Pressure-recovery adjustment of aortic valve area does not improve risk prediction in aortic stenosis

Background

In aortic stenosis (AS), overestimation of pressure gradients by Doppler compared to catheter-based measurement can be adjusted for by accounting for the pressure-recovery, expressed as an energy-loss index (ELI). However, evidence of improved risk assessment by ELI instead of aortic valve area indexed by body-surface area (AVAi) is scarce.

Purpose

We aimed to evaluate the prognostic performance of ELI and AVAi in a head-to-head comparison using large-scale, real-world data.

Methods

In the multi-center, mortality-data linked National Echocardiography Database of Australia (NEDA), patients with AS and requisite aortic area measurements were identified. The prognostic value of AVAi and ELI, respectively, was analyzed using Cox regression and compared by difference in C statistics.

Results

In patients with mild AS (n=3,179), moderate AS (n=4,194), and severe AS (n=3,120), there were 4,229 deaths of which 2,359 were cardiovascular deaths (median [interquartile range] follow-up 2.5 [1.1–4.5] years). There was an association with increased mortality risk per 0.1 cm2/m2 decrement for both AVAi (hazard ratio CV mortality [95% confidence interval] 1.18 [1.16–1.20]) and ELI (1.10 [1.09–1.12]). Prognostic performance for 5-year CV mortality did not improve by using ELI compared to AVAi (C statistic 0.626 vs 0.626, p=0.96).

Conclusions

In aortic stenosis, ELI was not associated with improved prognostic performance compared to AVAi using large-scale, real-world clinical data. These data do not support routine calculation of pressure recovery in echocardiographic reporting.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): Swedish Heart and Lung Foundationthe Swedish Cardiac Society

Incident pulmonary hypertension in 13,448 cases investigated with repeat echocardiography: insight from the National Echo Database of Australia (NEDA)

Background

We addressed the paucity of data describing the characteristics and natural history of incident pulmonary hypertension (PHT).

Methods

In this observational clinical cohort study, we examined 85,173 individuals undergoing routine echocardiography for heart disease without evidence of PHT (according to estimated right ventricular systolic pressure, eRVSP <30 mmHg). Subsequent age and sex-specific incidence of PHT were derived from repeat echocardiograms conducted a median of 3.43 (interquartile range, IQR 1.49 to 6.55) years apart in 13,448 adults. Progressive PHT severity with individual data linkage to all-cause mortality were examined per PHT stage: eRVSP 30–39 mmHg (mild PHT), 40–49 mmHg (moderate PHT) and ≥50 mmHg (severe PHT).

Results

A total of 6,169 men (45.9%, mean age 61.4±16.7 years) and 7,279 women (60.8±16.9 years) with no initial evidence of PHT were identified (first echocardiogram). Subsequently, 5,412 (40.2%,) developed PHT (repeat echocardiogram) – comprising 4,125 (30.7%, 65.0±14.3 years), 928 (6.9%, 69.2±13.5 years), and 359 (2.7%, 69.8±12.7 years) cases of mild, moderate, and severe PHT, respectively. The incidence of all stages of PHT was 15.2 cases per million men/annum and 12.5 cases per million women/annum. Overall, median eRVSP increased by +7.4 (IQR +4.6 to +10.1) and +30.7 (IQR +26.0 to +37.3) mmHg; median E:e' ratio increased by +1.0 (IQR −0.4 to +3.2) and +3.6 (IQR +2.0 to +8.2); and median LA volume increased by +5.0 (IQR +0.0 to +12.0) and +19.5 (IQR +9.0 to +31.0) ml/m2, respectively, in mild and severe PHT groups between first and last echocardiograms. During subsequent median 8.1 years follow-up, 2,776/13,448 (20.6%) individuals died from all-cause. Compared to no PHT, the age- and sex-adjusted hazard ratios for all-cause mortality increased to 1.35 (95% confidence interval, CI 1.23–1.47) in mild PHT, 1.94 (95% CI 1.73–2.18) in moderate PHT, and 2.43 (95% CI 2.09–2.83) in severe PHT (all p<0.001).

Conclusions

New onset of PHT is a common finding among individuals with heart disease followed-up with echocardiography. Even milder stage of PHT is associated with higher mortality, reinforcing the need for proactive evaluation for symptoms consistent with PHT.

Funding Acknowledgement

Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Johnson & Johnson: Investigator-initiated grant

Counting the cost of progressive aortic valve stenosis: an international observational clinical cohort study

Background

We sought, for the first time, to quantify the societal to healthcare burden associated with the entire spectrum of aortic stenosis (AS) - the most common cardiac valve disorder requiring clinical management.

Methods

In this observational clinical cohort study with individual data linkage to all-cause mortality, we examined the distribution of progressive stenosis in the native aortic valves of 98,565 men and 99,357 women aged ≥65 years within a multicentre cohort across 24 sites internationally. Individuals were grouped according to their peak aortic valve velocity on last electrocardiogram investigation: 1.0–1.99 m/s (no AS/reference groups) and then 0.5 m/s increments up to ≥4.0 m/s (severe AS). Sex-specific premature mortality and years of life-lost (YLL) during 5-year follow-up were calculated, along with willingness-to-pay (WTP) to regain quality-adjusted life-years (QALYs) and healthcare costs associated with concurrent heart failure (HF).

Results

Overall, 20,701 men (21.0%; 95% CI 20.8–21.3) and 18,576 women (18.7%; 95% CI 18.5–18.9) had AS at baseline. Five-year mortality in men and women with no AS was 38.6% and 36.8%, respectively. In both sexes, mortality rose from 46.8–49.0% to 59.5–65.7% in the least- to most-severe cases of AS. Premature mortality (53.5–59.1% in the no AS group) declined from 42.4–50.9% to 34.8–35.2% of all deaths. Per 1,000 individuals, AS in men and women was associated with 8 and 13 more premature deaths resulting in 36 and 62 more QALYs (societal cost of $AU1.54 and $AU2.68 million respectively) when compared to no AS. The additional 5-year cost of healthcare utilisation for the 25 and 35 more men and women with concurrent HF and AS was estimated to be $AU1.27 and $AU1.91 million, respectively.

Conclusions

The presence of any degree of aortic valve stenosis is associated with higher levels of premature mortality coupled with excess healthcare costs that impose a heavy societal burden. These findings indicate the need for proactive and cost-effective strategies to promptly detect this common condition.

Funding Acknowledgement

Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Edwards LifeSciences: Investigator-initiated grant

Relative prevalence and outcomes of low gradient severe aortic stenosis in routine clinical practice

Background

The prevalence and outcomes of low-gradient severe aortic stenosis (AS), in particular low-flow low-gradient (LFLG) severe AS, have not been well characterized.

Objectives

We sought to delineate the relative prevalence and outcomes of patients with low-gradient severe AS in routine clinical cardiology practice.

Methods

Data were derived from the National Echocardiography Database of Australia (NEDA), a national multicentre clinical registry with data linkage to mortality. A total of 192,060 adults with native aortic valve profiling from 2000–2019 were identified, of whom 12,013 (6.3%) had severe AS (aged 75.3±13.1 years, 53.5% men). Severe AS subtypes were determined using standard echocardiographic criteria. All-cause and cardiovascular-related mortality were assessed on an adjusted basis (age and sex) in 8,162 patients with classifiable severe AS subtypes, during mean follow-up of 88±45 months. Additionally, rates of recorded Aortic Valve Replacement (AVR) during follow-up were compared between AS groups.

Results

5,601 patients (47%) had high-gradient and 6,412 patients (53%) had low-gradient severe AS. In the low-gradient group, Stroke Volume Index data was recorded in 2,741 patients; 1,750 (64%) had LFLG and 991 (36%) had normal-flow low-gradient (NFLG). For LFLG patients, 1,570 had LV ejection fraction recorded; 959 (61%) had paradoxical LFLG and 651 (39%) had classical LFLG. Adjusted 5-year cardiovascular mortality rates were 28% in high-gradient, 24% in NFLG, 27% in paradoxical LFLG and 50% in classical LFLG patients (p<0.001). Rates of recorded AVR per mean follow-up years were 5.2% for high-gradient, 4.0% for NFLG, 3.1% for classical LFLG and 2.4% for paradoxical LFLG severe AS (p<0.001).

Conclusions

Approximately half the subjects with echocardiographic features of severe AS in routine clinical practice have low-gradient haemodynamics, associated with long-term mortality comparable to or worse than high-gradient severe AS. The poorest survival was associated with classical LFLG severe AS.

Funding Acknowledgement

Type of funding sources: None. Summary illustration5 year mortality curves

Right ventricular myocardial work characterisation in patients with pulmonary hypertension: association with invasive haemodynamic parameters

Background

Non-invasive evaluation of indices of right ventricular (RV) myocardial work derived from RV pressure-strain loops may provide novel insights into RV function in pre-capillary pulmonary hypertension.

Purpose

This study was designed to evaluate the association between the indices of RV myocardial work and invasive haemodynamic parameters in a patient cohort with pulmonary arterial hypertension (Group I) or chronic thromboembolism pulmonary hypertension (Group IV).

Methods

The non-invasive analysis of echocardiography-derived RV myocardial work (Figure 1, upper panel) was completed in 51 patients (mean age 58.1±12.7 years, 31% male) with Group I (78%) or Group IV (22%) pulmonary hypertension. Conventional echocardiographic measurements of RV systolic function, RV global work index (RV GWI), RV global constructive work (RV GCW), RV global wasted work (RV GWW) and RV global work efficiency (RV GWE) were compared with parameters derived invasively during right heart catheterisation (RHC).

Results

The median RV GWI, RV GCW, RV GWW and RV GWE were 620 (443 to 857) mmHg%, 830 (650 to 1206) mmHg%, 105 (54 to 169) mmHg% and 87 (82 to 93)%, respectively. Compared to pulmonary artery systolic pressure and conventional echocardiographic parameters of RV systolic function (RV global longitudinal strain [GLS], tricuspid annular plane systolic excursion and RV fractional area change), RV GCW and RV GWI correlated more closely with invasively-derived RV stroke work index (R=0.63, P<0.001 and R=0.60, P<0.001, respectively) (Figure 1, lower panels). Invasively-derived pulmonary vascular resistance (PVR) correlated with RV GWW (R=0.63, P<0.001), RV GWE (R=0.48, P<0.001) and RV GLS (R=0.58, P<0.001). RV GLS correlated more closely with invasively-derived stroke volume index (R=−0.57, P<0.001) than RV GCW, RV GWI and RV GWE (R=0.34, P=0.016, R=0.48, P<0.001 and R=0.47, P<0.001, respectively).

Conclusions

In a patient cohort with Group I and Group IV pulmonary hypertension, indices of RV myocardial work were more closely correlated with invasively-derived RV stroke work index and PVR than conventional echocardiographic parameters of RV systolic function.

Funding Acknowledgement

Type of funding sources: None. Figure 1. Method and correlations

Cardiac damage staging classification and prognosis in low flow low gradient severe aortic stenosis

Introduction

There are currently no established prognostic models for low-flow, low-Gradient (LFLG) severe aortic stenosis (AS). The Cardiac Damage Staging Classification, has been validated as a clinically useful prognostic tool in high-gradient severe AS, but not specifically in LFLG severe AS.

Objective

To assess and confirm the prognostic utility of the Cardiac Damage Staging Classification, in LFLG severe AS patients.

Methods

We analysed data from the National Echocardiography Database of Australia (NEDA), a national multicentre registry with individual data linkage to mortality. Of 192,060 adults (aged 62.8±17.8 years) with comprehensive ultrasound profiling of the native aortic valve studied between 2000–2019, 12,013 (6.3%) had severe AS. Based on standard echocardiographic parameters, 611 classical and 959 paradoxical LFLG patients were identified. Mean follow-up was 70±41 months. All-cause and cardiovascular-related mortality were assessed for each LFLG group on an adjusted basis (age and sex), according to cardiac damage classification staging.

Results

Paradoxical LFLG patients were younger (mean age 74.3 vs 76.2, p=0.006) and more often female (62.8% vs 36.3%, p<0.001). Classical LFLG patients had greater associated cardiac damage at diagnosis (76% vs. 49%, for Stage ≥2, p<0.001). Compared to Stage 0 paradoxical LFLG patients, adjusted 1- and 5-year cardiovascular-related mortality was 200% (HR 2.82, 95% CI 1.12–7.11) and 120% (HR 2.42, 95% CI 1.46–4.02) higher in Stage 2 patients and 350% (HR 4.23, 95% CI 1.68–10.63) and 175% (HR 3.18, 95% CI 1.90–5.34) higher in Stage 3/4 patients, respectively. Compared to Stage 1 classical LFLG patients, adjusted 1- and 5-year cardiovascular-related mortality was 58% (HR 1.66, 95% CI 0.95–2.90) and 24% (HR 1.35, 95% CI 0.91–1.99) higher in Stage 2 patients and 125% (HR 2.48, 95% CI 1.43–4.28) and 52% (HR 1.87, 95% CI 1.26–2.78) higher in Stage 3/4 patients, respectively.

Conclusion

In patients with LFLG severe AS identified by echocardiography, increasing severity denoted by the cardiac damage staging classification is strongly associated with increasing medium- to long-term mortality.

Funding Acknowledgement

Type of funding sources: None. Stages + Classical LFLG 5 year mortalityParadoxical LFLG 5 year mortality curves

Stroke volume index and outcomes in low gradient severe aortic stenosis

Background

Conventionally, stroke volume index (SVI) ≤35ml/m2 is considered as “low-flow” in severe aortic stenosis (AS). Recent evidence suggests a lower threshold SVI (30ml/m2) may be more prognostically informative.

Objectives

To assess the association between SVI and survival in patients with low-gradient severe AS, with either preserved (LVEF ≥50%) or reduced (LVEF<50%) ejection fraction.

Methods

Data were derived from the National Echocardiography Database of Australia (NEDA). Of 192,060 adults (aged 62.8±17.8 years) with comprehensive aortic valve profiling between 2000–2019, 12,013 (6.3%) had severe AS. Based on standard echocardiographic parameters, 1,623 patients had severe low-gradient AS with preserved LVEF and 744 patients had severe low-gradient AS with reduced LVEF. All-cause and cardiovascular-related mortality were assessed for each low-gradient group on an adjusted basis (age, sex, BMI, AVA indexed and RVSP), according to SVI. Mean follow-up was 81±42 months.

Results

Compared to patients with “normal” flow (SVI >35ml/m2), those with “low-flow” (SVI ≤35ml/m2) had significantly higher BMI, smaller AVA index and lower AV mean gradient and peak velocity (p≤0.001). In low-gradient with preserved LVEF patients, adjusted survival at 1- and 3-years was significantly lower only for SVI ≤30ml/m2 compared with SVI>35ml/m2 (p<0.001 and p<0.03, respectively). In low-gradient with reduced LVEF patients, adjusted survival at 1- and 3-years was significantly lower for SVI ≤35ml/m2 compared with SVI>35ml/m2 (p=0.015 and p=0.018, respectively).

Conclusions

Taken together with previous data, our results suggest that a SVI threshold of ≤30ml/m2 (rather than ≤35 ml/m2) is prognostically significant in severe low-gradient AS with preserved LVEF.

Funding Acknowledgement

Type of funding sources: None. Survival plots LGAS with preserved EFSummary of results

P2440Increased mortality with high versus normal ejection fraction in elderly women

Methods

NEDA contains echo data (>530,000, years 1997–2017) from laboratories (N=14) across Australia, linked to the National Death Index (NDI). 331,344 individuals aged >18 years (n=172,251 [52%] men, age=60.8±18.0 years and 152,095 [48%] women, 60.9±19.2 years) had a mean follow up of 5.4 person-years and 63,142 fatal events. LVEF using Method Of Discs (MOD) could be calculated in 116,544 individuals. We compared mortality in those with normal LVEF (52–72% for males and 54–74% for females) compared with high LVEF (>72% and >74% for males and females, respectively).

Results

Mortality was 11% higher in females with high LVEF vs females with normal LVEF. On multivariable logistic regression, high ejection fraction remained an independent predictor of mortality after correction for age, gender and body mass index. High LVEF was positively associated with female gender, age and diastolic dysfunction. Of 96,033 individuals in the study population, there were 46,876 males (48.8%), mean age 61.19±17.09 years, not significantly different from their female counterparts. 44,282 males had normal LVEF (52–72%) as opposed to 2,594 with high LVEF (>72%). In females, 46,186 had normal LVEF (54–74%) and 2,971 had high LVEF.InCox proportional hazard models, the hazard ratio (HR) for mortality in males with normal EF was 1.105 for age (95% CI 1.102 to 1.07) and 1.074 (95% CI 0.988 to 1.167) for high LVEF (p<0.001 for all outputs). In females, the corresponding adjusted HR was 1.107 for age (95% CI 1.105 to 1.110) and 1.140 for high LVEF (95% CI 1.057 to 1.230).

Conclusion

High LVEF, particularly in older women, is associated with increased mortality compared with an LVEF within the reference range. Further study is required to evaluate possible mechanisms for worse outcomes in patients with high LVEF.

Dissociation of visual and haptic vertical in two patients with vestibular nuclear lesions

The somatosensory (haptic) vertical (HV) and visual vertical (VV) were assessed in two patients with vestibular nuclear lesions. Patient 1 had paroxysmal nystagmus, and was tested "on" and "off." The HV was normal "on" and "off" but the VV was severely tilted during vestibular paroxysms. Patient 2, with a brainstem stroke, was tested at months 1 and 6. The VV was severely tilted on both occasions (>12 degrees) but the HV was marginally tilted (4 degrees) in the acute stage only. These VV-HV dissociations suggest that vestibular nuclear lesions influence gravity perception mostly via ocular torsional effects rather than by disrupting a single, internal representation of verticality.