2019 Volume 83 Issue 2 Pages 401-409
Background: We used dual Doppler echocardiography to measure the time interval between the mitral and tricuspid valve opening (MO-TO time), which we expected would reflect the balance between left and right ventricular hemodynamics.
Methods and Results: We prospectively enrolled 60 patients with heart failure (HF) and sinus rhythm. The MO-TO time was measured in addition to routine echocardiography parameters, invasive hemodynamic parameters and plasma B-type natriuretic peptide (BNP) level in all patients. Patients were divided into 2 groups based on the MO-TO time: MOP (mitral opening preceding tricuspid opening), and TOP (tricuspid opening preceding mitral opening) groups. We followed up the predefined adverse outcomes (cardiovascular [CV] death and hospitalization due to worsening HF) for 1 year. Pulmonary artery wedge pressure (PAWP) and mean pulmonary artery pressure (mPAP) were higher in the MOP than in the TOP group (P<0.001; P<0.001, respectively). The probability of an adverse CV outcome was higher in the MOP than in the TOP group (log-rank test; P=0.002). Addition of MOP improved the predictive power of univariate predictors (mitral E/A ratio and BNP) in the bivariate Cox analysis (P=0.017, P=0.024, respectively).
Conclusions: MOP reflects pulmonary hypertension caused by left heart disease and has prognostic value in predicting adverse CV events in patients with HF.
Heart failure (HF) is a major and growing problem with high mortality and morbidity in many countries.1,2 Risk stratification of patients with HF is important to reduce deaths and hospitalization for HF, and a simple echocardiographic method has been assessed for this purpose. Initially, mitral flow parameters were considered useful to assess the prognosis of HF; however, it is not necessarily easy to predict the outcome using any single parameter, particularly because the determinants of the prognosis of HF are complex and multifactorial. Recent studies have shown that both the right ventricular (RV) and left ventricular (LV) function are important as determinants of prognosis in patients with HF.3,4 However, noninvasive assessment of RV and LV function in an integrated fashion has not been established.
Editorial p 279
The recent advent of dual Doppler systems has enabled simultaneous recording of transmitral and transtricuspid flow velocity patterns. Tricuspid valve opening (TO) usually precedes mitral opening (MO) in healthy subjects.5 Earlier MO may have clinical utility for the diagnosis of pulmonary hypertension (PH) caused by elevated LV filling pressure, because MO shifts forward as a result of elevated left atrial (LA) pressure and TO shifts backward as a result of delayed RV relaxation. In addition, patients with earlier MO than TO may have a poor prognosis of HF. Thus, the objective of this study was to clarify whether we can estimate PH caused by left heart disease (PH-LHD) and predict cardiovascular (CV) events in patients with HF by assessing the time interval between MO and TO (i.e., MO-TO time).
From February 2013 to February 2014 we prospectively enrolled 77 patients who were admitted to the Hospital of Hyogo College of Medicine because of worsening of HF and who underwent an invasive hemodynamic study after stabilization of the acute phase of HF. After censoring the patients who met the exclusion criteria, the study population was finally 60 patients. The exclusion criteria were severe valvular insufficiency and/or stenosis (n=3), a history of open heart surgery (n=3) or pacemaker implantation (n=1), atrial fibrillation (AF: n=9), or an intraventricular conductance disturbance (n=1) as assessed on routine standard 12-lead ECG. The severity of HF was classified according to the New York Heart Association (NYHA) functional class within 24 h before the invasive hemodynamic study. All patients received optimal medical therapy for HF, including angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), β-blockers, and diuretics. The Hyogo College of Medicine Ethical Committee approved the study protocol, and informed consent was prospectively obtained from all patients.
Hemodynamic EvaluationRight heart catheterization was performed in all patients to measure the pulmonary artery wedge pressure (PAWP), pulmonary artery pressure (PAP), RV pressure and RA pressure, cardiac output (CO), cardiac index with Fick method, and pulmonary vascular resistance [(mean PAP−PAWP)/CO×80] in a recumbent position using 7-Fr Swan-Ganz catheters. The catheter wedge position was verified fluoroscopically. Each parameter was determined as an average of the values at the end of expiration over 7 cycles. The catheter operators were blinded to the echocardiographic data.
Echocardiographic StudyAll patients underwent routine transthoracic echocardiography by experienced sonographers within 24 h before or after the catheter study. A ProSound F75 system (Hitachi-Aloka Medical Corp., Tokyo, Japan) was used with a 3.0-MHz transducer, and the patients were studied in the left lateral decubitus position. All echocardiographic data were acquired digitally for off-line analysis according to the guidelines recommended by the American Society of Echocardiography.6 The LV ejection fraction (LVEF) and LA volume were determined by the modified Simpson’s rule using 2D images. The LA volume was indexed to body surface area (BSA). We regarded HF with an LVEF ≥50% as HF with preserved EF (HFpEF), and with LVEF <50% as HF with reduced EF (HFrEF). We also measured the following parameters using pulse-wave Doppler: early-diastolic ventricular inflow velocity (E wave), end-diastolic ventricular inflow velocity (A wave), and deceleration time of E wave (DcT). The mitral annular peak early-diastolic velocity of tissue Doppler imaging (e’) was measured at the basal septal myocardial segment in the apical 4-chamber view and was used to determine the ratio of the mitral E wave to e’ (E/e’).
The pressure gradient of tricuspid regurgitation (TRPG) was calculated using the following formula: TRPG=4(velocity of TR)2. The pulmonary artery systolic pressure (PASP) was determined as the sum of the TRPG and estimated RA pressure,6 and the tricuspid annular systolic excursion (TAPSE)/PASP ratio was calculated as a parameter of RV-pulmonary circulation coupling.7 The RV areas at end-diastole and end-systole were traced and indexed by BSA (RVEDAI, RVESAI) and used to determine the RV fractional area change.8
Transmitral and transtricuspid flows were simultaneously recorded using dual Doppler echocardiography in the apical 4-chamber view at a sweeping velocity of 300 mm/s (Figure 1). We measured the time intervals from the QRS wave onset to the mitral E wave onset (QRS-Em) and the tricuspid E wave onset (QRS-Et). The MO-TO time was calculated by subtracting QRS-Em from QRS-Et and it was corrected for the R-R interval (corrected MO-TO time=MO-TO time/R-R interval×103). All patients were divided into 2 groups based on the precedence of TO or MO: in the TOP group, TO preceded MO; in the MOP group, MO preceded TO (Figure 1). Inflow and outflow patterns were simultaneously recorded using dual Doppler echocardiography to determine the LV and RV isovolumic relaxation times (IRTs). IRT was measured as the time interval between the end of ejection and the onset of the inflow wave.
(A) Determination of the MO-TO time by simultaneous recording the left and right ventricular inflow patterns using dual Doppler echocardiography. Representative recordings of (B) tricuspid opening preceding mitral opening and (C) mitral opening preceding tricuspid opening. We measured the time interval from the QRS wave onset to the mitral E wave onset (QRS-EM) and tricuspid E wave onset (QRS-ET). The MO-TO time was calculated by subtracting QRS-ET from QRS-EM. MO-TO time, time interval between mitral and tricuspid valve opening.
To assess interobserver reproducibility, 2 observers analyzed the same randomly selected tracings independently of each other. For intra-observer reproducibility, an observer analyzed the same tracings twice at separate times. The interobserver and intra-observer variability of MO-TO time was assessed in 10 randomly selected tracings, and reproducibility was estimated with the Bland-Altman method of agreement.
Measurement of Plasma B-Type Natriuretic Peptide (BNP) ConcentrationBlood samples were collected just before the hemodynamic examinations to determine the BNP concentrations in all patients. The samples were placed in tubes containing ethylenediaminetetraacetic acid. The plasma BNP concentration was measured with a chemiluminescence immunoenzymatic assay.
Clinical OutcomesWe followed all patients’ CV events for 1 year after discharge. All patients were carefully followed at the hospital’s cardiology clinic or in regional community hospitals with titration medical therapy depending on the patient’s status. Adverse CV events were predefined as CV death or rehospitalization because of worsening HF. These data were confirmed by inspection of the electronic health record. If there was no electronic health record, the data of death and admission were verified by phone contact with either the patient or the patient’s family.
Statistical AnalysisStatistical analysis was performed using JMP Pro 13.1.0 software (SAS Institute, Inc., Cary, NC, USA). All comparable data were assessed using a t test and chi-square test. Linear regression analysis was used to evaluate the correlations between the corrected MO-TO time and PAWP and between E/E’ and PAWP. We also compared the correlation coefficients of the mean PAWP between the corrected MO-TO time and E/e’ using MedCalc Version 12.3.0.0 (MedCalc Software, Ostend, Belgium). Receiver-operating characteristic (ROC) analysis was used to determine cutoff values, sensitivities, and specificities associated with adverse CV events. A P-value <0.05 in the ROC analysis provided evidence that the value could distinguish the 2 patient groups. We performed a univariate Cox proportional hazards analysis to identify the significant predictors of adverse CV events. Based on the results of this analysis, we compared bivariate Cox models to examine incremental predictive values of MOP. Changes in chi-square values were determined in the likelihood ratio. Harrell’s c-index in each model was also determined to evaluate predictive utility. We constructed Kaplan-Meier curves to report the time from performing the echocardiographic study to the development of adverse CV events.
The patients’ baseline clinical characteristics are shown in Table 1; 28 patients (47%) had HFpEF (mean EF, 69%±8.3%), and 32 patients (53%) had HFrEF (mean EF, 31%±9.1%). Beta-blockers were administered to 68% of the patients, and an ACEI or ARB was administered to 77% of the patients. There were 26 patients with MOP and 34 patients with TOP. There was no difference between the MOP and TOP groups in age, sex, comorbidities, or basal administered medications. The NYHA functional class was more advanced in the MOP than in the TOP group (P=0.009). The echocardiographic parameters of the 2 groups are compared in Table 2. Patients in the MOP group had a shorter mitral DcT, higher mitral E/A ratio, higher mitral E/e’ ratio, greater TRPG, shorter LV IRT, longer RV IRT, greater RVEDAI, and greater RVESAI than the patients in the TOP group. Hemodynamic parameters of the 2 groups are presented in Table 3. The mean PAWP, mean PAP, systolic RVP, and RAP were higher in the MOP than in the TOP group. In contrast, there were no differences between the groups in the cardiac index or pulmonary vascular resistance. We therefore considered that the MO-TO time was closely associated with PH-LHD.
| TOP (n=34) |
MOP (n=26) |
P value | |
|---|---|---|---|
| Female | 19 (59) | 13 (41) | 0.651 |
| Age (years) | 67±12 | 64±17 | 0.366 |
| BSA | 1.55±0.18 | 1.56±0.21 | 0.825 |
| Heart rate | 68±11 | 73±18 | 0.177 |
| Systolic BP | 123±24 | 118±20 | 0.391 |
| Diastolic BP | 68±14 | 73±29 | 0.420 |
| HFrEF | 16 (47) | 16 (62) | 0.264 |
| Comorbidities | |||
| Hypertension | 15 (56) | 14 (54) | 0.455 |
| Diabetes | 8 (23) | 9 (34) | 0.346 |
| Dyslipidemia | 2 (6) | 1 (4) | 0.716 |
| CKD | 9 (26) | 10 (38) | 0.324 |
| QRS width (ms) | 102±12 | 106±11 | 0.167 |
| BNP (pg/mL) | 99 [IQR 36–186] | 413 [IQR 87–733] | <0.001 |
| Diagnosis | 0.919 | ||
| Hypertensive heart disease | 23 (68) | 17 (65) | |
| Ischemic heart disease | 7 (21) | 5 (19) | |
| Cardiomyopathy | 4 (12) | 4 (15) | |
| NYHA functional class | 0.009 | ||
| I | 11 (32) | 1 (4) | |
| II | 16 (47) | 15 (58) | |
| III | 7 (21) | 8 (31) | |
| IV | 0 (0) | 2 (8) | |
| Medical treatment | |||
| β-blocker | 20 (58) | 21 (81) | 0.066 |
| ACEI/ARB | 24 (71) | 22 (85) | 0.196 |
| Aldosterone blocker | 11 (32) | 8 (31) | 0.896 |
| Loop diuretic | 12 (35) | 10 (38) | 0.801 |
Values are presented as n (%), mean±standard deviation, or median with interquartile range (IQR). ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BNP, B-type natriuretic peptide; BP, blood pressure; BSA, body surface area; CKD, chronic kidney disease; HFrEF, heart failure with reduced ejection fraction; MOP, mitral opening preceding tricuspid opening; NYHA, New York Heart Association; TOP, tricuspid opening preceding mitral opening.
| TOP (n=34) |
MOP (n=26) |
P value | |
|---|---|---|---|
| LV dimension in end-diastolic period (mm) | 54±12 | 54±13 | 0.872 |
| LV dimension in end-systolic period (mm) | 41±17 | 45±15 | 0.312 |
| Interventricular septal thickness (mm) | 8.4±1.5 | 8.7±2.2 | 0.512 |
| LV posterior wall thickness (mm) | 8.4±1.2 | 8.6±1.7 | 0.586 |
| LVMI (g/m2) | 109±45 | 115±38 | 0.598 |
| RWT | 0.33±0.08 | 0.33±0.11 | 0.917 |
| LVEF (%) | 52±20 | 45±22 | 0.243 |
| LAVI (mL/m2) | 40±13 | 46±19 | 0.129 |
| Mitral E (cm/s) | 64±19 | 85±26 | <0.001 |
| Mitral A (cm/s) | 75±18 | 61±25 | 0.014 |
| Mitral DcT (ms) | 222±61 | 177±75 | 0.012 |
| Mitral E/A | 0.88±0.29 | 1.61±0.74 | <0.001 |
| e’ (cm/s) | 4.9±1.69 | 5.2±1.79 | 0.568 |
| Mitral E/e’ | 14.2±5.9 | 18.0±7.8 | 0.035 |
| Tricuspid E (cm/s) | 42±12 | 43±13 | 0.906 |
| Tricuspid A (cm/s) | 43±14 | 44±17 | 0.769 |
| Tricuspid DcT (ms) | 153±58 | 132±54 | 0.189 |
| Tricuspid E/A | 1.05±0.36 | 1.11±0.57 | 0.634 |
| TRPG (mmHg) | 26±5.9 | 33±10 | 0.003 |
| TAPSE (mm) | 18±4.9 | 19±6.3 | 0.778 |
| TAPSE/PASP | 0.54±0.15 | 0.46±0.16 | 0.105 |
| S’ (mm/s) | 10.2±3.3 | 10.5±2.7 | 0.767 |
| RVEDAI (cm2/m2) | 12.2±2.3 | 13.7±2.5 | 0.034 |
| RVESAI (cm2/m2) | 7.9±2.1 | 9.4±2.1 | 0.012 |
| RVFAC (%) | 36.1±10.0 | 31.6±6.3 | 0.064 |
| IVC (mm) | 14±3.7 | 15±4.3 | 0.206 |
| LV IRT (ms) | 104±35 | 73±26 | 0.003 |
| RV IRT (ms) | 79±35 | 104±41 | 0.041 |
| QRS-EM (ms) | 455±63 | 482±44 | 0.060 |
| QRS-ET (ms) | 496±62 | 459±48 | 0.014 |
| MO-TO time (ms) | 27±18 | −43±35 | <0.001 |
| Corrected MO-TO time (ms) | 30±19 | −52±45 | <0.001 |
Values are presented as n (%) or mean±standard deviation. A, end-diastolic ventricular inflow velocity; BSA, body surface area; DcT, deceleration time of E wave; E, early-diastolic ventricular inflow velocity; E’, early-diastolic mitral annular velocity; IRT, isovolumic relaxation time; IVC, inferior vena cava; LA, left atrial; LAVI, BSA-indexed LA volume; LV, left ventricular; LVEF, LV ejection fraction; MO-TO time, time interval between mitral and tricuspid valve opening; PASP, pulmonary artery systolic pressure; QRS-EM, time interval from QRS wave onset to mitral E wave onset; QRS-ET, time interval from QRS wave onset to tricuspid E wave onset; RVEDAI, BSA-indexed RV end-diastolic area; RVESAI, BSA-indexed RV end-systolic area; RV, right ventricular; RVFAC, RV fractional area change; RWT, relative wall thickness; S’, tricuspid lateral annular systolic velocity wave; TAPSE, tricuspid annular plane systolic excursion; TRGP, pressure gradient of tricuspid regurgitation.
| TOP (n=34) |
MOP (n=26) |
P value | |
|---|---|---|---|
| PAWP (mmHg) | 11±4.5 | 21±8.5 | <0.001 |
| Systolic PAP (mmHg) | 33±9.3 | 48±12 | <0.001 |
| Diastolic PAP (mmHg) | 13±4.1 | 22±6.5 | <0.001 |
| Mean PAP (mmHg) | 21±5.5 | 32±8.8 | <0.001 |
| Systolic RVP (mmHg) | 34±8.4 | 47±12 | <0.001 |
| RV end-diastolic pressure (mmHg) | 8.4±3.3 | 13±4.6 | <0.001 |
| Mean RAP (mmHg) | 5.8±2.8 | 9.3±4.9 | 0.001 |
| Cardiac index (L/min/m2) | 3.05±0.72 | 2.75±0.85 | 0.153 |
| PVR (dynes*s−1*cm−5) | 180±95 | 227±134 | 0.114 |
PAP, pulmonary artery pressure; PAWP, pulmonary artery wedge pressure; RVP, right ventricular pressure; RVEDP, right ventricular pressure in end-diastolic period; RAP, right atrial pressure; PVR, pulmonary vascular resistance.
We compared the corrected MO-TO time with the PAWP and mean PAP. The corrected MO-TO time inversely correlated with the PAWP (r=0.74, P<0.001) and mean PAP (r=0.70, P<0.001) (Figure 2). PAWP positively correlated with the E/e’ ratio as a conventional indicator of PAWP in this cohort (r=0.52, P<0.001). Thus, the corrected MO-TO time more closely correlated with PAWP than with the E/e’ ratio (P=0.046).
Comparison between corrected MO-TO time and PAWP and mean PAP. Results of the correlation analysis between the corrected MO-TO time and (A) PAWP and (B) PAP. MO-TO time, time interval between mitral and tricuspid valve opening; MOP, mitral opening preceding tricuspid opening; PAP, pulmonary artery pressure; PAWP, pulmonary artery wedge pressure; TOP, tricuspid opening preceding mitral opening.
Furthermore, we studied the correlations of the corrected MO-TO time with the PAWP, mean PAP, correlation of PAWP, and E/e’ ratio. In both the patients with HFpEF and those with HFrEF, the corrected MO-TO time inversely correlated with the PAWP and mean PAP (Supplementary Figure 1).
Detection of PH-LHD Using Echocardiographic ParametersWe performed a chi-square test based on whether the patients had PH-LHD (PAWP ≥15 mmHg and mean PAP ≥25 mmHg) and whether the patients had MOP or TOP. Of the 60 patients, there were 21 patients with PH-LHD. MOP had power to detect PH-LHD (chi-square=25.0, P>0.001). However, conventional echocardiographic parameters (E/e’ ratio ≥15 and TRPG ≥34 mmHg) were not able to stratify PH-LHD (chi-square=2.0, P=0.16). The MO-TO time may be a more useful marker for the detection of PH-LHD than the conventional echocardiographic parameters.
MOP as a Predictive Factor for Unfavourable CV EventsA total of 14 (23%) patients developed unfavourable CV events over 1 year (hospitalization for worsening HF, n=12; CV death, n=2). ROC analysis showed that the corrected MO-TO time with a discriminating value of ≥MO preceding by 4.5 ms (MO-TO time ≤−4.5 ms) [sensitivity, 0.70; specificity, 0.79; area under the curve (AUC), 0.72; 95% confidence interval (CI), 0.59–0.83; P=0.013]. However, PAWP and E/e’ were not associated with unfavourable CV events over 1 year (PAWP: AUC, 0.62; 95% CI, 0.49–0.74; P=0.21 and E/e’: AUC, 0.53; 95% CI, 0.40–0.67; P=0.69). Patients with HF with a corrected MO-TO time ≤−4.5 ms had worse outcomes than those with a corrected MO-TO time >−4.5 ms (58% vs. 89%, log-rank P=0.005). We also analyzed Kaplan-Meier curves using discrimination of an MO-TO time of ≤0 ms (MOP or TOP), which is simpler than using an MO-TO time of >−4.5 ms. Patients with HF and MOP had worse outcomes than those with TOP (57% vs. 91%, log-rank P=0.002) (Figure 3A). Although E/e’ is known to predict increased LV filling pressure,9 E/e’ ≥15 was not associated with an increased number of unfavourable CV events within 1 year (Figure 3B). Univariate analysis showed that the mitral E/A, plasma BNP concentration, and MO-TO time were associated with new-onset adverse CV events (Table 4). In the bivariate Cox models, predictive values were improved by the addition of MOP, both for the mitral E/A ratio and BNP (P=0.017, P=0.024, respectively). The improvement was also seen for the Harrell’s c-index (Figure 4).
Cumulative survival free from adverse cardiovascular events. Kaplan-Meier plots show the 1-year rate of survival free from adverse cardiovascular events (A) between MOP vs. TOP and (B) between an E/e’ ≥15 vs. <15. MOP, mitral valve opening preceding tricuspid valve opening; TOP, tricuspid valve opening preceding mitral valve opening.
| Variables | Univariate analysis | ||
|---|---|---|---|
| HR | 95% CI | P value | |
| Mitral E | 1.002 | 0.980–1.021 | 0.832 |
| Mitral A | 0.984 | 0.960–1.008 | 0.194 |
| Mitral E/A | 2.168 | 1.023–4.270 | 0.044 |
| Mitral deceleration time | 0.995 | 0.985–1.003 | 0.269 |
| Mitral E/e’ | 1.007 | 0.929–1.073 | 0.840 |
| LVEF | 0.984 | 0.957–1.010 | 0.231 |
| TRPG | 1.021 | 0.958–1.074 | 0.488 |
| BNP | 3.126 | 1.206–8.345 | 0.019 |
| PAWP | 1.042 | 0.979–1.099 | 0.185 |
| Systolic PAP | 1.030 | 0.992–1.065 | 0.118 |
| Diastolic PAP | 1.064 | 0.982–1.150 | 0.126 |
| Mean PAP | 1.038 | 0.981–1.089 | 0.184 |
| Systolic RVP | 1.013 | 0.969–1.051 | 0.551 |
| RV end-diastolic pressure | 0.983 | 0.862–1.111 | 0.792 |
| RAP | 0.988 | 0.854–1.116 | 0.854 |
| Corrected MO-TO time | 0.989 | 0.983–0.997 | 0.012 |
Abbreviations as in Tables 1–3.
Nested-model comparisons for predicting adverse cardiovascular outcomes. Addition of MOP to routine echocardiographic and clinical parameters with statistical significance for predicting adverse cardiovascular outcomes in the univariate analysis (E/A ratio and BNP). The global chi-square and Harrell’s c-indexes were improved by the addition of MOP. BNP, B-type natriuretic peptide; MOP, mitral valve opening preceding tricuspid valve opening.
The intra- and interobserver reproducibility assessed by the Bland-Altman plot of agreement showed good agreement between the 2 measurements. The 1.96 SD value for the difference was 5.9 and −9.7 ms for intra-observer reproducibility and 12.0 and −12.0 ms for interobserver reproducibility (Supplementary Figure 2).
In this study, we assessed the MO-TO time using a dual Doppler system to determine whether this time interval is useful for assessing PH-LHD and the prognosis in patients with HF regardless of their LVEF. To the best of our knowledge, this is the first study to assess the clinical significance of the MO-TO time using a dual Doppler system.
Interventricular Inflow Delay Reflects Biventricular HemodynamicsWe hypothesized that the MO-TO time would reflect both LV and RV hemodynamic conditions. When the LA pressure increases, the LV IRT may well be shortened.10 In contrast, the RV IRT may well be prolonged under conditions of RV systolic pressure elevation.11–13 The RA pressure was higher in the MOP than in the TOP group. An increase in RA pressure is certainly likely to work toward shifting the TO forward and hence shorten the RV IRT. Nevertheless, the RV IRT was prolonged in our patients in the MOP group. The effect of RV systolic pressure elevation may have overcome the effect of the RA pressure increase in our patients; however, the RV IRT may be shortened in patients with markedly elevated RA pressure. Additional studies are necessary to clarify the detailed effects of an increase in RA pressure.
Correlation of Interventricular Inflow Delay With PAWP and Plasma BNPTO precedes MO in the healthy heart,5 whereas MO precedes TO in the heart with an elevated LV filling pressure. The precedence of MO basically reflects an elevated LV filling pressure and is likely to be exaggerated by the association of PH. Therefore, the MO-TO time may be used as a simple echocardiographic marker of HF.
The MO-TO time positively correlated with elevation of the LV filling pressure, and in this study the correlation was better than that between E/e’ or other measures and LV filling pressure. We have no data to explain why; on the other hand, there is a wide range of values of PAWP and PAP for any given MO-TO time, and thus, we can not use MO-TO time for estimating these pressures. Various conventional echocardiographic parameters enable estimation of the LV end-diastolic pressure. For example, E/e’ is reportedly a useful parameter for estimation of the LV filling pressure,9 but its use can be limited in patients with severe LV dysfunction, significant mitral regurgitation, coronary artery disease, or constrictive pericarditis.14–17 We think the MO-TO time will be 1 such parameter because inclusion of this parameter improved the value of Doppler parameters in the noninvasive assessment of hemodynamics as well as prognosis. We also understand future studies are necessary to confirm the value of the MO-TO time in the management of patients with HF. Cardiomyocytes secrete BNP if stretched; thus, the BNP concentration reflects the degree of elevation of the LV end-diastolic pressure in patients with sinus rhythm.18 The plasma BNP concentration was higher in patients with a shorter MO-TO time; however, the MO-TO time correlated better with PAWP than with BNP in this study. It is well-known that the BNP concentration reflects not only an increase in LV pressure but also other factors such as LV hypertrophy and geometry.
It seemed that the MO-TO time might be useful for assessing PAP, even in patients with primary PH, but this is unlikely because the beginning of LV filling does not shift forward, while the beginning of RV filling shifts backward, reflecting the elevation of RV pressure in such patients. In other words, an increase in LA pressure is very important for the beginning of LV filling to shift forward, and LA pressure is usually normal or even low in patients with primary PH.
In terms of the issue about observer variability, one might consider the variability was larger than the value of the threshold. It would be a serious problem if the MO-TO time always took a positive value, or ranged from 0 to 50 or 100 ms or so. However, that was not the case, and the MO-TO time ranged more widely. In addition, the variability of 12 ms is comparable to the value reported as the variability of IRT.19
Clinical ImplicationsPH is associated with severe symptoms and a poor prognosis in patients with chronic HF.3,20,21 Elevation of the LA pressure because of left heart disease is the main cause of PH in many patients, and an invasive assessment using right heart catheterization is necessary to confirm this type of PH. A persistently high residual LV filling pressure after a medical intervention in patients with HF suggests an adverse prognosis.22–24 In such patients, it is beneficial to frequently measure the right heart hemodynamics, but repeated invasive measurement of hemodynamics is not desirable in the clinical setting. Conventional Doppler echocardiography provides indexes that reflect PH or elevated LA pressure. E/e’ is the most widely used of such indexes, but has many limitations; in fact, the results were suboptimal in this study. The E/A ratio is another well-known predictor of HF, but has inherent problems such as fusion of the E and A waves, effects of significant mitral regurgitation, and others.25 Although the PASP is a recently well-established parameter of PH-LHD, it is calculated by adding the estimated RAP to the TRPG. The TAPSE/PASP ratio was proposed to predict PH-LHD, but still has some limitations.7,26 Thus, any of these parameters has significant limitations. The MO-TO time is an integrated parameter that reflects elevation of the LV filling pressure and PH more accurately than the combination of 2 other indexes (TRPG ≥34 mmHg and E/e’ ≥15). The corrected MO-TO time can be measured simply and noninvasively; thus, it may be useful for the assessment of HF, although a dual Doppler system is necessary.
Study LimitationsFirst, the number of enrolled patients was small, and this was a single-centre study. The number of the adverse CV events was too small to draw a conclusion; thus, future studies are needed to confirm the value of the MO-TO time in the prediction of prognosis. Second, we defined the start of ventricular filling as the instance of valve opening, at least partially, because some of the dual Doppler data lacked a valve click signal. Whether the valve click Doppler signal or the start of ventricular filling is a better signal of valve opening remains unknown, but we believe that the difference is subtle. Third, only patients with sinus rhythm were studied, whereas our preliminary study suggested that MOP is associated with elevated PAWP even in patients with AF. Evaluation of additional patients with AF is needed to establish the value of this method in such patients. Fourth, in this study we selected patients with HF and normal intraventricular conduction. When HF worsens, LV electrical conductance may newly emerge, and patients may receive implantable devices. The effect of an intraventricular conduction delay on the MO-TO time should be studied in the future. Fifth, strain data were not collected because this study focused on time phases. The frame rate of images often requires 40–70 Hz for strain analysis.27 In other words, frame-by-frame time differences equal to 14–25 ms, and much higher time resolution, is desirable for our purpose. Sixth, we did not measure temporal variability in patients with HF because there is no established method of confirming if the biological variability is insignificant in each patient.
The MO-TO time correlated with hemodynamic parameters, and MO reflected concomitant elevation of the LV filling pressure and PAP. The MO-TO time may be used to predict clinical outcomes in patients with HF.
We thank Ms. Masumi Tanaka, Ms. Sachiko Makihara, Ms. Chika Misumi, Ms. Kumiko Matsunaga, Mr. Taiki Maki, Ms. Miho Makihara, and Ms. Manami Hiramoto for helping carry out this study.
None.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-18-0999
