PARADISE-MI – event rates and treatment effect of sacubitril/valsartan v ramipril by the presence or absence of transient pulmonary congestion and/or LVEF less or greater than 40

Background/Introduction

Sacubitril/valsartan was compared to ramipril in patients with acute myocardial infarction in the PARADISE-MI trial. In the whole trial population sacubitril/valsartan did not reduce the composite primary outcome of CV death or incident heart failure compared to ramipril. Whether or not event rates and/or treatment effects vary in patients with different baseline characteristics is unknown.

Purpose

To investigate a) event rates b) the treatment effect of sacubitril/valsartan compared to ramipril and c) safety by the presence or absence of transient pulmonary congestion and/or left ventricular ejection fraction (LVEF) ≤40%.

Methods

PARADISE-MI was a double-blind, randomised clinical trial that compared sacubitril/valsartan to ramipril in 5661 patients with an acute myocardial infarction with either LVEF ≤40% and/or transient pulmonary congestion. 3 groups were investigated: 1) LVEF ≤40% with pulmonary congestion (n=2012) and 2) LVEF ≤40% without pulmonary congestion (n=2596) and 3) LVEF not ≤40% with pulmonary congestion (n=1044).

Results

Patients with pulmonary congestion (with and without LVEF ≤40%) were more likely to have had a prior MI, prior CABG or PCI, had more atrial fibrillation and were more often treated with mineralocorticoid receptor antagonists and diuretics than patients with no pulmonary congestion and LVEF ≤40%. Patients with LVEF ≤40% and pulmonary congestion had more than twice the rate of the primary composite outcome compared to those with LVEF ≤40% without pulmonary congestion: 10.2 (95% CI 9.2–11.3) vs. 4.8 (4.3–5.5) events per 100 patient-years, respectively). Patients with pulmonary congestion and LVEF not ≤40% had an intermediate event rate (6.6, 5.5–7.9, events per 100 patient-years). A similar pattern of event rates was seen for the components of the primary outcome and for all secondary outcomes whether Clinical Events Committee or investigator-reported events were analysed. The treatment effect of sacubitril/ valsartan versus ramipril did not vary between the 3 congestion/ LVEF subgroups. The safety of sacubitril/valsartan compared to ramipril did not vary between congestion/LVEF subgroups.

Conclusion

Patients with pulmonary congestion with or without LVEF ≤40% had higher rates of primary and all secondary outcomes than those without pulmonary congestion and LVEF ≤40%. The treatment effect, and safety, of sacubitril/valsartan compared to ramipril was consistent in patients with or without pulmonary congestion and with or without LVEF ≤40%.

Funding Acknowledgement

Type of funding sources: Private company. Main funding source(s): Novartis

Increased circumferential deformation predicts cardiac events when left ventricular ejection fraction is normal

Background

Recent research suggests that early left ventricular (LV) systolic dysfunction is characterized by compromised longitudinal function and a concomitant augmentation of circumferential deformation expressed as increased global circumferential strain (GCS). It is hypothesized that increased GCS could be used to identify people at higher risk of cardiovascular events when left ventricular ejection fraction (LVEF) is preserved.

Purpose

The present study aimed to investigate whether elevated GCS in individuals with preserved LVEF predicts major adverse cardiovascular events (MACE).

Methods

A total of 2,874 participants from the general population were included in this prospective cohort study. All participants had echocardiography performed and analyzed. Exclusion criteria were HF at baseline, non-sinus rhythm during echocardiography, and inadequate image quality for GCS assessment. Outcome was MACE including incident heart failure, myocardial infarction, and/or cardiovascular death. Absolute values of strain were used. To determine the optimal transition point of a potential effect modification, Cox regression models with different LVEF thresholds were created. The model with the optimal hazard ratio (HR) and P values of the interaction terms and Akaike information criterion was chosen.

Results

Mean age was 53±18 years and 60% were female. Mean LVEF and GCS were 57±6% and 21.6±4.0%, respectively. Median follow-up was 3.5 years [IQR: 2.6; 4.4] and a total of 92 (3.2%) developed MACE. A U-shaped relationship was observed between GCS and MACE. The lowest incidence rate was within the GCS range of 20 to 25% (Figure 1a). A significant interaction (P<0.001) between GCS and LVEF was observed regarding MACE. The optimal transition point for this effect modification was determined to be LVEF=50%. In multivariable Cox regressions (including conventional cardiovascular risk factors and global longitudinal strain (GLS)), increasing GCS was significantly associated with future MACE in participants with LVEF ≥50% (HR=1.09 [95% CI: 1.01; 1.17] per 1% increase, n=2420) while decreasing GCS were associated with a higher risk of MACE in individuals with abnormal LVEF (HR=1.16 [95% CI: 1.04; 1.29] per 1% decrease, n=289) (Figure 1b). A total of 50 (2.1%) developed the MACE amongst the participants with LVEF ≥50%. This group had significantly lower GLS (19.0±2.7% vs 19.9±2.0%, P=0.013) and ratio of peak early transmitral filling velocity to peak early diastolic tissue velocity (E/e') (10.7 [IQR: 7.9; 14.3] vs 6.5 [5.3; 8.4], P<0.001) compared to the participants that did not develop MACE with normal LVEF. Figure 2 illustrates a model of the development of GCS, LVEF, GLS, and E/e' as LV function declines.

Conclusion

In the general population, the prognostic value of GCS is modified by LVEF. In participants with normal LVEF, higher GCS was associated with increased risk of MACE, while the opposite was observed in participants with abnormal LVEF.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): The Copenhagen City Heart Study is funded by The Danish Heart FoundationThe Metropolitan Region of Denmark (public funding).

A prospective validation of a deep learning-based automated workflow for the interpretation of the echocardiogram

Funding Acknowledgements

Type of funding sources: Private company. Main funding source(s): Us2.ai

Background. Deep learning can automate the interpretation of medical imaging tests. This study aimed to prospectively assess the interchangeability of deep learning algorithms with expert human measurements for interpreting echocardiographic studies, the primary method for assessing cardiac structure and function.

Methods. We compared a deep learning interpretation of 23 echocardiographic parameters—including cardiac volumes, ejection fraction, and Doppler measurements—with three repeated measurements by core lab human experts in a prospective study for submission to the United States Food and Drug Administration (FDA). The primary outcome metric was the individual equivalence coefficient (IEC), which compares the disagreement between deep learning and human readers relative to the disagreement among human readers. The pre-determined non-inferiority criterion was 0.25 for the upper bound of the 95% confidence interval (CI). Secondary outcomes included measures of agreement, including the mean absolute deviation.

Results. We included 602 studies from 600 participants (421 with heart failure, 179 controls, 69% women) with a mean age of 57 ± 16 years. The point estimates of IEC were all <0, indicating that the disagreement between the deep learning and human measures were lower than the disagreement among three core lab readers, and the upper bound of the 95% CI of IECs fell below the prespecified success criterion of 0.25. Secondary endpoints showed good agreement of automated with human expert measurements (Figure), with comparable or lower mean absolute deviations between automated and human experts relative to the mean absolute deviation among human experts.

Conclusion. This prospective validation study demonstrated excellent agreement between deep learning and expert human interpretation for a wide range of echocardiographic measurements. These results highlight the potential of deep learning algorithms to improve efficiency and reduce costs of echocardiography. Abstract Figure.

Lung ultrasound in acute heart failure: association between quality of life, symptoms and B-lines

Background

Impaired health-related quality of life (HRQL) and pulmonary congestion are common and important findings among patients hospitalized for acute heart failure (AHF). There are few data describing the association between HQRL, symptoms and pulmonary congestion in AHF.

Purpose

This study investigates whether worse HRQL and symptoms is associated with more pulmonary congestion. Pulmonary congestion measured by lung ultrasound (LUS) in patients with AHF is a marker of worse prognosis at baseline and pre-discharge.

Methods

In this 2-site, prospective, observational study, four-zone LUS was performed at baseline (LUS1) and within 72h of hospital discharge (LUS2) in patients hospitalized for AHF. B-lines were quantified off-line, blinded to clinical findings and outcomes, by a core laboratory. Clinicians managing the patients were blinded to LUS findings. HRQL was assessed at baseline using the patient-reported Kansas City Cardiomyopathy Questionnaire Total Symptom Score (KCCQ-TSS). Physician assessment of functional limitation at baseline was reported using the NYHA classification. In a subset of participants, patient-reported dyspnea at rest was also examined, at baseline and pre-discharge, using a numeric ranking scale (range 0–10; 10 worst). Dyspnea on exertion was also examined at baseline.

Results

Among 322 patients (mean age 72, 60% men, mean LVEF 39%) the median [interquartile range] KCCQ-TSS score was 33 [18–48]. Those with worse KCCQ-TSS scores, analyzed in tertiles, were younger, more likely to be obese and have diabetes mellitus and asthma/COPD, more likely to be on diuretics and report worse dyspnea at rest. At baseline, worse KCCQ-TSS was associated with worse NYHA class (Spearman's rho = −0.33, p<0.0001), dyspnea at rest (Spearman's rho = −0.41, p<0.0001) and dyspnea on exertion (Spearman's rho = −0.44, p<0.0001). A higher number of B-lines on LUS1 was weakly associated with worse NYHA class (Spearman's rho = 0.15, p=0.007) (Figure 1) but was not significantly associated with KCCQ-TSS (p=0.91), dyspnea at rest (p=0.74) or dyspnea on exertion (p=0.96). Among 118 patients with LUS2 performed within 72h of hospital discharge, pre-discharge dyspnea at rest was not significantly associated with B-lines (p=0.98).

Conclusion

Among patients hospitalized for AHF, at baseline, worse KCCQ-TSS was associated with worse NYHA class, dyspnea at rest and dyspnea on exertion but was not significantly associated with pulmonary congestion assessed by LUS. A higher number of B-lines at baseline was associated with worse NYHA class. Patient-reported breathlessness and HQRL measures, although important, may not consistently reflect the degree of pulmonary congestion in patients with AHF.

Funding Acknowledgement

Type of funding sources: Public Institution(s). Main funding source(s): National Institutes of Health/National Heart, Lung, and Blood Institute (NIH/NHLBI) (K23HL123533 to E.P.) and the British Heart Foundation (PG/13/17/30050 to R.T.C. and J.J.V.M.) Association of LUS1 B-lines & NYHA class