Left and right ventricular echocardiographic patterns in cardiogenic pulmonary edema: An exploratory retrospective study

Abstract

BACKGROUND

Cardiogenic pulmonary edema (CPE) is a severe manifestation of acute heart failure (AHF), typically driven by elevated left ventricular (LV) filling pressures. Although LV dysfunction has traditionally been emphasized, right ventricular (RV) function and biventricular interaction may also contribute to the hemodynamic profile of acute decompensation.

AIM

To describe the clinical and echocardiographic profile of patients with CPE and to explore associations between left and RV echocardiographic parameters and the in-hospital course.

METHODS

We retrospectively analyzed 28 patients with CPE from a larger AHF registry, excluding those with incomplete echocardiographic data or missing N-terminal pro-B-type natriuretic peptide measurement. Comprehensive echocardiographic assessment at admission included tricuspid annular plane systolic excursion, RV systolic annular velocity, RV-right atrial gradient, LV ejection fraction, LV outflow tract velocity-time integral, LV systolic annular velocity, mitral E velocity, and E/E’ ratio. Clinical variables, N-terminal pro-B-type natriuretic peptide levels, and in-hospital variables, including mortality, ventilatory support, inotropic support, and intensive care unit stay, were recorded. Group comparisons, correlation analyses, and exploratory receiver operating characteristic curve analyses were performed.

RESULTS

The mean LV systolic annular velocity was significantly lower than the RV systolic annular velocity (7.34 ± 2.38 cm/second vs 11.49 ± 2.70 cm/second; P < 0.001). Four patients (14.3%) died during hospitalization. No echocardiographic parameter showed a statistically significant association with in-hospital mortality. Exploratory analyses showed numerical differences in mitral E velocity, ventricular systolic velocities, and the ratio of LV to RV systolic annular velocity across in-hospital severity measures, but these findings were not statistically significant.

CONCLUSION

In this small retrospective cohort of patients with CPE, LV systolic annular velocity was lower than RV systolic annular velocity at presentation, while no echocardiographic parameter showed a significant association with in-hospital mortality. These findings should be considered exploratory and hypothesis-generating and require confirmation in larger prospective studies.

Key Words: Cardiogenic pulmonary edema; Acute heart failure; Echocardiography; Right ventricular function; Left ventricular function; Left ventricular outflow tract velocity-time integral; E/E’ ratio; Mortality

Core Tip: This single-center retrospective exploratory study describes the clinical and echocardiographic profile of patients presenting with cardiogenic pulmonary edema, with particular focus on left and right ventricular systolic function. Admission echocardiographic parameters were examined in relation to the in-hospital course, including mortality, ventilatory support, inotropic support, and intensive care unit stay. In this small cohort, left ventricular systolic annular velocity was significantly lower than right ventricular systolic annular velocity, whereas no echocardiographic parameter showed a significant association with in-hospital mortality. These findings should be interpreted cautiously and considered hypothesis-generating.

INTRODUCTION

Cardiogenic pulmonary edema (CPE) is a severe and potentially life-threatening presentation of acute heart failure (AHF), caused by a rapid increase in left-sided cardiac filling pressures and pulmonary capillary hydrostatic pressure, leading to acute pulmonary congestion and respiratory compromise[1,2]. Despite major advances in the treatment of heart failure, AHF continues to be associated with substantial short-term morbidity and mortality, particularly in patients requiring urgent hospitalization and advanced supportive care[3,4]. Early clinical assessment remains essential in order to identify patients at higher risk of complicated in-hospital evolution and to guide initial therapeutic decisions[5,6].

Transthoracic echocardiography (TTE) is a central component of the bedside evaluation of AHF because it provides immediate, non-invasive information on ventricular systolic function, filling pressures, valvular abnormalities, and overall hemodynamic status[7,8]. In patients with CPE, echocardiography may be particularly useful because this clinical syndrome can result from multiple overlapping mechanisms, including impaired left ventricular (LV) systolic reserve, diastolic dysfunction with elevated filling pressures, acute blood pressure changes, and concomitant valvular disease[9]. Although LV ejection fraction (LVEF) remains widely used in routine practice, it does not fully capture the hemodynamic complexity of acute decompensation and may be insufficient when used in isolation[10].

Additional echocardiographic parameters may offer a more nuanced description of ventricular performance in this setting. LV outflow tract velocity-time integral (LVOT VTI) is commonly used as a surrogate of forward stroke volume, while tissue Doppler-derived systolic velocity of the lateral mitral annulus (S’ LV) may provide complementary information on longitudinal LV systolic function[11,12]. Similarly, right ventricular (RV) performance can be assessed using simple and reproducible measurements such as tricuspid annular plane systolic excursion (TAPSE), tissue Doppler-derived systolic velocity of the lateral tricuspid annulus (S’ RV), and the RV-RA pressure gradient. In AHF, RV dysfunction has been associated with adverse outcomes in broader patient populations, but its role specifically in patients presenting with CPE remains less clearly characterized[13-17].

The interaction between the right and left ventricles may also be clinically relevant in acute decompensated states. However, interpretation of relative differences between LV and RV longitudinal systolic parameters requires caution, since systolic annular velocities differ physiologically between the two ventricles because of their distinct geometry, loading conditions, and myocardial fiber architecture[18,19]. In the absence of an appropriate control group, such differences should not be interpreted as disease-specific markers, but they may still provide descriptive insight into ventricular functional patterns observed during acute CPE.

Given the heterogeneous hemodynamic substrate of CPE, a broader echocardiographic characterization of both ventricles may help improve understanding of this presentation at admission. At the same time, because available data in this specific phenotype remain limited, small retrospective cohorts are more suitable for exploratory analyses than for definitive prognostic modeling[20].

In this study, we aimed to describe the clinical and echocardiographic profile of patients with CPE and to explore associations between left and RV echocardiographic parameters and the in-hospital course.

MATERIALS AND METHODS

Study design and patients

This single-center retrospective exploratory observational study was conducted between January 2024 and July 2025. From a registry of 222 consecutive patients hospitalized for AHF, we identified patients presenting with CPE as the predominant clinical phenotype.

CPE was diagnosed on the basis of acute respiratory distress supported by at least two of the following criteria: Bilateral alveolar infiltrates on chest radiography, multiple B-lines on lung ultrasound consistent with interstitial-alveolar fluid, pulmonary crackles on auscultation, and hypoxemia on arterial blood gas analysis. Rapid clinical and radiographic improvement after intravenous diuretics and/or vasodilator therapy was considered supportive of the diagnosis[21,22].

Patients were excluded if they had incomplete echocardiographic datasets, concomitant acute coronary syndrome, Takotsubo cardiomyopathy, myocarditis, primary valvular emergencies, or missing NT-proBNP measurement at admission. The final analytic cohort included 28 patients with complete echocardiographic and laboratory data available for exploratory comparative analysis of left and RV function. Inclusion criteria were age ≥ 18 years, primary presentation with CPE as the dominant AHF phenotype, and complete echocardiographic and laboratory evaluation within the first 6 hours of admission, including NT-proBNP measurement.

Because the analysis was restricted to patients with complete datasets, the final study population represents a selected subgroup of the initial AHF registry. This approach was necessary to ensure internal consistency of the exploratory analysis, but it may limit the representativeness of the cohort.

The study complied with the Declaration of Helsinki and received approval from the institutional ethics committee[23].

Data collection

Demographic, clinical, laboratory, and outcome data were extracted from the hospital’s electronic medical records. Clinical variables included age, sex, cardiovascular risk factors, blood pressure and heart rate at presentation, and signs of congestion. Laboratory parameters included serum NT-proBNP, creatinine, hemoglobin, electrolytes, and high-sensitivity troponin[24-26].

The in-hospital outcomes of interest included all-cause in-hospital mortality, need for respiratory support, need for inotropic support, and total duration of intensive care unit (ICU) stay. Mortality was considered a hard clinical endpoint, whereas respiratory support requirement, inotropic support, and ICU stay were considered indicators of in-hospital clinical course and treatment intensity.

Echocardiographic assessment, outcomes and definitions

Comprehensive TTE was performed within 6 hours of admission using a GE Vivid E95 ultrasound system. All measurements adhered to current ASE/EACVI recommendations[27,28] and were interpreted by two experienced echocardiographers blinded to patient outcomes.

The analysis focused on conventional echocardiographic parameters that were consistently available in the emergency setting and could be retrospectively retrieved in a uniform manner.

RV systolic function was assessed using TAPSE (mm) measured by M-mode in the apical four-chamber view, tissue Doppler-derived systolic velocity at the lateral tricuspid annulus (S RV, cm/second), and the peak systolic RV-right atrial pressure gradient (RV-RA gradient, mmHg) derived from tricuspid regurgitation velocity using the modified Bernoulli equation[29,30].

LV systolic function was assessed using LVEF (%) calculated by the biplane Simpson method, LVOT VTI (cm) measured by pulsed-wave Doppler in the apical five-chamber view, and tissue Doppler-derived systolic velocity at the lateral mitral annulus (S LV, cm/second)[31,32].

Diastolic function was characterized by early diastolic mitral inflow velocity (E, cm/second), early diastolic mitral annular velocity (E’ LV, cm/second), and the ratio of early mitral inflow velocity to early diastolic mitral annular velocity (E/E’) ratio as a surrogate marker of LV filling pressures[33]. The S LV/S RV ratio was calculated as a descriptive marker of the relative relationship between left and RV longitudinal systolic velocities[19].

Outcomes and definitions

All-cause in-hospital mortality was recorded during the index hospitalization. Respiratory support was defined as the requirement for non-invasive ventilation or invasive mechanical ventilation (IMV) during hospitalization. Inotropic support was defined as the administration of intravenous inotropic therapy for hemodynamic support, as documented in the medical record. ICU stay was defined as the total duration of ICU hospitalization, expressed in days.

Given the retrospective design, these variables were analyzed as markers of in-hospital clinical course. In particular, the need for ventilatory or inotropic support and the duration of ICU stay may also reflect clinician judgment and local management practices in addition to disease severity.

Statistical analysis

Length of ICU stay was analyzed as a continuous numerical variable.

Continuous variables were tested for normality using the Shapiro-Wilk test and expressed as mean ± SD or median with interquartile range (IQR), as appropriate. Categorical variables were expressed as n (%). Group comparisons (survivors vs non-survivors; IMV vs no-IMV; inotrope vs no-inotrope) were made using the independent-samples t-test or Mann-Whitney U test, depending on data distribution. Categorical variables were compared using the χ² or Fisher’s exact test[34,35].

Paired within-patient comparisons (e.g., S LV vs S RV) employed the Wilcoxon signed-rank test. Correlations between echocardiographic parameters and NT-proBNP, as well as with the total number of days spent in intensive care, were assessed using Spearman’s rank correlation coefficient.

Exploratory receiver operating characteristic (ROC) curve analyses were performed for selected echocardiographic parameters, including TAPSE, S RV, RV-RA gradient, TAPSE/RV-RA gradient ratio, LVEF, LVOT VTI, E/E’ ratio, S LV, and S LV/S RV ratio, in relation to in-hospital mortality and ventilatory support requirement. The area under the curve (AUC) was calculated with 95% confidence intervals using bootstrap resampling. Thresholds derived from the Youden index were considered exploratory only and were not intended for clinical application or formal prognostic classification[36].

Given the small sample size and low number of outcome events, all statistical analyses were considered exploratory and hypothesis-generating. No multivariable modeling was performed, and the study was not designed or powered to identify independent predictors of outcome.

Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 26.0 (IBM Corp., Armonk, NY, United States) and JASP, Version 0.19 (University of Amsterdam, Amsterdam, The Netherlands). A two-sided P value < 0.05 was considered statistically significant.

RESULTS

Patient’s characteristics at baseline

A total of 28 patients with CPE at presentation fulfilled the predefined inclusion criteria, all undergoing comprehensive echocardiographic evaluation of right and LV function. Baseline demographic, clinical, and echocardiographic characteristics are presented in Table 1.

Table 1 Baseline demographic, clinical, and echocardiographic characteristics of the study cohort (n = 28).

Variable	Total (n = 28)	Survivors (n = 24)	Non-survivors (n = 4)	P value
Age (years)	73.82 ± 12.14	74.12 ± 11.68	72.00 ± 16.57	0.819
Male sex, %	61.9	64.7	50.0	0.618
Smoking, %	21.4	20.8	25.0	1.000
Hypertension, %	100.0	100.0	100.0	1.000
Dyslipidemia, %	96.4	100.0	75.0	0.143
Obesity, %	29.6	30.4	25.0	1.000
Diabetes mellitus, %	53.6	50.0	75.0	0.600
Valvular heart disease, %	71.4	75.0	50.0	0.555
Systolic BP (mmHg)	173.39 ± 38.52	172.58 ± 39.94	178.25 ± 33.05	0.772
Diastolic BP (mmHg)	99.39 ± 21.23	99.75 ± 22.59	97.25 ± 11.70	0.746
Heart rate (bpm)	105.61 ± 23.04	105.75 ± 24.69	104.75 ± 10.24	0.892
NT-proBNP (pg/mL)	10513.96 ± 10381.23	9955.83 ± 10293.27	13862.75 ± 11840.24	0.569
TAPSE (mm)	22.61 ± 8.20	23.29 ± 8.61	18.50 ± 3.32	0.072
LVOT VTI (cm)	16.20 ± 4.52	16.03 ± 4.72	17.23 ± 3.34	0.561
RV-RA gradient (mmHg)	28.93 ± 9.74	29.00 ± 10.51	28.50 ± 2.65	0.845
LVEF (%)	38.68 ± 15.42	38.83 ± 15.07	37.75 ± 19.92	0.923

Continuous variables are presented as mean ± SD, and categorical variables as percentages. Comparisons between survivors and non-survivors were performed using the Student’s t-test or Mann-Whitney U test, as appropriate, for continuous variables, and the χ² test or Fisher’s exact test for categorical variables. A two-sided P value < 0.05 was considered statistically significant. Valvular heart disease: ≥ moderate dysfunction of any cardiac valve. BP: Blood pressure; LVOT VTI: Left ventricular outflow tract velocity-time integral; LVEF: Left ventricular ejection fraction; NT-proBNP: N-terminal pro-B-type natriuretic peptide; RV-RA: Right ventricular to right atrial pressure gradient; TAPSE: Tricuspid annular plane systolic excursion.

The mean age of the study population was 73.8 ± 12.1 years, and 61.9% were male. Active smoking was reported in 21.4% of cases. Hypertension was present in all patients (100%), while dyslipidemia was highly prevalent (96.4%). Obesity was recorded in 29.6%, and diabetes mellitus in 53.6%.

At admission, the mean systolic blood pressure was 173.39 ± 38.52 mmHg and the mean diastolic blood pressure was 99.39 ± 21.23 mmHg, with a mean heart rate of 105.61 ± 23.04 bpm. The average NT-proBNP concentration was 10514 ± 10381 pg/mL. Echocardiographic measurements showed a mean TAPSE of 22.61 ± 8.20, LVOT VTI of 16.20 ± 4.52 cm, RV-RA gradient of 28.93 ± 9.74 mmHg, and LVEF of 38.68% ± 15.42%. Valvular heart disease was present in 71.4% of patients.

Echocardiographic parameters in relation to in-hospital mortality

Four patients (14.3%) died during hospitalization. Comparative analysis of echocardiographic parameters between survivors (n = 24) and non-survivors (n = 4) is presented in Table 2.

Table 2 Comparison of echocardiographic parameters between survivors and non-survivors with cardiogenic pulmonary edema.

Parameter	Survivors, median (IQR)	Non-survivors, median (IQR)	P value
E/E’	11.37 (10.15-14.33)	15.41 (12.70-18.17)	0.291
TAPSE	20.00 (17.75-26.25)	19.00 (17.75-19.75)	0.323
S RV	12.00 (10.00-13.20)	10.00 (8.25-11.75)	0.339
S LV	6.90 (5.97-7.62)	6.75 (6.12-7.12)	0.669
S LV/S RV	0.60 (0.50-0.71)	0.73 (0.59-0.83)	0.533
LVEF	37.50 (26.75-47.25)	38.00 (20.75-55.00)	0.844
LVOT VTI	16.00 (13.38-18.25)	17.00 (14.72-19.50)	0.645

Continuous variables are presented as median and interquartile range. Comparisons between groups were performed using the Mann-Whitney U test due to non-normal data distribution. A two-sided P value < 0.05 was considered statistically significant. E/E’ is ratio of early mitral inflow velocity to early diastolic mitral annular velocity. LVEF: Left ventricular ejection fraction; S LV: Systolic velocity of the lateral mitral annulus; S RV: Systolic velocity of the lateral tricuspid annulus; TAPSE: Tricuspid annular plane systolic excursion; LVOT VTI: Left ventricular outflow tract velocity-time integral; IQR: Interquartile range.

Although none of the evaluated parameters reached statistical significance, numerical differences between groups were observed.

The E/E’ ratio, a surrogate of LV filling pressures[33,37], was higher in non-survivors compared with survivors (median: 15.4 vs 11.3; P = 0.291). This distribution is illustrated in Figure 1, where the median E/E’ is higher in the non-survivor group, although the IQRs overlap.

Figure 1 Comparison of ratio of early mitral inflow velocity to early diastolic mitral annular velocity ratio between survivors and non-survivors. E/E’ is ratio of early mitral inflow velocity to early diastolic mitral annular velocity. The authors used ChatGPT (version GPT-5.1, OpenAI, San Francisco, CA, United States) to assist in generating the graphical representations of the study data included in Figure 1. All AI-assisted graphical content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy, originality, and integrity of the final figures.

TAPSE and S RV, markers of RV longitudinal systolic function[13,15], were numerically lower in non-survivors (TAPSE: 19.0 mm vs 20.0 mm, P = 0.323; S RV: 10.0 cm/second vs 12.0 cm/second, P = 0.339). S LV was slightly lower in non-survivors (6.75 cm/second vs 6.90 cm/second; P = 0.669), while the S LV/S RV ratio was numerically higher in non-survivors (0.73 vs 0.61; P = 0.533).

Exploratory ROC curve analysis was performed for key RV parameters (Figure 2A), as well as for LV parameters and the S LV/S RV ratio (Figure 2B).

Figure 2 Exploratory receiver-operating characteristic curves in relation to in-hospital mortality. A: Exploratory receiver-operating characteristic (ROC) curves for right ventricular parameters in relation to in-hospital mortality; B: Exploratory ROC curves for left ventricular parameters and systolic velocity at the lateral mitral annulus/systolic velocity at the lateral tricuspid annulus ratio in relation to in-hospital mortality. E/E’ is ratio of early mitral inflow velocity to early diastolic mitral annular velocity. AUC: Area under the curve; TAPSE: Tricuspid annular plane systolic excursion; S RV: Right ventricular systolic annular velocity; RV-RA gradient: Right ventricle-right atrium pressure gradient; TAPSE/RV-RA gradient: Ratio of tricuspid annular plane systolic excursion to right ventricle-right atrium pressure gradient; LVOT: Left ventricular outflow tract; VTI: Velocity-time integral; LVEF: Left ventricular ejection fraction; S LV/S RV: Ratio of left ventricular to right ventricular systolic annular velocity. The authors used ChatGPT (version GPT-5.1, OpenAI, San Francisco, CA, United States) to assist in generating the graphical representations of the study data included in Figure 2. All AI-assisted graphical content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy, originality, and integrity of the final figures.

Figure 2A displays ROC curves for TAPSE, S RV, the RV-RA gradient, and the TAPSE/RV-RA gradient ratio. All curves were close to the diagonal, and none of the evaluated RV parameters reached statistical significance.

Figure 2B depicts the ROC curves for LVOT VTI, LVEF, E/E’, and the S LV/S RV ratio. Among the evaluated parameters, the E/E’ ratio had the highest AUC (0.69), although this did not reach statistical significance. The S LV/S RV ratio had an AUC of 0.60, with an exploratory threshold of 0.83 according to Youden’s index. LVOT VTI yielded an AUC of 0.59, with an exploratory cut-off value of 13.9 cm (P = 0.47). LVEF showed limited separation from the reference line. Previously reported studies have described LVOT VTI values below 14-17 cm in association with low cardiac output and adverse outcomes in AHF[11,12].

Contrary to our initial hypothesis, LVOT VTI was not lower in non-survivors: Median 17.0 cm (IQR: 14.7-19.5) vs 16.0 cm (IQR: 13.4-18.3) in survivors (P = 0.645). The ROC-derived cut-off of 13.9 cm showed limited separation in this cohort. However, prior data suggest that LVOT VTI < 17 cm is associated with low cardiac output and poor outcomes in AHF[11,12].

To further illustrate the distribution of LVOT VTI according to in-hospital survival status, we performed a boxplot-based comparative analysis between survivors and non-survivors (Figure 3).

Figure 3 Distribution of left ventricular outflow tract velocity-time integral according to in-hospital mortality. LVOT: Left ventricular outflow tract; VTI: Velocity-time integral. The authors used ChatGPT (version GPT-5.1, OpenAI, San Francisco, CA, United States) to assist in generating the graphical representations of the study data included in Figure 3. All AI-assisted graphical content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy, originality, and integrity of the final figures.

None of the evaluated ROC curves reached statistical significance, and all ROC-derived results should therefore be interpreted as exploratory, given the limited sample size.

Ventilation and intensive care requirements

In the analyzed cohort, 7 patients (25%) required IMV, while 11 patients (39.3%) underwent at least one episode of non-invasive ventilation (NIV/CPAP). The median duration of cardiac ICU stay was 4 days (IQR: 2-6), with a range from 0 days to 33 days.

To evaluate the relationship between ventricular function and the need for invasive ventilation, echocardiographic parameters were compared according to ventilatory support category, as shown in Table 3. No parameter reached statistical significance across groups. Among the evaluated variables, S RV showed the lowest P value, although this finding did not reach statistical significance. LVEF values were numerically lower in patients who required IMV than in the other ventilatory support groups.

Table 3 Echocardiographic parameters according to ventilatory support requirements (n = 28).

Parameter	No ventilation, median (IQR)	NIV, median (IQR)	IMV, median (IQR)	P value
TAPSE	19.00 (15.75-28.00)	20.00 (17.50-24.00)	20.50 (17.75-23.25)	0.916
S RV	12.35 (10.50-13.60)	11.70 (9.00-12.20)	11.45 (9.75-12.43)	0.483
S LV	6.70 (5.83-8.20)	7.00 (6.00-7.75)	6.75 (6.12-7.05)	0.683
S LV/S RV	0.62 (0.48-0.73)	0.62 (0.52-0.80)	0.62 (0.57-0.69)	0.865
RV-RA gradient	31.0 (17.75-40.25)	29.0 (21.50-34.50)	29.5 (26.75-35.50)	0.816
E/E’	12.29 (10.89-15.48)	11.44 (9.97-13.92)	13.61 (9.85-18.17)	0.680
LVOT VTI	16.40 (13.55-21.75)	15.00 (13.50-18.50)	15.75 (14.72-17.62)	0.800
LVEF	38.0 (27.25-54.00)	38.0 (30.50-45.00)	28.0 (20.75-40.00)	0.580

Continuous variables are presented as median and interquartile range. Patients were categorized into three groups: No ventilatory support, non-invasive ventilation, and invasive mechanical ventilation. Comparisons across groups were performed using the Kruskal-Wallis test. A two-sided P value < 0.05 was considered statistically significant. E/E’: Ratio of early mitral inflow velocity to early diastolic mitral annular velocity; RV-RA gradient: Right ventricle-right atrium systolic pressure gradient; IMV: Invasive mechanical ventilation; LVEF: Left ventricular ejection fraction; LVOT VTI: Left ventricular outflow tract velocity-time integral; NIV: Non-invasive ventilation; S LV: Systolic velocity of the lateral mitral annulus; S RV: Systolic velocity of the lateral tricuspid annulus; IQR: Interquartile range.

Spearman correlation analysis showed no statistically significant associations between echocardiographic parameters and ICU length of stay. As presented in Table 4, S LV and E/E’ showed the lowest P values, but neither correlation reached statistical significance. Numerically, lower S LV values and higher E/E’ ratios were observed in patients with longer ICU stay. A similar but weaker pattern was observed for LVEF and TAPSE.

Table 4 Spearman correlation analysis between echocardiographic parameters and intensive care unit length of stay.

Parameter	Spearman ρ	P value
TAPSE	-0.148	0.444
S RV	-0.078	0.686
S LV	-0.189	0.326
S LV/S RV	-0.001	0.995
RV-RA gradient	-0.055	0.777
E-wave	0.175	0.363
E/E’	0.198	0.302
LVOT VTI	0.031	0.874
LVEF	-0.175	0.365

Spearman’s rank correlation coefficient (ρ) was used to assess the association between echocardiographic parameters and duration of intensive care unit stay. A two-sided P value < 0.05 was considered statistically significant. E/E’ is ratio of early mitral inflow velocity to early diastolic mitral annular velocity. E-wave: Peak early mitral inflow velocity; RV-RA gradient: Right ventricle-right atrium systolic pressure gradient; LVEF: Left ventricular ejection fraction (biplane Simpson method); LVOT VTI: Left ventricular outflow tract velocity-time integral; S LV: Systolic velocity of the lateral mitral annulus; S RV: Systolic velocity of the lateral tricuspid annulus; TAPSE: Tricuspid annular plane systolic excursion.

From a descriptive perspective, these exploratory findings indicate that variations in LV systolic velocity, LV filling pressure surrogates, and biventricular systolic parameters may accompany differences in the in-hospital course of patients with CPE. However, given the absence of statistically significant associations in this limited cohort, these observations should be interpreted cautiously and not as evidence of established relationships. Previous studies have explored the potential contribution of integrated hemodynamic and echocardiographic assessment in AHF settings[12,22].

Comparative analysis of S LV and S RV

To explore potential differences in longitudinal systolic performance between the left and right ventricles in patients with CPE, we compared the tissue Doppler-derived S LV with that of the lateral tricuspid annulus (S RV). This analysis included all 28 patients with complete data for both parameters.

The mean S LV was 7.34 ± 2.38 cm/second, significantly lower than the mean S RV of 11.49 ± 2.70 cm/second (Wilcoxon signed-rank test, P < 0.001)[18]. Figure 4 shows the paired distribution of S LV and S RV values, with the median S RV exceeding the median S LV across the study cohort.

Figure 4 Paired comparison of left and right ventricular systolic annular velocities (systolic velocity at the lateral mitral annulus and systolic velocity at the lateral tricuspid annulus) in patients with cardiogenic pulmonary edema. S LV: Systolic velocity of the lateral mitral annulus; S RV: Systolic velocity of the lateral tricuspid annulus. The authors used ChatGPT (version GPT-5.1, OpenAI, San Francisco, CA, United States) to assist in generating the graphical representations of the study data included in Figure 4. All AI-assisted graphical content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy, originality, and integrity of the final figures.

This finding indicates that, in this cohort, longitudinal systolic velocity measured at the lateral mitral annulus was lower than that measured at the lateral tricuspid annulus at presentation. Given the absence of a control group, this difference should be interpreted as a descriptive observation within the study population. Further studies with larger cohorts are warranted to determine whether the degree of S LV to S RV disparity has clinical or prognostic relevance.

Association between echocardiographic parameters and inotropic support

In the study cohort, 7 patients (24.1%) required inotropic support during hospitalization. Comparative analysis of key echocardiographic markers of right and LV function between patients with and without inotropic therapy is shown in Table 5. No statistically significant differences were observed for any of the measured parameters.

Table 5 Echocardiographic parameters according to inotropic support requirement.

Parameter	No inotrope, median (IQR)	Inotrope, median (IQR)	P value
E/E’	12.02 (10.62-14.33)	13.61 (9.38-18.17)	0.874
TAPSE (mm)	19.50 (17.75-26.25)	19.50 (17.75-20.50)	0.448
S RV (cm/second)	12.00 (10.00-13.20)	10.00 (8.25-11.75)	0.339
S LV (cm/second)	6.90 (5.97-7.50)	6.75 (6.12-8.00)	0.973
S LV/S RV	0.60 (0.50-0.71)	0.73 (0.59-0.93)	0.393
LVEF (%)	38.50 (26.75-52.00)	29.50 (20.75-42.25)	0.449
LVOT VTI (cm)	16.20 (13.38-19.00)	14.50 (13.97-16.50)	0.792

Continuous variables are presented as median and interquartile range. Patients were classified according to the need for inotropic support during hospitalization (no inotropic support vs inotropic support). Comparisons between groups were performed using the Mann-Whitney U test. A two-sided P value < 0.05 was considered statistically significant. E/E’ is ratio of early mitral inflow velocity to early diastolic mitral annular velocity. LVEF: Left ventricular ejection fraction (biplane Simpson method); LVOT VTI: Left ventricular outflow tract velocity-time integral; S LV: Systolic velocity of the lateral mitral annulus; S RV: Systolic velocity of the lateral tricuspid annulus; TAPSE: Tricuspid annular plane systolic excursion; IQR: Interquartile range.

The S LV/S RV ratio was numerically higher among patients who required inotropic support, as shown in Figure 5 (P = 0.393). LVEF was also numerically lower in the inotrope group, although this difference did not reach statistical significance (P = 0.449).

Figure 5 Distribution of the systolic velocity of the lateral mitral annulus/systolic velocity of the lateral tricuspid annulus ratio according to inotropic support requirement. S LV: Systolic velocity of the lateral mitral annulus; S RV: Systolic velocity of the lateral tricuspid annulus. The authors used ChatGPT (version GPT-5.1, OpenAI, San Francisco, CA, United States) to assist in generating the graphical representations of the study data included in Figure 5. All AI-assisted graphical content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy, originality, and integrity of the final figures.

These findings indicate that, in this small cohort, the need for inotropic therapy was not significantly associated with any single echocardiographic measure of systolic function. The observed numerical differences should be considered exploratory and interpreted with caution. Previous studies have examined the relationship between RV systolic function and hemodynamic status in acute cardiovascular settings[17,30].

DISCUSSION

This study explored the comparative systolic function of the left and right ventricles in patients presenting with CPE, using tissue Doppler-derived systolic velocities (S LV and S RV), conventional echocardiographic indices (TAPSE, LVEF, LVOT VTI, E/E’), and selected outcome measures, including in-hospital mortality, need for ventilatory and inotropic support, and intensive care stay.

The main finding of the present study was that S LV was significantly lower than S RV at presentation. In this cohort, this observation indicates that longitudinal systolic velocity measured at the lateral mitral annulus was lower than that measured at the lateral tricuspid annulus. However, this difference should be interpreted cautiously, as the study did not include a control group and physiological differences between left- and right-sided longitudinal annular velocities are well recognized. No echocardiographic parameter showed a statistically significant association with in-hospital mortality. Although the E/E’ ratio had the highest area under the ROC curve (AUC = 0.69), this result did not reach statistical significance and should be considered exploratory. Similarly, LVOT VTI was not lower in non-survivors, although the exploratory threshold of 13.9 cm was numerically close to values previously reported in AHF cohorts[12,20].

In addition, no echocardiographic parameter was significantly associated with the need for invasive ventilation or inotropic support. Exploratory analyses showed numerical differences in S RV, LVEF, and the S LV/S RV ratio across these in-hospital variables, but these findings were not statistically significant and should not be interpreted as established associations. Likewise, none of the echocardiographic indices correlated significantly with intensive care stay. Although higher mitral E velocity and lower S LV showed the lowest P values in relation to longer ICU hospitalization, these observations remain descriptive and hypothesis-generating in this limited cohort[33,37].

Comparison with previous studies

Previous work has emphasized the central role of elevated LV filling pressures in the pathophysiology of CPE, often in the setting of preserved or mildly reduced LVEF[1,2].

In the present cohort, S LV was significantly lower than S RV at presentation. This finding is consistent with the possibility that LV longitudinal systolic performance may be more reduced than the corresponding RV annular systolic velocity in the acute setting. However, this observation should be interpreted cautiously. Higher S’ RV values compared with S’ LV do not necessarily indicate pathological interventricular imbalance, as longitudinal systolic velocities differ physiologically between the right and left ventricles because of their distinct myocardial architecture and loading conditions. In the absence of a contemporaneous control group without pulmonary edema, the present study cannot determine whether this pattern is specific to CPE or reflects a broader physiological or hemodynamic relationship.

Studies in broader AHF populations have reported that tissue Doppler velocities, particularly S RV, may be associated with clinical outcomes and may reflect RV-pulmonary artery coupling[13,15]. In our cohort, S RV values were numerically lower in non-survivors and in patients requiring ventilatory or inotropic support, but these differences were not statistically significant. Therefore, our findings do not establish a prognostic role for S RV in CPE, but they support its inclusion in future prospective studies of biventricular function in this clinical setting.

The exploratory LVOT VTI threshold of 13.9 cm identified in our analysis is numerically close to previously reported values associated with reduced forward flow and poorer outcomes in AHF[11,12]. However, in our cohort, LVOT VTI was not lower in non-survivors, and the corresponding ROC findings were not statistically significant. Accordingly, this result should be viewed as descriptive rather than confirmatory. Similarly, although the S LV/S RV ratio showed numerical differences across some in-hospital variables, it was not significantly associated with outcome in this study, and its potential clinical relevance remains to be clarified in larger cohorts[20].

Influence of age and sex on ventricular function

Patient demographics may also influence ventricular function in the setting of CPE. The advanced age of the study population is consistent with the typical epidemiological profile of CPE and may be associated with age-related myocardial changes, including increased ventricular stiffness, impaired diastolic relaxation, and reduced longitudinal systolic reserve, all of which may affect LV performance during acute decompensation[2]. In addition, sex-related differences in cardiac structure and function have been reported, with men more often showing reduced systolic reserve, whereas women more frequently present with higher filling pressures and preserved ejection fraction phenotypes[4]. Although the present study was not powered to perform age- or sex-stratified analyses, these demographic characteristics may represent potential sources of variability in echocardiographic measurements and should be considered in future studies.

Clinical implications

From a hemodynamic perspective, the observed difference between S LV and S RV may be compatible with the hypothesis that right-LV interaction contributes to the acute presentation of CPE, including in hypertensive forms[18,19]. However, in the absence of a control group and direct hemodynamic measurements, the present data do not allow firm conclusions regarding RV-LV stroke volume mismatch as a specific mechanism in this cohort.

Tissue Doppler-derived systolic velocities are simple and rapidly obtainable in the emergency setting and may provide a useful descriptive assessment of ventricular function at presentation. Although no single echocardiographic parameter showed significant prognostic associations in this dataset, the combined evaluation of LV and RV systolic velocities, LVOT VTI, and diastolic filling pressure surrogates may still offer a broader non-invasive characterization of patients with CPE. In this context, multiparametric echocardiographic assessment may be useful for clinical profiling and for generating hypotheses to be tested in larger prospective studies.

Limitations

This study has several limitations. First, it was a retrospective, single-center exploratory analysis with a small sample size and a low number of outcome events, which substantially limited statistical power and precluded robust prognostic inference. Accordingly, the findings should be interpreted as descriptive and hypothesis-generating rather than as evidence of independent predictive relationships.

Second, the final analytic cohort represented a highly selected subgroup of the initial AHF registry, as only patients with complete echocardiographic and laboratory datasets were included. Although this approach was necessary for analytic consistency, it may have introduced selection bias and may limit the representativeness and generalizability of the findings.

Third, echocardiographic measurements were performed at a single time point within the first 6 hours of admission, precluding assessment of dynamic changes in ventricular function during treatment and hospitalization. Serial echocardiographic evaluation might better characterize the temporal evolution of left and RV performance in CPE and should be incorporated into future prospective studies. In such studies, a structured approach including assessment at presentation, after initial hemodynamic stabilization, and prior to discharge may provide additional insight into ventricular adaptation during the acute phase.

Fourth, invasive hemodynamic data were not available to corroborate the echocardiographic findings. In addition, cardiac rhythm disturbances were not systematically analyzed, and advanced echocardiographic parameters such as RV fractional area change or strain-derived indices were not consistently available for uniform retrospective assessment. These factors may have influenced both echocardiographic measurements and clinical course and should be addressed in future studies. Nevertheless, the echocardiographic indices evaluated in this study are widely used, non-invasive measures of ventricular function and filling pressure surrogates and remain relevant in routine clinical practice.

Fifth, the analysis was limited to in-hospital variables, and no long-term survival, rehospitalization, or post-discharge functional outcome data were available. Moreover, some of the in-hospital variables examined, particularly inotropic support and ICU stay, may also reflect clinician judgment and local management practices rather than disease severity alone.

Finally, the absence of a control group without pulmonary edema precluded direct comparison with reference ventricular patterns outside the acute CPE setting, thereby limiting interpretation of the observed differences between S LV and S RV. Larger prospective studies incorporating serial imaging, rhythm assessment, advanced RV and strain parameters, invasive hemodynamic data, and longer follow-up are needed to better define the clinical relevance of these exploratory findings.

Future directions

Prospective multicenter studies with larger patient populations are needed to further investigate the clinical relevance of the ventricular functional patterns observed in CPE, including the relationship between left and RV longitudinal systolic parameters. Incorporating advanced imaging techniques such as speckle-tracking strain, indices of RV-pulmonary artery coupling, and integrated hemodynamic measures such as the cardiac power index may provide a more comprehensive characterization of this syndrome[22,29].

Future studies should also examine whether combining echocardiographic parameters with biomarkers such as NT-proBNP and relevant clinical variables can improve phenotypic characterization and support the development of more robust multiparametric models in patients with CPE. External validation in larger prospective cohorts will be necessary to determine whether biventricular echocardiographic assessment provides incremental value beyond conventional clinical and imaging parameters.

CONCLUSION

In this study of patients presenting with CPE, S LV was significantly lower than S RV at presentation, indicating a difference between left and RV longitudinal systolic velocities within this cohort. However, in the absence of a control group, this finding should be interpreted as a descriptive observation and not as evidence of a disease-specific interventricular imbalance. Elevated LV filling pressures remain central to the pathophysiology of pulmonary edema, and the present findings are consistent with the importance of left-sided dysfunction in this setting[38].

No single echocardiographic parameter showed a statistically significant association with in-hospital mortality, ventilatory support, inotropic support, or intensive care stay in this limited cohort. Exploratory numerical differences were observed for LVOT VTI, S LV/S RV ratio, S RV, mitral E velocity, and S LV across several in-hospital variables, but these findings were not statistically significant and should be considered hypothesis-generating only[11,33,38].

Taken together, these results suggest that a broader echocardiographic assessment incorporating LV systolic function, RV systolic indices, and diastolic filling pressure surrogates may provide a more complete descriptive characterization of patients with CPE than LVEF alone. In this context, the present observations are conceptually compatible with multiparametric approaches that integrate systolic and diastolic markers, including previously proposed combined indices such as the Virtue Index[20].

Given its retrospective single-center design, small sample size, and low number of outcome events, this study should be viewed as exploratory. Larger prospective multicenter studies incorporating serial imaging, biomarkers, and integrated hemodynamic assessment are needed to determine the clinical relevance of these findings and to clarify whether combined biventricular echocardiographic evaluation offers incremental value in the assessment of patients with CPE[37].