Evaluation of left ventricular filling pressure by echocardiography: incremental diagnostic information from left atrial strain

Funding Acknowledgements

Type of funding sources: Public Institution(s). Main funding source(s): South-Eastern Norway Regional Health Authority

Background

Elevated left ventricular (LV) filling pressure is an important diagnostic feature of heart failure.

Objectives

To investigate determinants of left atrial (LA) reservoir and pump strain and if these parameters may serve as markers of LV filling pressure.

Methods

In a multicenter study of 322 patients with cardiovascular disease of different etiologies, LA strain by speckle tracking echocardiography was compared to conventional echocardiographic markers using invasive pressure as reference.

Results

Left ventricular filling pressure correlated well with LA reservoir and pump strain (r-values

‑0.52 and ‑0.57, respectively) (Figure). However, LV global longitudinal strain (GLS) was the strongest determinant of LA reservoir strain (r = 0.64), and correlated well with LA pump strain (r = 0.51).  For both LA strains, association with filling pressure was strongest in patients with reduced LV ejection fraction. In patients with normal GLS (≥18%), atrial strains provided no information regarding filling pressure (Figure). Reservoir strain <18% and pump strain <8% predicted elevated LV filling pressure better (p < 0.05) than the conventional indices LA volume, ratio of mitral early filling velocity/annular velocity and tricuspid regurgitation velocity. Accuracy to classify filling pressure as normal or elevated was 75% for both LA strains . When any one of the conventional indices were missing, and were replaced by LA strains, the combination of indices had accuracy 82% to correctly classify filling pressure.

Conclusions

Left atrial reservoir and pump strain may serve as clinical markers of LV filling pressure, but will be useful predominantly in patients with reduced systolic function.  Due to limited diagnostic accuracy, LA strain should be used in combination with other indices.

Abstract Figure

Assessment of left ventricular filling pressure: left atrial reservoir strain is an excellent replacement for missing tricuspid regurgitation velocity

Background

When evaluating left ventricular filling pressure (LVFP) according to current guidelines, tricuspid regurgitation (TR) velocity is often not available.

Purpose

In the present study we investigate if left atrial (LA) reservoir strain may be used instead of TR velocity for evaluation of LVFP.

Methods

We performed a prospective, multicenter, multinational and multivendor study in an all comer population of 322 patients with suspected heart failure or other cardiovascular disease where LVFP was measured by right- or left heart catheterization, as pulmonary capillary wedge pressure or pre-A LV diastolic pressure, respectively. Echocardiography was performed within 1 day of catheterization.

101 patients classified as special populations in the 2016 ASE/EACVI recommendations (i.e. non-cardiac pulmonary hypertension, atrial fibrillation, hypertrophic and restrictive cardiomyopathies) were excluded. Of the remaining 221 patients, 118 patients had EF ≥50% and 103 patients had EF <50%. Regression analysis was performed for LA reservoir strain and TR velocity against LVFP. LA reservoir strain at a cut-off value of <18% was applied instead of TR velocity in the 2016 ASE/EACVI algorithm and compared with the current algorithm.

Results

LA reservoir strain correlated better with LVFP than TR velocity, r=0.62 vs 0.40 (p<0.01) (Figure 1). When replacing TR velocity with LA reservoir strain, the feasibility of the ASE/EACVI 2016 algorithm increased from 91.8% to 98.1%. The accuracy of the algorithm was not significantly altered (80% vs 79%).

An accuracy of 80% for the algorithm is lower than what has been reported in earlier publications, this may be due to inclusion of patients without suspected heart failure and no assessment of clinical data, which in turn may have influenced the accuracy of the algorithm.

Conclusion

LA reservoir strain has better correlation to LVFP than TR velocity, and can be used in the ASE/EACVI 2016 algorithm for estimation of LVFP as a replacement when TR velocity is missing.

Funding Acknowledgement

Type of funding source: Public Institution(s). Main funding source(s): South-Eastern Norway Regional Health Authority

Which single echo parameter is the best marker of left ventricular filling pressure?

Background

Estimation of left ventricular filling pressure (LVFP) is highly relevant in clinical practice. Invasive pressure remains the gold standard, but a number of echocardiographic parameters that correlate with LVFP are used as non-invasive markers of pressure.

Purpose

We investigated how different echocardiographic parameters correlated with invasively measured LVFP, and how accurately those parameters could differentiate between normal or elevated LVFP.

Method

We performed a prospective, multicenter, multinational and multivendor study in an all comer population of 322 patients with suspected heart failure or other cardiovascular disease. 194 patients had EF ≥50% and 129 had EF <50%. LVFP was measured by right- or left heart catheterization, as pulmonary capillary wedge pressure or pre-A LV diastolic pressure, respectively.

When excluding all special patient populations defined in the 2016 recommendations for echocardiographic evaluation of LV diastolic function, 213 patients remained. Of these 135 had EF ≥50% and 74 had EF <50%.

Echocardiography was performed within 1 day of catheterization. Previously recommended cut-off values for established parameters were used to determine the accuracy of classifying LVFP as normal or elevated. For left atrial (LA) reservoir strain, based on ROC analysis, a cut-off value of <18% was used as marker of elevated LVFP.

Results

LA reservoir strain and the ratio of peak mitral early flow velocity (E) and LA reservoir strain (E/LA strain) showed the best correlations to LVFP (Table 1, Figure 1). They also had the highest accuracy, 75% for both, in classifying LVFP as normal or elevated in the whole patient population. E/LA reservoir strain provided no additional diagnostic value to using LA reservoir strain alone.

In HFpEF patients accuracy was essentially similar for LA strain, E/LA strain and E/e', whereas in HFrEF patients the two former tended to be better than E/e'.

Conclusion

Parameters containing LA reservoir strain showed the best correlation to LVFP. This indicates that LA reservoir strain may have a role in evaluation of LVFP.

Funding Acknowledgement

Type of funding source: Public Institution(s). Main funding source(s): South-Eastern Norway Regional Health Authority

1231 Left atrial septal and lateral wall strains contain different pressure information: Utility in pulmonary hypertension

Funding Acknowledgements

South-Eastern Norway Regional Health Authority

Background

Reduced left atrial (LA) reservoir strain is a marker of elevated LA pressure. Thus it could be a potential non-invasive marker to differentiate pre- and post-capillary pulmonary hypertension (PH) as the latter is defined by elevated pulmonary capillary wedge pressure (PCWP) > 15 mmHg. However, in pre-capillary PH patients with elevated right atrial pressure (RAP), the atrial septal geometry may be abnormal. This could lead to lower regional LA septal strain, making LA lateral wall strain more accurately reflect PCWP.

Purpose

We investigated if LA lateral wall strain can differentiate between pre- and post-capillary PH, and how LA lateral wall strain and LA septal strain are both affected by elevated RAP in pre-capillary PH. Furthermore we investigated if LA septal strain can be used in pre-capillary PH patients to identify those with elevated RAP.

Methods

We analysed 63 patients with PH, 28 pre-capillary and 35 post-capillary, who underwent right heart catheterisation. Echocardiography was performed simultaneously with or within 24 hours of the invasive pressure measurements. Regional LA septal strain and lateral wall strain were measured from the apical four chamber view.

Results

Pulmonary artery pressure was 39.5 ± 11.1 mmHg (mean ± SD) in the pre-capillary PH patients and 37.0 ± 10.1 mmHg in the post-capillary PH patients (p = ns). Mean PCWP was 9.9 ± 2.5 mmHg and 24.5 ± 6.0 mmHg (p < 0.001), respectively.

LA lateral wall strain was significantly lower in patients with post-capillary PH compared to pre-capillary PH (11.9 ± 7.7% vs 26.6 ± 9.9%, p < 0.001) (Fig. a,b). At a cut-off value of 18.0%, LA lateral wall strain could predict elevated PCWP > 15 mmHg with AUC = 0.88, sensitivity = 85.7% and specificity = 76.3%.

In the 28 patients with pre-capillary PH, we classified mean RAP ≥ 10 mmHg as elevated and < 10 mmHg as normal. Seven of these patients had elevated RAP and showed significantly reduced LA septal strain compared to the 21 patients with normal RAP (13.0 ± 6.2% vs 22.1 ± 7.6%, p < 0.01). LA lateral wall strain showed no difference in these groups of pre-capillary PH patients (25.8 ± 10.1% vs 28.9 ± 9.4%) (Fig. c).

Conclusions

LA lateral wall strain can be used for differentiating between pre- and post-capillary PH. In addition, LA septal strain may be used in pre-capillary PH patients to identify those with elevated RAP.

Abstract 1231 Figure

P4368Estimation of pulmonary artery pressure from right atrial strain and tricuspid regurgitation velocity

Background

Systolic pulmonary artery pressure (SPAP) can be estimated non-invasively as the sum of indices for right atrial (RA) pressure and tricuspid regurgitation (TR) pressure gradient. Although echocardiographic evaluation of inferior vena cava diameter and collapsibility is currently being used to estimate RA pressure (IVC method), RA strain may be an alternative since atrial strain is related to atrial pressure.

Objective

We tested if RA strain by speckle tracking echocardiography can be used as a surrogate of mean RA pressure (RA strain method), and by adding the TR pressure gradient, be used to estimate SPAP.

Methods

We retrospectively analyzed 91 patients (mean age, 58 years) referred to right heart catheterization due to unexplained dyspnea or suspected pulmonary hypertension. Echocardiography was performed within 24 hours of the invasive procedure. RA reservoir strain was calculated from apical four-chamber view. SPAP was calculated as the sum of peak TR pressure gradient and estimated RA pressure by the IVC or RA strain methods.

Results

Right heart catheterization showed SPAP and mean RA pressures of 51±20 mmHg and 9±6 mmHg, respectively. RA reservoir strain was inversely correlated with mean RA pressure (r=−0.61, p<0.01). Thus, we set mean RA pressure as 5, 10 and 15 mmHg depending on high (≥25%), middle (10–25%) and low (≤10%) values of RA reservoir strain. As shown in the figure, both the RA strain and IVC methods when combined with peak TR velocity, provided good estimates of invasively measured SPAP.

Conclusions

RA strain provides a semiquantitative measure of RA pressure, which can be used in combination with peak TR velocity to estimate SPAP. This approach can be used as an alternative when the IVC method is not available in cases with poor subcostal window.

P2452Application of left atrial strain for differentiation between pre- and post-capillary pulmonary hypertension

Background

Pulmonary hypertension (PH) is classified as pre- or post-capillary PH, and pulmonary capillary wedge pressure (PCWP) >15 mmHg is used as criterion for post-capillary PH. Elevated left atrial (LA) pressure is associated with reduced LA reservoir strain. Thus, LA strain may potentially serve to differentiate between these diagnoses.

Objectives

This study tested the hypothesis that LA strain can be used as a noninvasive parameter to differentiate between pre- and post-capillary PH.

Methods

We analyzed 103 patients (mean age: 58 years, 51 female) referred to right heart catheterization due to unexplained dyspnea or suspected heart failure. Echocardiography was performed within 24 hours of the invasive procedure. Mean pulmonary artery pressure (PAP) was noninvasively estimated from tricuspid regurgitation (TR) velocity and inferior vena cava (IVC) diameter and collapsibility. LA reservoir strain was calculated from apical four-chamber view by speckle tracking echocardiography, and was feasible in 101 patients.

Results

Twenty-eight patients were invasively confirmed with pre-capillary PH and 43 patients with post-capillary PH. The remaining 32 patients had no PH. LA reservoir strain was significantly lower in patients with post-capillary PH than patients with pre-capillary PH (9.9±5.5% vs. 24.6±8.2%, p<0.01). At a cut-off value of 15.4%, LA reservoir strain could predict elevated PCWP >15 mmHg with AUC=0.88, sensitivity=84.8% and specificity=81.8%. As shown in the figure, echocardiography with LA reservoir strain correctly differentiated 82% of patients into pre- and post-capillary PH.

Conclusions

These results suggest that LA reservoir strain can be used to predict elevated PCWP, thus allowing discrimination between pre- and post-capillary PH.

P1480Left atrial strain improves estimation of left ventricular filling pressure

Background

The current algorithm in the 2016 recommendations for echocardiographic estimation of left ventricular filling pressure (LVFP) as normal or elevated, combines traditional indices of mitral inflow velocities, tissue Doppler, left atrial volume and tricuspid regurgitation velocity (Figure A). Some of the patients remain unclassified by this algorithm. Left atrial (LA) strain is a novel index that correlates well with LVFP and may improve estimation of LVFP in these patients.

Purpose

We tested if LA strain can improve estimation of LVFP for the patients that are unclassified by the 2016 algorithm.

Methods

We analyzed data from 100 patients who were referred to right heart catheterization due to unexplained dyspnea or suspected heart failure. Echocardiography was performed simultaneously with or within 24 hours of right heart catheterization. Pulmonary capillary wedge pressure (PCWP) was used as an estimate for LVFP and defined as elevated if above 12 mmHg. Elevated LVFP was first estimated using the 2016 algorithm. In patients who were unclassified by the algorithm due to conflicting indices or unattainable indices, LA strain was subsequently used to detect elevated LVFP using a cut-off found from ROC analysis of the whole cohort.

Results

Six patients were unclassified by the 2016 algorithm. The ROC analysis of all 100 patients showed that at an LA strain cut-off of above or below 16.2%, LVFP was correctly classified as normal or elevated, respectively, with a sensitivity of 83% and specificity of 88%. All 6 unclassified patients by the 2016 algorithm were correctly classified using the LA strain cut-off, effectively increasing the accuracy of the algorithm by 6 percentage points.

Conclusions

LA strain may have a role in non-invasive estimation of LVFP, particularly in patients who remain unclassified when using the conventional echocardiographic indices.

Acknowledgement/Funding

South-Eastern Norway Regional Health Authority

P2458Restricted left atrial motion as a result of atrial stiffening in patients with cardiac amyloidosis

Background

Left atrial (LA) involvement of abnormal amyloid fibrils could induce LA dysfunction and stiffening in patients with cardiac amyloidosis (CA). Thus, the assessments of LA function and stiffness might be a potential approach to diagnose CA phenotype among patients with hypertrophied hearts.

Purpose

We sought to determine whether LA reservoir strain with speckle tracking echocardiography could be used a marker of LA stiffness in a derivation cohort. Furthermore, we tested to our hypothesis that LA reservoir strain could differentiate CA patients from hypertrophic cardiomyopathy (HCM) in an independent validation cohort.

Methods

In the derivation cohort, echocardiography was performed simultaneously with measurements of pulmonary capillary wedge pressure (PCWP) in 50 patients with suspected or established heart failure and relatively preserved left ventricular (LV) ejection fraction (56±10%). LA maximum and minimum volume index, and reservoir strain were measured from apical four-chamber view. LA stiffness index was computed as a pressure rise from x-trough to v-wave divided by an increase from minimum to maximum indexed LA volume (Figure A). In an independent validation group, we studied a total of 33 biopsy-proved CA patients and 127 HCM patients (LV ejection fraction: 57±11% vs. 66±10%, P<0.01) in sinus rhythm on the date of comprehensive echocardiographic study. Among them, cardiac magnetic resonance imaging (CMR) could be evaluated in 17 CA patients and 98 HCM patients. Furthermore, right heart catheterization was performed with 12 CA patients and 12 HCM patients in the CMR group.

Results

The derivation cohort study found that there was a significant curvilinear correlation of LA reservoir strain to LA stiffness index (Figure B). In the validation cohort, LA reservoir strain was reduced in patients with CA compared with HCM in all participants (11.6±5.6% vs. 18.5±6.9%, P<0.01), although there was no significant difference of LA maximal volume index between 2 groups (37±16 ml/m2 vs. 37±12 ml/m2, p=0.89). In the CMR group, the late gadolinium enhancement was observed in the LA wall in 16 patients with CA (94.1%) as shown in Figure C. In contrast, the LA enhancement revealed only in 1 patient with HCM (1.0%). Among patients with invasive measures, LA stiffness index [median (interquartile range)] was higher in patients with CA than that in patients with HCM [1.1 (0.4–2.8) vs. 0.2 (0.1–0.6), P=0.01].

Conclusions

LA reservoir function was fairly limited in patients with CA compared with HCM. Restricted LA motion might be related to atrial amyloid deposits or fibrosis, which potentially provokes atrial chamber stiffening.

Everolimus Initiation With Early Calcineurin Inhibitor Withdrawal in De Novo Heart Transplant Recipients: Three-Year Results From the Randomized SCHEDULE Study

In a randomized, open-label trial, de novo heart transplant recipients were randomized to everolimus (3-6 ng/mL) with reduced-exposure calcineurin inhibitor (CNI; cyclosporine) to weeks 7-11 after transplant, followed by increased everolimus exposure (target 6-10 ng/mL) with cyclosporine withdrawal or standard-exposure cyclosporine. All patients received mycophenolate mofetil and corticosteroids. A total of 110 of 115 patients completed the 12-month study, and 102 attended a follow-up visit at month 36. Mean measured GFR (mGFR) at month 36 was 77.4 mL/min (standard deviation [SD] 20.2 mL/min) versus 59.2 mL/min (SD 17.4 mL/min) in the everolimus and CNI groups, respectively, a difference of 18.3 mL/min (95% CI 11.1-25.6 mL/min; p < 0.001) in the intention to treat population. Multivariate analysis showed treatment to be an independent determinant of mGFR at month 36. Coronary intravascular ultrasound at 36 months revealed significantly reduced progression of allograft vasculopathy in the everolimus group compared with the CNI group. Biopsy-proven acute rejection grade ≥2R occurred in 10.2% and 5.9% of everolimus- and CNI-treated patients, respectively, during months 12-36. Serious adverse events occurred in 37.3% and 19.6% of everolimus- and CNI-treated patients, respectively (p = 0.078). These results suggest that early CNI withdrawal after heart transplantation supported by everolimus, mycophenolic acid and steroids with lymphocyte-depleting induction is safe at intermediate follow-up. This regimen, used selectively, may offer adequate immunosuppressive potency with a sustained renal advantage.

The Effect of Everolimus Initiation and Calcineurin Inhibitor Elimination on Cardiac Allograft Vasculopathy in De Novo Recipients: One-Year Results of a Scandinavian Randomized Trial

Early initiation of everolimus with calcineurin inhibitor therapy has been shown to reduce the progression of cardiac allograft vasculopathy (CAV) in de novo heart transplant recipients. The effect of de novo everolimus therapy and early total elimination of calcineurin inhibitor therapy has, however, not been investigated and is relevant given the morbidity and lack of efficacy of current protocols in preventing CAV. This 12-month multicenter Scandinavian trial randomized 115 de novo heart transplant recipients to everolimus with complete calcineurin inhibitor elimination 7-11 weeks after HTx or standard cyclosporine immunosuppression. Ninety-five (83%) patients had matched intravascular ultrasound examinations at baseline and 12 months. Mean (± SD) recipient age was 49.9 ± 13.1 years. The everolimus group (n = 47) demonstrated significantly reduced CAV progression as compared to the calcineurin inhibitor group (n = 48) (ΔMaximal Intimal Thickness 0.03 ± 0.06 and 0.08 ± 0.12 mm, ΔPercent Atheroma Volume 1.3 ± 2.3 and 4.2 ± 5.0%, ΔTotal Atheroma Volume 1.1 ± 19.2 mm(3) and 13.8 ± 28.0 mm(3) [all p-values ≤ 0.01]). Everolimus patients also had a significantly greater decline in levels of soluble tumor necrosis factor receptor-1 as compared to the calcineurin inhibitor group (p = 0.02). These preliminary results suggest that an everolimus-based CNI-free can potentially be considered in suitable de novo HTx recipients.

Microbiota-dependent metabolite trimethylamine-N-oxide is associated with disease severity and survival of patients with chronic heart failure

OBJECTIVES

Recent metabolomic, experimental and clinical studies have demonstrated that trimethylamine-N-oxide (TMAO), a microbiota-dependent metabolite from dietary phosphatidylcholine and carnitine, is a strong predictor of coronary artery disease (CAD). This finding suggests a link between the gut microbiota and atherosclerosis. The potential impact of TMAO in chronic heart failure (HF) is unknown. We hypothesized that TMAO levels would provide prognostic information about adverse outcomes in chronic HF.

DESIGN

Prospective, observational study including 155 consecutive patients with chronic HF. In addition, 100 patients with stable CAD without HF and 33 matched healthy individuals were included as controls. Plasma levels of TMAO and its precursors choline and betaine were measured, and associations with symptoms, aetiology and transplant-free survival in the patients with HF were explored.

RESULTS

Plasma levels of TMAO (P = 0.01), choline (P < 0.001) and betaine (P < 0.001) were elevated in patients with chronic HF compared to control subjects, with the highest levels in patients with New York Heart Association (NYHA) classes III and IV. Furthermore, TMAO levels were highest in individuals with ischaemic HF, followed by those with stable CAD and nonischaemic HF. TMAO, but not choline or betaine, was associated with reduced transplant-free survival: approximately 50% of patients in the upper tertile of TMAO levels died or received a heart transplant during 5.2 years of follow-up (unadjusted Cox-regression: hazard ratio 2.24, 95% confidence interval 1.28-3.92, P = 0.005).

CONCLUSIONS

TMAO levels were elevated in patients with HF and associated with NYHA class, ischaemic aetiology and adverse outcomes. Future studies should focus on gut microbiota, dietary composition and intestinal dysfunction in relation to TMAO levels and clinical outcome in HF.

Everolimus initiation and early calcineurin inhibitor withdrawal in heart transplant recipients: a randomized trial

In a randomized, open-label trial, everolimus was compared to cyclosporine in 115 de novo heart transplant recipients. Patients were assigned within 5 days posttransplant to low-exposure everolimus (3–6 ng/mL) with reduced-exposure cyclosporine (n = 56), or standard-exposure cyclosporine (n = 59), with both mycophenolate mofetil and corticosteroids. In the everolimus group, cyclosporine was withdrawn after 7–11 weeks and everolimus exposure increased (6–10 ng/mL). The primary efficacy end point, measured GFR at 12 months posttransplant, was significantly higher with everolimus versus cyclosporine (mean ± SD: 79.8 ± 17.7 mL/min/1.73 m2 vs. 61.5 ± 19.6 mL/min/1.73 m2; p < 0.001). Coronary intravascular ultrasound showed that the mean increase in maximal intimal thickness was smaller (0.03 mm [95% CI 0.01, 0.05 mm] vs. 0.08 mm [95% CI 0.05, 0.12 mm], p = 0.03), and the incidence of cardiac allograft vasculopathy (CAV) was lower (50.0% vs. 64.6%, p = 0.003), with everolimus versus cyclosporine at month 12. Biopsy-proven acute rejection after weeks 7–11 was more frequent with everolimus (p = 0.03). Left ventricular function was not inferior with everolimus versus cyclosporine. Cytomegalovirus infection was less common with everolimus (5.4% vs. 30.5%, p < 0.001); the incidence of bacterial infection was similar. In conclusion, everolimus-based immunosuppression with early elimination of cyclosporine markedly improved renal function after heart transplantation. Since postoperative safety was not jeopardized and development of CAV was attenuated, this strategy may benefit long-term outcome.

Virtual histology assessment of cardiac allograft vasculopathy following introduction of everolimus--results of a multicenter trial

In this 12-month multicenter Scandinavian study, 78 maintenance heart transplant (HTx) recipients randomized to everolimus with reduced calcineurin inhibitor (CNI) exposure or continued standard CNI-therapy underwent matched virtual histology (VH) examination to evaluate morphological progression of cardiac allograft vasculopathy (CAV). Parallel measurement of a range of inflammatory markers was also performed. A similar rate of quantitative CAV progression was observed in the everolimus (n = 30) and standard CNI group (n = 48) (plaque index 1.9 ± 3.8% and 1.6 ± 3.9%, respectively; p = 0.65). However, VH analysis revealed a significant increase in calcified (2.4 ± 4.0 vs. 0.3 ± 3.1%; p = 0.02) and necrotic component (6.5 ± 8.5 vs. 1.1 ± 8.6%; p = 0.01) among everolimus patients compared to controls. The increase in necrotic and calcified components was most prominent in everolimus patients with time since HTx >5.1 years and was accompanied by a significant increase in levels of von Willebrand (vWF) factor (p = 0.04) and vascular cell adhesion molecule (VCAM) (p = 0.03). Conversion to everolimus and reduced CNI is associated with a significant increase in calcified and necrotic intimal components and is more prominent in patients with a longer time since HTx. A significant increase in vWF and VCAM accompanied these qualitative changes and the prognostic implication of these findings requires further investigation.