Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure

Abstract

Background:

Afterload-related cardiac performance (ACP), a diagnostic parameter for septic cardiomyopathy, integrates both cardiac performance and vascular effects and could predict prognosis in septic shock.

Objectives:

We hypothesized that ACP would also correlate with clinical outcomes in patients with chronic heart failure (HF).

Design:

A retrospective study.

Methods:

We retrospectively studied consecutive patients with chronic HF who underwent right heart catheterization and established an expected cardiac output-systemic vascular resistance (CO-SVR) curve model in chronic HF for the first time. ACP was calculated as CO_measured/CO_predicted × 100%. ACP > 80%, 60% < ACP ⩽ 80%, and ACP ⩽ 60% represented less impaired, mildly impaired, and severely impaired cardiovascular function, respectively. The primary outcome was all-cause mortality, and the secondary outcome was event-free survival.

Results:

A total of 965 individual measurements from 290 eligible patients were used to establish the expected CO-SVR curve model (CO_predicted = 53.468 × SVR ^−0.799). Patients with ACP ⩽ 60% had higher serum NT-proBNP levels (P < 0.001), lower left ventricular ejection fraction (P = 0.001), and required dopamine more frequently (P < 0.001). Complete follow-up data were available in 263 of 290 patients (90.7%). After multivariate adjustment, ACP remained associated with both primary outcome (hazard ratio (HR) 0.956, 95% confidence interval (CI) 0.927–0.987) and secondary outcome (HR 0.977, 95% CI 0.963–0.992). Patients with ACP ⩽ 60% had the worst prognosis (all P < 0.001). ACP was significantly more discriminating (area under the curve of 0.770) than other conventional hemodynamic parameters in predicting mortality (Delong test, all P < 0.05).

Conclusion:

ACP is a powerful independent hemodynamic predictor of mortality in patients with chronic HF. ACP and the novel CO-SVR two-dimensional graph could be useful in assessing cardiovascular function and making clinical decisions.

Clinical trial registration:

URL: https://www.clinicaltrials.gov. Unique identifier: NCT02664818.

Graphical Abstract:

Keywords

Introduction

Hemodynamic monitoring is central to our understanding of heart failure (HF) pathophysiology. Right heart catheterization (RHC), as an integral aspect of the diagnosis and management of HF, provides clinicians with important hemodynamic profiles.¹ Among various hemodynamic parameters measured by RHC, cardiac output (CO) is the most popular one used to characterize the clinical phenotype of patients with cardiovascular destabilization. However, attempts to predict the prognosis of patients with HF using CO have always been controversial, as many analyses demonstrated that resting CO had less prognostic utility and was not associated with outcomes.^2–5 Another important hemodynamic parameter, systemic vascular resistance (SVR), is often used to represent afterload and is one of the key factors in determining CO.⁶ Previous studies have demonstrated that there was a nonlinear relationship between CO and SVR.^7,8 As CO decreases, SVR increases to maintain blood pressure. However, the increase in SVR in response to CO decreases may not be consistent, as SVR may increase appropriately (a compensated state), increase only slightly (a decompensated state, often leading to low blood pressure and cardiogenic shock), or increase excessively (an afterload mismatch state, often leading to pulmonary edema).⁹ In order to correct SVR and better assess the degree of impairment of cardiovascular function, it seems necessary to create a parameter that integrates both cardiac and vascular factors.

In 2011, Werdan et al. established an expected CO-SVR nonlinear regression model in a cohort of septic shock patients who might present with septic cardiomyopathy and innovatively developed the parameter ‘‘afterload-related cardiac performance’’ (ACP) to judge whether a particular CO is ‘normal’ or not when focusing on a given SVR value.⁷ This parameter integrates information from SVR and CO and is defined as CO_measured/CO_predicted × 100%. Studies have shown that ACP had a good correlation with cardiac index and cardiac power index, and an abnormal ACP was associated with higher mortality in septic patients.^7,10,11 However, the concept of ACP has never been introduced for risk assessment in patients with HF, and the expected CO-SVR relationship curve in HF patients remains uncertain.

In this study, we hypothesized that ACP could quantify the impairment of cardiovascular function and predict outcomes in chronic HF patients. Considering that the hemodynamic profile of patients with chronic HF is different from septic shock, we used RHC-derived hemodynamic data measured from patients with chronic HF to develop a completely new CO-SVR regression model to test this concept.

Methods

Study population

We conducted a retrospective cohort study including chronic HF patients who underwent RHC being considered for advanced therapy, evaluated for heart transplantation, or suspected of exertional dyspnea at Heart Failure Care Unit (HFCU) in our hospital between September 2013 and February 2022. Chronic HF was diagnosed according to the HF guidelines.^12,13 Inclusion criteria for patients in this study: (1) the presence of symptoms and/or signs of HF; (2) N-terminal pro-B-type natriuretic peptide (NT-proBNP) > 125 pg/ml; (3) abnormal findings of echocardiography. Patients who had revascularization or acute coronary syndrome within 7 days prior, had acute pulmonary embolism, had large pericardial effusion or constrictive pericarditis leading to cardiac tamponade, had septic shock, had hypertensive crisis, were receiving mechanical circulatory support or were acutely decompensated were excluded. Clinical characteristics, laboratory and echocardiographic tests, and HF therapies were recorded. Patients were followed up by telephone or clinic visits. Clinical outcomes including death, durable left ventricular assist device (LVAD) implantation, heart transplantation, and HF rehospitalization were collected. The primary outcome was all-cause mortality. The secondary outcome was event-free survival, which was considered as freedom from death, heart transplantation, LVAD implantation, or HF rehospitalization.

RHC and hemodynamic assessment

RHC was performed by Swan-Ganz catheter (Edwards Lifesciences, USA) following previously published standard operating procedures.¹⁴ The external pressure transducer was zeroed at the mid-thoracic level in each patient. Pressure measurements were recorded at end-expiration during spontaneous breathing and CO was measured by a thermodilution method. Blood pressure was measured noninvasively and was recorded during RHC.

Key hemodynamic measures recorded at the time of RHC included heart rate (HR), systolic/diastolic/mean arterial pressure (SAP/DAP/MAP), right atrial pressure (RAP), systolic/diastolic/mean pulmonary arterial pressure (s/d/m PAP), pulmonary arterial wedge pressure (PAWP), stroke volume (SV), and CO. SVR was calculated in Wood units as (MAP-RAP)/CO, and pulmonary vascular resistance (PVR) was calculated as (mPAP-PAWP)/CO. Pulmonary arterial compliance (PAC) was estimated as SV/(sPAP-dPAP). Cardiac power output (CPO) was defined as MAP × CO/451. Pulmonary artery pulsatility index (PAPI) was calculated as (sPAP-dPAP)/RAP.

Expected CO-SVR curve model and afterload-related cardiac performance

The idea to establish the expected CO-SVR curve model in chronic HF patients was consistent with the idea proposed in septic patients by Werdan et al.⁷ After the screening, 290 eligible patients with a total of 965 individual measurements were included in the analysis. We first constructed a CO-SVR nonlinear regression model (CO = 38.512 × SVR^−0.799, R² = 0.769) using all hemodynamic measurements from these patients (Figure 1). There was a large variation in CO values for each given SVR value. We hypothesized that for each SVR value, the corresponding maximum CO value belongs to the patients with relatively less impaired cardiovascular function. Therefore, we defined an expected CO-SVR relationship curve to be the upper limit of the 95% tolerance range (CO_predicted = 53.468 × SVR^−0.799) for this nonlinear regression model. The CO_predicted value calculated from this ‘expected’ curve was defined as the relatively ‘ideal’ CO for each given SVR value. ACP was calculated as CO_measured/CO_predicted × 100%. We considered ACP > 80% as a relatively less impaired cardiovascular function, 60%<ACP ⩽ 80% as mildly impaired cardiovascular function, and ACP ⩽ 60% as severely impaired cardiovascular function. In this study, the hemodynamic data measured for the first time after RHC in each patient were used to calculate ACP and perform subsequent analysis.

Figure 1.

Correlation of cardiac output and systemic vascular resistance.

Statistical analysis

Categorical values were expressed as numbers (%) and continuous variables as median with interquartile range or as mean ± SD. Shapiro–Wilk test was used to assess normality. Differences were evaluated by one-way analysis of variance (ANOVA) for continuous variables or Kruskal–Wallis if non-normally distributed, and by the Pearson χ² test for categorical variables. The association for frequencies was examined by calculating Cramer’s V coefficients. Linear regression was performed and correlations were measured with Spearman’s rank correlation coefficient. Based on our clinical expertise as well as previous studies,^15–17 age, sex, New York Heart Association (NYHA) functional class, left ventricular ejection fraction (LVEF), NT-proBNP, history of ischemic cardiomyopathy (ICM), atrial fibrillation (AF), hypertension, diabetes, hyperlipidemia, use of dopamine, and hemodynamic variables were selected as possible confounders of the ACP association and were assessed in the univariate model. The variables that remained significant at the 0.10 level in univariable analysis were considered for inclusion in the multivariate model. A forward stepwise method was used to remove variables with a p value > 0.10 and enter variables that met a 0.05 significance level for the selection of the final multivariate model. Subgroup analyses for the primary outcome included age, sex, NYHA functional class, LVEF, NT-proBNP, CO, MAP, RAP, SVR, ICM, AF, and hypertension. The Kaplan–Meier analysis was used to compare the different groups for the estimation of outcomes with the log-rank test. We also performed survival analyses when all-cause mortality was censored at the time of heart transplantation or LVAD implantation. We assessed the ability of hemodynamic parameters including ACP, CO, CPO, RAP, mPAP, PAWP, and PAC to discriminate between patients who had died and those still alive by the close of follow-up by calculating the area under the curve (AUC) and compared performance using the Delong method. Besides, we calculated that with a total sample size of 263 patients, the study would have more than 90% power to detect the difference in the primary outcome. A two-sided p < 0.05 was considered statistically significant. SPSS Statistics 25 (IBM, USA), R version 4.0.2 (The R Foundation, Austria), and Prism 8 (GraphPad Software, USA) were used for statistical analyses.

Results

ACP in relation to clinical characteristics and hemodynamic parameters

A total of 290 first measurements after RHC were analyzed from 290 individual patients. Baseline characteristics and hemodynamic profiles, for the whole study population and for each subset categorized by the value of ACP, are described in Table 1. Overall, the median age was 51 (35–61) years, and 210 (72.4%) patients were male. Sixty (20.7%) patients had a history of ICM, while 95 (32.8%) had AF, 71 (24.5%) had hypertension, 45 (15.5%) had diabetes, and 74 (25.5%) had hyperlipidemia. In patients with ACP > 80%, the proportion of hypertension was relatively higher (42.3% vs 19.1% vs 12.8%, P < 0.001), and they were more likely to be taking a renin-angiotensin-aldosterone system inhibitor (52.6% vs 38.2% vs 17.9%, P = 0.001). In contrast, patients with ACP ⩽ 60% were more likely to be taking loop diuretics (94.9% vs 93.6% vs 84.6%, P = 0.044) and digoxin (51.3% vs 46.8% vs 29.5%, P = 0.019). Support with dopamine also tended to be used more frequently in patients with ACP ⩽ 60% (48.7% vs 28.3% vs 10.3%, P < 0.001). In addition, patients with ACP ⩽ 60% were more likely to present with higher NYHA functional class (Cramer’s V [V]=0.176, P = 0.001).

Table 1.

Clinical and hemodynamic characteristics stratified by ACP.

	All (n = 290)	ACP ⩽ 60% (n = 39)	60%<ACP ⩽ 80% (n = 173)	ACP > 80% (n = 78)	P value
Clinical characteristics
Age, y	51 [35–61]	43 [34–58]	51 [35–61]	54 [37–64]	0.115
Men, n (%)	210 (72.4%)	27 (69.2%)	123 (71.1%)	60 (76.9%)	0.565
ICM, n (%)	60 (20.7%)	6 (15.4%)	36 (20.8%)	18 (23.1%)	0.625
Atrial fibrillation, n (%)	95 (32.8%)	11 (28.2%)	60 (34.7%)	24 (30.8%)	0.671
Hypertension, n (%)	71 (24.5%)	5 (12.8%)	33 (19.1%)	33 (42.3%)	<0.001
Diabetes, n (%)	45 (15.5%)	4 (10.3%)	23 (13.3%)	18 (23.1%)	0.087
Hyperlipidemia, n (%)	74 (25.5%)	12 (30.8%)	38 (22.0%)	24 (30.8%)	0.241
NYHA functional class					0.001
Ⅱ, n (%)	65 (22.4%)	5 (12.8%)	31 (17.9%)	29 (37.2%)
Ⅲ, n (%)	120 (41.4%)	13 (33.3%)	79 (45.7%)	28 (35.9%)
Ⅳ, n (%)	105 (36.2%)	21 (53.8%)	63 (36.4%)	21 (26.9%)
Medications
Beta blocker, n (%)	247 (85.2%)	31 (79.5%)	147 (85.0%)	69 (88.5%)	0.433
ACEI/ARB/ARNI, n (%)	114 (39.3%)	7 (17.9%)	66 (38.2%)	41 (52.6%)	0.001
Loop diuretic, n (%)	265 (91.4%)	37 (94.9%)	162 (93.6%)	66 (84.6%)	0.044
Digoxin, n (%)	124 (42.8%)	20 (51.3%)	81 (46.8%)	23 (29.5%)	0.019
Dopamine, n (%)	76 (26.2%)	19 (48.7%)	49 (28.3%)	8 (10.3%)	<0.001
Laboratory values
Hemoglobin, g/l	137 ± 22	131 ± 23	136 ± 22	142 ± 20	0.027
Creatinine, μmol/l	93 [78–116]	96 [82–123]	90 [73–112]	100 [84–124]	0.056
NT-proBNP, pg/ml	2877 [894–6367]	7314 [4334–13,507]	2949 [1306–5676]	834 [352–3207]	<0.001
Echocardiography
LVEF, %	37 [28–55]	33 [27–48]	35 [25–52]	48 [32–60]	0.001
Hemodynamics
CO, l/min	3.9 [2.9–4.4]	2.6 [1.9–3.8]	3.7 [2.9–4.3]	4.3 [4.0–5.4]	<0.001
SV, ml	52 [37–65]	33 [25–54]	51 [36–63]	61 [49–79]	<0.001
HR, bpm	73 [65–83]	75 [66–87]	72 [65–82]	74 [66–84]	0.583
SAP, mmHg	102 [94–112]	93 [87–97]	99 [94–107]	122 [111–128]	<0.001
MAP, mmHg	79 [74–86]	73 [69–77]	77 [73–82]	92 [86–100]	<0.001
DAP, mmHg	68 [62–74]	63 [59–68]	65 [60–72]	78 [70–85]	<0.001
SVR, WU	18.8 [15.6–25.0]	21.6 [13.7–30.0]	18.6 [15.7–25.0]	19.1 [15.6–23.7]	0.563
RAP, mmHg	8 [4–13]	18 [14–24]	8 [4–12]	5 [2–9]	<0.001
sPAP, mmHg	39 [27–52]	44 [35–62]	41 [28–54]	32 [22–45]	<0.001
mPAP, mmHg	27 [18–37]	33 [25–43]	29 [20–38]	23 [14–30]	<0.001
dPAP, mmHg	20 [12–27]	27 [16–33]	20 [13–28]	14 [9–22]	<0.001
PAWP, mmHg	18 [11–26]	26 [17–33]	20 [12–27]	12 [8–19]	<0.001
PVR, Wood	2.3 [1.2–3.5]	3.2 [2.3–4.4]	2.5 [1.4–3.6]	1.5 [0.9–2.9]	<0.001
PAC, ml/mmHg	2.5 [1.6–4.2]	1.5 [1.3–2.6]	2.4 [1.5–3.5]	3.9 [2.3–6.0]	<0.001
ACP, %	71.9 [64.2–80.7]	56.5 [53.3–58.8]	70.6 [65.3–74.3]	88.1 [83.3–94.0]	<0.001
CPO, W	0.67 [0.50–0.82]	0.41 [0.32–0.58]	0.61 [0.49–0.74]	0.90 [0.79–1.09]	<0.001
PAPI	2.6 [1.5–4.8]	1.1 [0.9–1.8]	2.6 [1.7–4.6]	3.5 [2.0–8.3]	<0.001

ACEI, angiotensin-converting enzyme inhibitor; ACP, afterload-related cardiac performance; ARB, angiotensin-receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; CO, cardiac output; CPO, cardiac power output; DAP, diastolic arterial pressure; dPAP, diastolic pulmonary arterial pressure; HR, heart rate; ICM, ischemic cardiomyopathy; LVEF, left ventricular ejection fraction; MAP, mean arterial pressure; mPAP, mean pulmonary arterial pressure; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; PAC, pulmonary arterial compliance; PAPI, pulmonary artery pulsatility index; PAWP, pulmonary arterial wedge pressure; PVR, pulmonary vascular resistance; RAP, right atrial pressure; SAP, systolic arterial pressure; sPAP, systolic pulmonary arterial pressure; SV, stroke volume; SVR, systemic vascular resistance; SVR, WU→SVR, Wood.

Regarding laboratory tests and echocardiography, patients with ACP ⩽ 60% had higher serum NT-proBNP values (P < 0.001) as well as lower LVEF (P = 0.001). Conventional hemodynamic indexes also exhibited significant differences among the three groups. Patients with ACP ⩽ 60% had lower CO and SV, lower blood pressure, lower CPO, lower PAC and PAPI, higher RAP, higher PAP, higher PAWP, and higher PVR (all P < 0.001). HR and SVR did not differ among the three groups (all P > 0.05).

We also analyzed the correlation between ACP and traditional hemodynamic parameters reflecting cardiac performance (Figure 2). ACP and SV were significantly correlated (r = 0.4669, P < 0.001). Furthermore, there was a strong correlation between ACP and CO (r = 0.5773, P < 0.001). The correlation between ACP and CPO was even more significant (r = 0.7730, P < 0.001).

Figure 2.

Correlation of ACP with stroke volume (a), cardiac output (b), and cardiac power output (c).

Clinical outcomes associated with ACP

Complete follow-up data were available in 263 of 290 patients (90.7%). During the follow-up period of 393 (149–831) days, 47 (17.9%) patients died, 63 (24.0%) patients underwent heart transplantation, 6 (2.3%) patients received an LVAD, 52 (19.8%) patients were rehospitalized for HF, and 95 (36.1%) patients were event-free at the end of follow-up. Kaplan–Meier survival analysis revealed a significant difference in event-free survival and all-cause mortality among the three groups (all P < 0.001; Figure 3). Patients with ACP ⩽ 60% tended to experience worse clinical outcomes.

Figure 3.

Survival analyses. Kaplan–Meier estimates of time to all-cause mortality (a), all-cause mortality censored (b), and event-free survival (c) stratified by ACP.

In univariable Cox regression analysis, ACP was independently associated with all-cause mortality as well as event-free survival (all P < 0.001). After multivariate adjustment, ACP remained significantly associated with both primary outcome (hazard ratio (HR) 0.956, 95% confidence interval (CI) 0.927–0.987) and secondary outcome (HR 0.977, 95% CI 0.963–0.992; Table 2). Lower ACP was independently associated with increased all-cause mortality across most subgroups. Although the predictive value of ACP was not statistically significant in patients with ICM, these patients with lower ACP were still more likely to reach a primary outcome (HR 0.932, P = 0.080). No significant interactions were identified between ACP as a continuous variable and age, sex, NYHA functional class, LVEF, NT-proBNP, CO, MAP, RAP, SVR, ICM, AF, and hypertension (Figure 4).

Table 2.

Univariate and multivariate Cox regression results for primary and secondary outcomes.

	All-cause mortality				All-cause mortality (censored)^a				Event-free survival
	Univariate		Multivariate		Univariate		Multivariate		Univariate		Multivariate
	HR (95%CI)	P value	HR (95% CI)	P value	HR (95% CI)	P value	HR (95% CI)	P value	HR (95% CI)	P value	HR (95% CI)	P value
Age, y	1.013 (0.995–1.032)	0.160			1.009 (0.992–1.027)	0.306			1.001 (0.992–1.011)	0.753
Sex, male versus female	0.677 (0.370–1.238)	0.205			0.715 (0.391–1.308)	0.277			1.199 (0.847–1.697)	0.305
NYHA Class, IV versus II–III	3.136 (1.749–5.622)	<0.001	2.004 (1.108–3.626)	0.022	3.993 (2.212–7.207)	<0.001	2.510 (1.367–4.610)	0.003	2.868 (2.080–3.955)	<0.001	2.100 (1.505–2.930)	<0.001
LVEF, %	0.986 (0.968–1.004)	0.126			0.970 (0.951–0.989)	0.003	–	0.106	0.972 (0.962–0.982)	<0.001	0.982 (0.971–0.992)	0.001
Log NT-proBNP, pg/ml	10.777 (5.455–21.290)	<0.001	7.300 (3.434–15.517)	<0.001	11.335 (5.858–21.932)	<0.001	7.268 (3.445–15.334)	<0.001	2.883 (2.176–3.820)	<0.001	1.710 (1.221–2.393)	0.002
History of ICM	0.977 (0.472–2.025)	0.951			0.891 (0.430–1.848)	0.757			0.994 (0.677–1.459)	0.975
History of AF	1.112 (0.613–2.017)	0.726			1.243 (0.685–2.254)	0.475			1.591 (1.164–2.174)	0.004	1.403 (1.017–1.935)	0.039
History of hypertension	0.523 (0.234–1.168)	0.114			0.466 (0.209–1.042)	0.063	–	0.252	0.726 (0.503–1.048)	0.088	–	0.129
History of diabetes	0.626 (0.248–1.582)	0.322			0.582 (0.230–1.472)	0.253			1.145 (0.771–1.701)	0.501
History of hyperlipidemia	0.797 (0.396–1.603)	0.524			0.763 (0.379–1.536)	0.449			1.043 (0.737–1.477)	0.812
Use of dopamine	1.642 (0.910–2.964)	0.100	–	0.484	2.431 (1.346–4.391)	0.003	–	0.860	2.453 (1.786–3.368)	<0.001	–	0.260
CO, l/min	0.707 (0.551–0.907)	0.006	–	0.574	0.647 (0.502–0.834)	0.001	–	0.818	0.734 (0.650–0.830)	<0.001	–	0.894
Heart rate, bpm	1.021 (1.001–1.042)	0.037	–	0.484	1.020 (1.001–1.040)	0.042	–	0.392	1.010 (0.999–1.022)	0.087	–	0.805
MAP, mmHg	0.941 (0.908–0.975)	0.001	–	0.595	0.937 (0.906–0.970)	<0.001	–	0.796	0.971 (0.956–0.986)	<0.001	–	0.584
SVR, WU	1.005 (0.974–1.037)	0.753			1.013 (0.982–1.044)	0.415			1.019 (1.003–1.035)	0.019	–	0.796
RAP, mmHg	1.084 (1.043–1.126)	<0.001	–	0.981	1.085 (1.044–1.128)	<0.001	–	0.910	1.047 (1.024–1.071)	<0.001	–	0.698
mPAP, mmHg	1.028 (1.006–1.050)	0.013	–	0.663	1.037 (1.014–1.059)	0.001	–	0.842	1.035 (1.023–1.047)	<0.001	–	0.548
PAWP, mmHg	1.039 (1.012–1.067)	0.004	–	0.573	1.056 (1.027–1.086)	<0.001	–	0.952	1.048 (1.033–1.064)	<0.001	–	0.745
PVR, Wood	1.058 (0.941–1.189)	0.349			1.085 (0.972–1.211)	0.148			1.101 (1.040–1.164)	0.001	–	0.929
PAC, ml/mmHg	0.686 (0.545–0.864)	0.001	–	0.702	0.635 (0.503–0.802)	<0.001	–	0.322	0.759 (0.689–0.835)	<0.001	0.895 (0.812–0.987)	0.026
ACP, %	0.919 (0.892–0.948)	<0.001	0.956 (0.927–0.987)	0.005	0.911 (0.884–0.939)	<0.001	0.951 (0.921–0.982)	0.002	0.956 (0.942–0.969)	<0.001	0.977 (0.963–0.992)	0.003
CPO, W	0.093 (0.023–0.379)	0.001	–	0.637	0.053 (0.012–0.224)	<0.001	–	0.888	0.167 (0.088–0.317)	<0.001	–	0.896
PAPI	0.866 (0.766–0.979)	0.021	–	0.812	0.872 (0.773–0.984)	0.027	–	0.931	0.993 (0.963–1.023)	0.628

ACP, afterload-related cardiac performance; AF, atrial fibrillation; CI, confidence interval; CO, cardiac output; CPO, cardiac power output; HR, heart rate; ICM, ischemic cardiomyopathy; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; MAP, mean arterial pressure; mPAP, mean pulmonary arterial pressure; NT-proBNP, N-terminal pro-B-type natriuretic peptide; PAC, pulmonary arterial compliance; PAPI, pulmonary artery pulsatility index; PAWP, pulmonary arterial wedge pressure; PVR, pulmonary vascular resistance; RAP, right atrial pressure; SVR, systemic vascular resistance; SVR, WU→SVR, Wood.

All-cause mortality was censored at the time of heart transplantation or LVAD implantation.

Figure 4.

Subgroup analyses for associations between ACP and all-cause mortality.

ACP was significantly more discriminating than CO, CPO, RAP, mPAP, PAWP, and PAC for the prediction of all-cause mortality, with an AUC of 0.770 for ACP (95% CI: 0.701–0.840), 0.666 for CO (95% CI: 0.581–0.751), 0.688 for CPO (95% CI: 0.604–0.771), 0.678 for RAP (95% CI: 0.592–0.763), 0.618 for mPAP (95% CI: 0.532–0.703), 0.640 for PAWP (95% CI: 0.555–0.726), and 0.678 for PAC (95% CI: 0.601–0.755; Delong test, P = 0.017, 0.019, 0.044, 0.005, 0.012, and 0.038, respectively, for the comparison with the ACP receiver operating characteristic curve; Figure 5).

Figure 5.

Receiver operating characteristic curves for prediction of all-cause mortality.

Discussion

To the best of our knowledge, this was the first study to analyze the prognostic value of ACP in chronic HF and was also the first time to establish the expected CO-SVR curve model in a cohort of patients with chronic HF. Our results demonstrated that this relatively old but powerful parameter was strongly associated with adverse clinical outcomes and exhibited an excellent ability to predict mortality.

Patients with HF are often characterized by decreased CO from a variety of causes. The decrease in CO activates a series of compensatory adaptations, including activation of the adrenergic nervous systemic system and the renin-angiotensin system, leading to vasoconstriction and an elevated SVR, thereby maintaining cardiovascular homeostasis.¹⁸ If SVR increases appropriately in response to CO decreases, the patients then are in a relative hemodynamic stabilization, suggesting that the cardiovascular function of the patients is relatively preserved. In the present study, we quantified this condition as ACP > 80%, and patients who met this criterion had lower NYHA functional class, lower serum NT-proBNP levels, a lower proportion requiring inotropes, and a significantly better prognosis. However, if CO decreases significantly while SVR only slightly increases, that is, CO_measured is much lower than CO_predicted when focusing on a given SVR value, then ACP might be less than 80% or even <60%. The data suggested that these patients with lower ACP had worse outcomes and often required intervention compared with those with higher ACP. This might be due to the imbalance between extremely low CO and insufficient increase in SVR leading to hypotension and reduced perfusion pressure of vital organs including the heart.⁹ Therefore, ACP is an advanced quantitative hemodynamic parameter that could integrate both cardiac function and vascular response, thus assessing cardiovascular function more comprehensively and accurately.

Over the past decades, our perception of the hemodynamics in HF has transformed from the traditional measurement of pressure/flow (e.g. MAP, PAWP, CO) to a more comprehensive understanding of the importance of the interaction between preload, afterload, contractility, and lusitropy. Therefore, it would make sense to establish a parameter that could represent multiple elements and enable a comprehensive assessment of hemodynamics. In essence, ACP incorporates four fundamental hemodynamic parameters (SV, HR, MAP, and RAP) and takes both cardiac performance and vascular effects into account. CPO is the product of simultaneously measured MAP and CO and is considered a powerful predictor of outcomes both in HF and cardiogenic shock patients.^15,19 In comparison with CPO, ACP demonstrated better prognostic performance for adverse outcomes. This could be explained by the fact that ACP is also corrected by RAP. In addition, ACP was found to be the only independent hemodynamic parameter correlated with all-cause mortality on multivariate analysis. Subgroup analysis further suggested that ACP was a robust parameter and that other clinical factors could hardly affect its prognostic performance. These results demonstrated that the predictive value of ACP was superior to other traditional parameters, which was our strongest reason to recommend ACP as one of the routine hemodynamic parameters for the evaluation of patients with chronic HF.

The idea to integrate both cardiac and vascular factors has provided new insights into hemodynamic assessment. Given this, we established a novel CO-SVR two-dimensional graph for a more convenient assessment of the hemodynamic status (Figure 6). By plotting the CO and SVR results of individual patients on a two-dimensional graph, clinicians could visually assess the hemodynamics of a given patient and evaluate cardiovascular function according to ACP. Of note, when SVR is significantly elevated, a slight reduction in CO would result in a substantial reduction in ACP, which may imply a poor clinical outcome. This suggested that more efforts should be made to maintain CO in patients especially when their SVR is elevated. Conversely, a compensatory increase in CO due to low SVR may mask the fact that cardiovascular function has already been impaired. Therefore, the value of CO alone as a predictor of adverse outcomes may be compromised when HF patients presented with low SVR, whereas ACP could provide a more comprehensive assessment of cardiovascular function. More importantly, the CO-SVR two-dimensional graph and ACP enable a better understanding of the pathophysiology of chronic HF and might guide physicians in making treatment decisions, such as whether to prescribe inotropes, vasodilators, or vasopressors.

Figure 6.

CO-SVR two-dimensional graph.

Overall, ACP might be a valuable parameter to assess the hemodynamic status and predict outcomes in patients with chronic HF. Nevertheless, it must be noted that the predictive value of ACP might be challenged in patients with acute decompensated HF with pulmonary edema or those with hypertensive crisis.²⁰ In these clinical scenarios, patients are no doubt in an unstable clinical condition, however, the calculated ACP would be extremely high due to the significantly increased SVR and the relatively non-significant decrease in CO. Therefore, abnormally elevated ACP might also indicate unstable hemodynamic status, which suggests a poor prognosis. In the present study, since most of the patients in our HFCU were not with acute decompensated HF or hypertensive crisis, we did not perform additional analyses on these patients and excluded them during the screening process. Undoubtedly, this is the next issue to be discussed in further investigations. Although the concept of ACP has been proposed for more than a decade, it has never been employed in the field of HF. More clinical studies are needed to verify its prognostic information and analyze whether ACP-guided therapy could improve symptoms as well as outcomes in patients with various types of HF.

Limitations

Our study has several limitations which are worth noting. First, this was a single-center retrospective study. The data used to establish the expected CO-SVR curve model were derived from a single center. Multicenter studies are warranted to reproduce the regression model and to verify the prognostic information of ACP. Second, as the calculation of SVR depends on CO, MAP, and RAP, there might be a mathematical coupling of measurement errors in shared variables if measurements were not accurate.²¹ Third, there was a small proportion (9.3%) of patients lost to follow-up in our study. Although the ACP values of patients lost to follow-up were slightly higher, their clinical characteristics, medications, and laboratory values did not differ from patients with complete follow-up data (Supplemental Table 1). Fourth, about 30% of patients were receiving inotropes at the time of measurements, which may increase ACP. However, supportive treatments were unlikely to overestimate the prognostic value of ACP. On the contrary, ACP might be a more powerful parameter to predict outcomes in the absence of any intervention. This is due to the possibility that supportive treatments may narrow the differences in ACP between patients with various degrees of cardiovascular impairment.

Conclusion

ACP is a robust hemodynamic parameter and a powerful predictor of mortality in our cohort of patients with chronic HF. By incorporating CO and SVR, it integrates both cardiac performance and vascular effects and could represent the cardiovascular function of patients in a quantitative manner. In addition, ACP and the CO-SVR two-dimensional graph demonstrate their potential ability in making clinical decisions. Future studies should validate its prognostic performance in patients with different types of HF and focus more on the feasibility of ACP-guided therapy.

Supplemental Material

sj-docx-1-taj-10.1177_20406223231171554 – Supplemental material for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure

Supplemental material, sj-docx-1-taj-10.1177_20406223231171554 for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure by Yihang Wu, Pengchao Tian, Lin Liang, Yuyi Chen, Jiayu Feng, Boping Huang, Liyan Huang, Xuemei Zhao, Jing Wang, Jingyuan Guan, Xinqing Li, Yuhui Zhang and Jian Zhang in Therapeutic Advances in Chronic Disease

Supplemental Material

sj-tif-2-taj-10.1177_20406223231171554 – Supplemental material for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure

Supplemental material, sj-tif-2-taj-10.1177_20406223231171554 for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure by Yihang Wu, Pengchao Tian, Lin Liang, Yuyi Chen, Jiayu Feng, Boping Huang, Liyan Huang, Xuemei Zhao, Jing Wang, Jingyuan Guan, Xinqing Li, Yuhui Zhang and Jian Zhang in Therapeutic Advances in Chronic Disease

Supplemental Material

sj-tif-3-taj-10.1177_20406223231171554 – Supplemental material for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure

Supplemental material, sj-tif-3-taj-10.1177_20406223231171554 for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure by Yihang Wu, Pengchao Tian, Lin Liang, Yuyi Chen, Jiayu Feng, Boping Huang, Liyan Huang, Xuemei Zhao, Jing Wang, Jingyuan Guan, Xinqing Li, Yuhui Zhang and Jian Zhang in Therapeutic Advances in Chronic Disease

Supplemental Material

sj-tif-4-taj-10.1177_20406223231171554 – Supplemental material for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure

Supplemental material, sj-tif-4-taj-10.1177_20406223231171554 for Afterload-related cardiac performance is a powerful hemodynamic predictor of mortality in patients with chronic heart failure by Yihang Wu, Pengchao Tian, Lin Liang, Yuyi Chen, Jiayu Feng, Boping Huang, Liyan Huang, Xuemei Zhao, Jing Wang, Jingyuan Guan, Xinqing Li, Yuhui Zhang and Jian Zhang in Therapeutic Advances in Chronic Disease

Footnotes

Acknowledgements

None.

Declarations

ORCID iD

Yihang Wu

Supplemental material

Supplemental material for this article is available online.

References

Thakkar

Desai

. Swan, Ganz, and their Catheter: its evolution over the past half century. Ann Intern Med 2018; 169: 636–642.

Lim

Dubois

De Backer

, et al. Do all nonsurvivors of cardiogenic shock die with a low cardiac index? Chest 2003; 124: 1885–1891.

Gardner

Henderson

McDonagh

. The prognostic use of right heart catheterization data in patients with advanced heart failure: how relevant are invasive procedures in the risk stratification of advanced heart failure in the era of neurohormones? J Heart Lung Transplant 2005; 24: 303–309.

Cooper

Mentz

Stevens

, et al. Hemodynamic predictors of heart failure morbidity and mortality: fluid or flow? J Card Fail 2016; 22: 182–189.

Aalders

Kok

. Comparison of hemodynamic factors predicting prognosis in heart failure: a systematic review. J Clin Med 2019; 8: 1757.

Parmley

. Pathophysiology of congestive heart failure. Am J Cardiol 1985; 56: 7A–11A.

Werdan

Oelke

Hettwer

, et al. Septic cardiomyopathy: hemodynamic quantification, occurrence, and prognostic implications. Clin Res Cardiol 2011; 100: 661–668.

Mahajan

, et al. Hemodynamic phenotypes of hypertension based on cardiac output and systemic vascular resistance. Am J Med 2020; 133: e127–e139.

Cotter

Moshkovitz

Kaluski

, et al. The role of cardiac power and systemic vascular resistance in the pathophysiology and diagnosis of patients with acute congestive heart failure. Eur J Heart Fail 2003; 5: 443–451.

10.

Wilhelm

Hettwer

Schuermann

, et al. Severity of cardiac impairment in the early stage of community-acquired sepsis determines worse prognosis. Clin Res Cardiol 2013; 102: 735–744.

11.

Chen

Zhang

Tao

, et al. Afterload-related cardiac performance identifies cardiac impairment and associates with outcome in patients with septic shock: a retrospective cohort study. J Intensive Care 2021; 9: 33.

12.

Yancy

Jessup

Bozkurt

, et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation 2013; 128: e240–e327.

13.

McDonagh

Metra

Adamo

, et al. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 2021; 42: 3599–3726.

14.

D’Alto

Dimopoulos

Coghlan

, et al. Right heart catheterization for the diagnosis of pulmonary hypertension: controversies and practical issues. Heart Fail Clin 2018; 14: 467–477.

15.

Fincke

Hochman

Lowe

, et al. Cardiac power is the strongest hemodynamic correlate of mortality in cardiogenic shock: a report from the SHOCK trial registry. J Am Coll Cardiol 2004; 44: 340–348.

16.

Nakagawa

Yasumura

Yoshida

, et al. Prognostic importance of pulmonary arterial capacitance in acute decompensated heart failure with preserved ejection fraction. J Am Heart Assoc 2021; 10: e023043.

17.

Zern

Wang

Rambarat

, et al. Association of Pulmonary Artery Pulsatility Index with adverse cardiovascular events across a hospital-based sample. Circ Heart Fail 2022; 15: e009085.

18.

Hartupee

Mann

. Neurohormonal activation in heart failure with reduced ejection fraction. Nat Rev Cardiol 2017; 14: 30–38.

19.

Yildiz

Aslan

Demirozu

, et al. Evaluation of resting cardiac power output as a prognostic factor in patients with advanced heart failure. Am J Cardiol 2017; 120: 973–979.

20.

Tan

Williams

Wright

. Ventriculo-arterial function curves – a new dimension in characterising acute heart failure. Eur J Heart Fail 2003; 5: 407–410.

21.

Hollenberg

Singer

. Pathophysiology of sepsis-induced cardiomyopathy. Nat Rev Cardiol 2021; 18: 424–434.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB

1.55 MB

1.58 MB

2.26 MB