtained in all patients at rest (feasibility 100%), but it did not discriminate between HFpEF and non-HF. (B) E/e′20W predicted normal PCWP during exercise (<25 mm Hg) with the cutoff value of ≤12.4 showing high specificity. (C) Although the E/e′ ratio during peak exercise showed the largest AUC in predicting normal exercise PCWP, the feasibility of E/e′peak was low (66.7%). ∗The cutoff values were obtained in protocol 1 (E/e′rest ≤10.4 and E/e′20W ≤12.4). HF = heart failure; other abbreviations as in Figures 1 and 3.
Discussion
The diagnosis of HFpEF in euvolemic patients presenting with chronic dyspnea is challenging because many patients develop an abnormal elevation in the left-side filling pressures only during exercise. The E/e′ ratio is the central parameter in estimating LV filling pressure in exercise stress echocardiography,5,10-14 but the primary limitation is that its diagnostic value is substantially limited when the diastolic Doppler velocities are fused if the heart rate is elevated. In this study, we found that the E/e′ ratio could be obtained in the majority of the patients during the 20-W exercise (96.3%), but the feasibility decreased to 74.9% at peak exercise because of E-A fusion. The E/e′20W predicted normal peak E/e′ ratio during exercise (E/e′peak ≤15) (AUC: 0.91; P < 0.0001) with the cutoff of ≤12.4 providing excellent specificity (94%) and PPV (98%). Using an invasive exercise hemodynamic test, we further confirmed that E/e′20W predicted normal PCWP during exercise in patients with dyspnea. These data suggest that E/e′20W might serve as an alternative to the peak exercise value in ruling out HFpEF in patients with dyspnea (Central Illustration).
The Utility of E/e′ Ratio During Low-Level Exercise in Diagnosing HFpEF
In patients with E/e′peak >15, an abnormal increase in the E/e′ ratio occurred during the 20-W exercise, with high feasibility. The feasibility decreased to 74.9% at peak exercise because of E-A fusion. The E/e′20W predicted E/e′peak ≤15 and normal pulmonary capillary wedge pressure (PCWP) during exercise. AUC = area under the receiver-operating characteristic curve; HFpEF = heart failure with preserved ejection fraction; PLR = passive leg raise; Rec = recovery.
Accumulated data have demonstrated that many patients with HFpEF develop pathologic elevations in LV filling pressures only during exercise-induced stress, which makes it difficult to identify diseases exclusively with the use of resting assessments.4-7 The consensus recommendations from the Heart Failure Association of the European Society of Cardiology have proposed an algorithm that emphasizes the importance of exercise stress testing for the diagnosis of HFpEF.8,18 Invasive hemodynamic exercise testing is considered to be the criterion standard in the diagnosis of HFpEF, but its requirement for specialized training and equipment and measurable risk may limit its broad application in practice. Although diastolic stress echocardiography is widely performed, evidence to support this practice remains limited and relies on expert consensus opinion.
The E/e′ ratio is the most important parameter for estimating LV filling pressure at rest and during exercise,5 and contemporary guidelines have proposed its use in identifying HFpEF during exercise stress echocardiography.9,17,18 However, the primary limitation is the inability to assess the E/e′ ratio when diastolic Doppler velocities are fused. In this study, we observed that the E/e′ ratio could not be measured in 25% of the patients during peak exercise because of E-A fusion, which is similar to a previous study.5 The European Association of Cardiovascular Imaging and the American Society of Echocardiography recommend the acquisition of the E/e′ ratio at a heart rate of 100-110 beats/min to avoid fusion, but this has not been rigorously tested.17,19 Previous invasive studies have shown that abnormal increases in the PCWP occur early during 20-W exercises in HFpEF.4,5,7,20,21 The diagnostic criteria of HFpEF through exercise testing are not limited to peak exercise, and the observation that hemodynamic abnormalities are apparent during low-level exercise is relevant for its application in exercise echocardiography.
We found that the E/e′ ratio could be obtained in the majority of the patients (96.3%) during the 20-W exercise. Although the mean value of E/e′ ratio was >15 (16.3 ± 3.5) during the 20-W exercise in patients with E/e′peak >15, it showed a limited increase during peak exercise (mean: 17.7 ± 2.8). These findings were invasively validated in protocol 2, wherein most of the increase in the PCWP occurred during the 20-W exercise (accounting for 96% of peak exercise), and 25 patients with HFpEF (93%) developed PCWP ≥25 mm Hg during that low-level workload exercise. We extended the data and found that E/e′20W ≤12.4 accurately predicted normal E/e′ ratio during peak exercise (E/e′peak ≤15) in protocol 1 and normal PCWP during exercise (<25 mm Hg) in protocol 2. These data suggest that E/e′20W can be a suitable alternative to the peak exercise value for ruling out HFpEF.
Post-exercise assessment during the recovery stage may be performed to obtain the E/e′ ratio. Some recommendations propose image acquisition during the first 2 minutes of the recovery phase when the mitral E and A velocities are no longer fused.17-19 This is based on the speculation that LV filling pressures remain elevated during this period. However, the PCWP may immediately return to the baseline value in the recovery phase in HFpEF (1 minute after exercise),4 which is consistent with the current results (Figure 4). Although E/e′recovery presented a modest discriminative ability to diagnose HFpEF (AUC: 0.71; P = 0.04) in protocol 2, its use may require caution. Because E/e′recovery was recorded 30 seconds after the exercise, further studies are warranted to examine the diagnostic utility of the post-exercise E/e′ ratio using the E-A at the time of separation. On the other hand, E/e′PLR showed a similar diagnostic accuracy to E/e′recovery in protocol 2 (AUC: 0.72; P = 0.02). E/e′PLR ≤10.8 may be useful to rule out HFpEF, which is in agreement with a recent invasive study.26
Clinical implications
From a diagnostic perspective, it is vital to consider the sensitivity, specificity, PPV, and NPV of the exercise E/e′ ratio to distinguish HFpEF from control subjects rather than using a diagnostic value alone. E/e′20W ≤12.4 demonstrated high specificity (94%) and PPV (98%) for predicting normal E/e′peak in protocol 1. The observation was invasively confirmed in protocol 2, where E/e′20W ≤12.4 provided excellent diagnostic performance to predict normal PCWP during exercise. The previous invasive study by Obokata et al5 reported that normal E/e′ ratio during peak exercise (≤15) was useful to rule out HFpEF.5 In addition to high feasibility, the observed high specificity of E/e′20W to predict normal E/e′peak suggests that E/e′20W could be an alternative to the peak exercise value in ruling out HFpEF among patients with dyspnea. The low-level exercise (20 W) is almost equivalent to the usual daily activities in which patients might have shortness of breath, and this may be one of the potential reasons for the reasonable diagnostic performance of E/e′20W.
Study limitations
This was a single-center study from a tertiary referral center, which may have drawbacks related to selection and referral bias. Our study design required the exclusion of patients who did not exercise more than or equal to 40 W, which might have biased the overall results. The sample size of protocol 2 was small, and this limited additional subgroup analyses. However, the inclusion of invasive exercise hemodynamic testing, which is considered to be the criterion standard in diagnosing HFpEF, was the strength of this study. The feasibility of E/e′ ratio during 20-W or peak exercise was numerically lower in protocol 1 than in protocol 2. This might be related to the fact that cases with equivocal exercise echocardiographic tests were more likely to be referred for invasive exercise hemodynamic testing. Participants underwent supine ergometry exercise. The exercise position might have influenced the overall results, and our findings may not apply to exercises performed in a different posture, such as upright exercises. The image acquisition of E/e′ during the recovery period was performed 30 seconds after the cessation of exercise, and the E and A waves remained fused in some participants. In protocol 2, echocardiography was not performed simultaneously with cardiac catheterization. The AUC was determined only in participants in whom the E/e′ ratio could be obtained. The peak exercise workload was different between protocols 1 and 2 for some patients. The current results in protocol 2 and others suggest the potential diagnostic value of exercise LA reservoir strain in HFpEF.27 This should be investigated in a larger cohort using invasive exercise hemodynamic testing.
Conclusions
The E/e′ ratio during the 20-W exercise was highly feasible and predicted E/e′ ratio during peak exercise to be ≤15, which indicates normal LV filling pressure. In addition, E/e′20W ≤12.4 helped to predict normal PCWP during exercise, but the diagnostic value of E/e′peak was limited because of low feasibility. These data suggest that E/e′20W might serve as an alternative to the peak exercise value in ruling out HFpEF in patients with dyspnea.
Perspectives
COMPETENCY IN MEDICAL KNOWLEDGE: The E/e′ ratio during a low-level (20 W) workload exercise (E/e′20W) is feasible and can predict normal invasive and noninvasive estimates of LV filling pressure during peak exercise. E/e′20W may be used as an alternative to the peak exercise value to rule out HFpEF in patients with dyspnea.
TRANSLATIONAL OUTLOOK: This study focused on exercise E/e′ ratio rather than the combination of other parameters, such as e′ velocity or tricuspid regurgitant velocity. Further studies are warranted to determine the optimal approach in exercise stress echocardiography for the diagnosis of HFpEF.
Funding Support and Author Disclosures
Dr Harada has received research grants from the Bayer Academic Support. Dr Obokata has received research grants from the Fukuda Foundation for Medical Technology, Mochida Memorial Foundation for Medical and Pharmaceutical Research, Nippon Shinyaku, Takeda Science Foundation, Japanese Circulation Society, and the Japanese College of Cardiology. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Abbreviations and Acronyms
| AF | atrial fibrillation |
| E/e′ ratio | ratio of early diastolic mitral inflow velocity to early diastolic mitral annular tissue velocity |
| E/e′20W | E/e′ ratio during 20-W exercise |
| HF | heart failure |
| HFpEF | heart failure with preserved ejection fraction |
| LA | left atrial |
| LV | left ventricular |
| PCWP | pulmonary capillary wedge pressure |
| PLR | passive leg raise |
| RHC | right heart catheterization |
| TR | tricuspid regurgitation |
References
1. "Evaluation and management of heart failure with preserved ejection fraction". Nat Rev Cardiol .2020;17:559-573.
2. "Echocardiography in the diagnostic evaluation and phenotyping of heart failure with preserved ejection fraction". J Cardiol .2022;79:679-690.
3. "Prior heart failure hospitalization and outcomes in patients with heart failure with preserved and reduced ejection fraction". Am J Med .2019;133:84-94.
4. "Exercise hemodynamics enhance diagnosis of early heart failure with preserved ejection fraction". Circ Heart Fail .2010;3:588-595.
5. "Role of diastolic stress testing in the evaluation for heart failure with preserved ejection fraction: a simultaneous invasive-echocardiographic study". Circulation .2017;135:825-838.
6. "A simple, evidence-based approach to help guide diagnosis of heart failure with preserved ejection fraction". Circulation .2018;138:861-870.
7. "Haemodynamics, dyspnoea, and pulmonary reserve in heart failure with preserved ejection fraction". Eur Heart J .2018;39:2810-2821.
8. "Multimodality imaging in patients with heart failure and preserved ejection fraction: an expert consensus document of the European Association of Cardiovascular Imaging". Eur Heart J Cardiovasc Imaging .2022;23:e34-e61.
9. "Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging". J Am Soc Echocardiogr .2016;29:277-314.
10. "The strengths and limitations of E/e′ in heart failure with preserved ejection fraction". Eur J Heart Fail .2018;20:1312-1314.
11. "Diastolic dysfunction and heart failure with preserved ejection fraction: understanding mechanisms by using noninvasive methods". J Am Coll Cardiol Img .2020;13:245-257.
12. "Estimation of left ventricular filling pressure with exercise by doppler echocardiography in patients with normal systolic function: a simultaneous echocardiographic-cardiac catheterization study". J Am Soc Echocardiogr .2007;20:477-479.
13. "Hemodynamic basis of exercise limitation in patients with heart failure and normal ejection fraction". J Am Coll Cardiol .2010;56:855-863.
14. "Diastolic stress echocardiography: hemodynamic validation and clinical significance of estimation of ventricular filling pressure with exercise". J Am Coll Cardiol .2006;47:1891-1900.
15. "Effects of treadmill exercise on mitral inflow and annular velocities in healthy adults". Am J Cardiol .2003;91:114-115.
16. "Diastolic stress echocardiography: a novel noninvasive diagnostic test for diastolic dysfunction using supine bicycle exercise Doppler echocardiography". J Am Soc Echocardiogr .2005;18:63-68.
17. "The clinical use of stress echocardiography in non-ischaemic heart disease: recommendations from the European Association of Cardiovascular Imaging and the American Society of Echocardiography". J Am Soc Echocardiogr .2017;30:101-138.
18. "How to diagnose heart failure with preserved ejection fraction: the HFA-PEFF diagnostic algorithm: a consensus recommendation from the Heart Failure Association (HFA) of the European Society of Cardiology (ESC)". Eur Heart J .2019;40:3297-3317.
19. "A systematic review of diastolic stress tests in heart failure with preserved ejection fraction, with proposals from the EU-FP7 MEDIA study group". Eur J Heart Fail .2014;16:1345-1361.
20. "A transcatheter intracardiac shunt device for heart failure with preserved ejection fraction (REDUCE LAP-HF): a multicentre, open-label, single-arm, phase 1 trial". Lancet .2016;387:1298-1304.
21. "Low level exercise echocardiography helps diagnose early stage heart failure with preserved ejection fraction: a study of echocardiography versus catheterization". Clin Res Cardiol .2017;106:192-201.
22. "Association between lung ultrasound B-lines and exercise-induced pulmonary hypertension in patients with connective tissue disease". Echocardiography .2021;38:1297-1306.
23. "Pathophysiological and diagnostic importance of fatty acid-binding protein 1 in heart failure with preserved ejection fraction". Sci Rep .2021;11:21175.
24. "The H2FPEF and HFA-PEFF algorithms for predicting exercise intolerance and abnormal hemodynamics in heart failure with preserved ejection fraction". Sci Rep .2022;12:13.
25. "Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach". Biometrics .1988;44:837-845.
26. "The value of passive leg raise during right heart catheterization in diagnosing heart failure with preserved ejection fraction". Circ Heart Fail .2022;15:e008935.
27. "Exercise stress real-time cardiac magnetic resonance imaging for noninvasive characterization of heart failure with preserved ejection fraction". Circulation .2021;143:1484-1498.
Appendix
Supplemental MethodsFootnotes
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
