CA125 but not NT-proBNP predicts the presence of a congestive intrarenal venous flow in patients with acute heart failure

Abstract

Background

Intrarenal venous flow (IRVF) measured by Doppler ultrasound has gained interest as a potential surrogate marker of renal congestion and adverse outcomes in heart failure. In this work, we aimed to determine if antigen carbohydrate 125 (CA125) and plasma amino-terminal pro-B-type natriuretic peptide (NT-proBNP) are associated with congestive IRVF patterns (i.e., biphasic and monophasic) in acute heart failure (AHF).

Methods and results

We prospectively enrolled a consecutive cohort of 70 patients hospitalized for AHF. Renal Doppler ultrasound was assessed within the first 24-h of hospital admission. The mean age of the sample was 73.5 ± 12.3 years; 47.1% were female, and 42.9% exhibited heart failure with preserved ejection fraction. The median (interquartile range) for NT-proBNP and CA125 were 6149 (3604–12 330) pg/mL and 64 (37–122) U/mL, respectively. The diagnostic performance of both exposures for identifying congestive IRVF patterns was tested using the receiving operating curve (ROC). The cut-off for CA125 of 63.5 U/mL showed a sensibility and specificity of 67% and 74% and an area under the ROC curve of 0.71. After multivariate adjustment, CA125 remained non-linearly and positively associated with congestive IRVF (P-value = 0.008) and emerged as the most important covariate explaining the variability of the model (R²: 47.5%). Under the same multivariate setting, NT-proBNP did not show to be associated with congestive IRVF patterns (P-value = 0.847).

Conclusions

CA125 and not NT-proBNP is a useful marker for identifying patients with AHF and congestive IRVF patterns.

Aims

Congestion is a hallmark feature of worsening heart failure (HF); however, its severity and organ distribution vary widely between patients.^1,2 Consequently, there is a growing interest in searching for instruments that may help physicians identify the predominant congestion phenotype and determine the adequacy and intensity of decongestive therapies.^1,3

In this context, intrarenal venous flow (IRVF) measured by Doppler ultrasound has emerged as a promising imaging tool to evaluate renal congestion^4–6; with discontinuous IRVF patterns (i.e. biphasic and monophasic) having been shown to predict a reduced diuretic response, worsening renal function, and adverse outcomes.^7–12 However, the assessment of IRVF in the often dyspneic patient during the early phase of hospital admission may be challenging and time-consuming.¹³ Therefore, we aimed to evaluate whether plasma antigen carbohydrate 125 (CA125) and the amino-terminal pro-B-type natriuretic peptide (NT-proBNP), two widely available surrogate markers of congestion,¹ were associated with discontinuous IRVF patterns in patients with AHF.

Methods

Study design and patient population

We prospectively included a consecutive cohort of 73 patients hospitalized for acute heart failure (AHF) at the Cardiology department of a tertiary care teaching hospital (Hospital Clínico Universitario de Valencia, Spain) from January 2020 to December 2020. Trained cardiologists performed the diagnosis of AHF according to current guidelines.¹⁴ Either patients with new-onset or decompensated chronic HF were eligible. The main exclusion criteria were dialysis, kidney transplantation, autosomal dominant polycystic kidney disease, and postrenal obstruction. Within the first 24 h of hospital admission, all patients underwent a complete physical examination, blood test, transthoracic echocardiography, and intrarenal Doppler ultrasonography (IRD). This study complies with the Declaration of Helsinki and was approved by the local institutional review committees. All patients provided written informed consent.

Procedures and measurements

Intrarenal Doppler ultrasonography study

After obtaining informed consent, an investigator not involved in the patient's clinical care performed the IRD at the time of echocardiography using the same ultrasound system (EPIQ 7, Philips Healthcare) with a sector transducer frequency range of 2.5–5 MHz. IRD was recorded in the right kidney with the patient in the left lateral decubitus position. The velocity range of the colour Doppler was set to 15 cm/s. Colour Doppler images were used to determine interlobar vessels, and the sample volume was set on the best-visualized colour Doppler signal. Pulsed Doppler waveforms of vein and artery were recorded simultaneously at the end of expiration during at least three cardiac cycles. Doppler waveforms were divided into continuous (non-congestive) and discontinuous (congestive) IRVF patterns. Discontinuous IRVF was defined as a pattern in which velocity at the nadir was zero.⁸ Two types of discontinuous IRVF were recorded: biphasic (with venous peaks during systole and diastole) and monophasic (with a venous peak only during diastole).

Echocardiography

Comprehensive transthoracic echocardiography was performed according to published guidelines using standard views and techniques.^15,16 Left heart parameters included left ventricular ejection fraction (LVEF), mitral inflow velocity to averaged (medial and lateral) annular relaxation velocity (E/e′), and more than moderate mitral regurgitation. Right heart parameters included basal diameter of the right ventricle, tricuspid annular plane systolic excursion, more than moderate tricuspid regurgitation (TR), inferior vena cava (IVC) diameter, and the percentage of inspiratory collapsibility, right atrial pressure (RAP), and systolic pulmonary arterial pressure (SPAP). RAP (as a surrogate of CVP) was estimated based on IVC parameters according to current recommendations.¹⁷ SPAP was estimated by adding RAP to the peak TR systolic pressure gradient.¹⁶

Congestion assessment

Dyspnoea, orthopnoea, fatigue, jugular venous pressure (JVP), rales, and pedal oedema were assessed at admission based on a standardized 4-point scale ranging from 0 to 3 as described by Ambrosy et al.¹⁸ A composite congestion score was calculated by summing the individual scores of orthopnoea (0–3), JVP (0–3), and pedal oedema (0–3).¹⁸

Laboratory data

Blood samples were taken within the first 24 h of hospital admission. The estimated glomerular filtration rate (eGFR) was calculated according to the Chronic Kidney Disease Epidemiology Collaboration equation (CKD-EPI). Plasma concentrations of NT-proBNP and CA125 were measured with the commercially available electrochemiluminescent sandwich immunoassay (Roche Elecsys^® NT-proBNP assay; Roche Elecsys^® CA 125 assay). For NT-proBNP, the intra-assay precision (coefficient of variation) is 1.2–1.5%, and the inter-assay precision (coefficient of variation) is 4.4–5.0%, with an analytical range of 5–35 000 pg/mL.¹⁹ For CA125, the intra-assay precision (coefficient of variation) is 1.4–2.0%, and the inter-assay precision (coefficient of variation) is 0.0–0.9%, with an analytical range of 0.6–5000 U/mL.²⁰

Reproducibility

To assess intraobserver variability, a single observer analysed data from 15 examinations twice: at the moment of acquisition and 4 weeks later. To test interobserver reproducibility, a single sonographer analysed data from 15 examinations performed by the other sonographer without knowing the other observer's initial interpretation.

Statistical analysis

Continuous variables were expressed as mean (± standard deviation) or medians [interquartile range (IQR)], as appropriate. Discrete variables were summarized as percentages. Baseline characteristics among IRVF categories (continuous vs. discontinuous IRVF patterns) were compared using the Student’s t-test for continuous variables and chi-square test for categorical variables.

The diagnostic performance of both exposures (CA125 and NT-proBNP) for identifying a congestive IRVF was tested using the receiving operating curve (ROC) and compared using the roccomp command. The optimal cut-point (best balance between sensitivity and specificity) was tested with the Youden method. Independent factors associated with congestive IRVF were assessed by multivariate logistic regression analyses using univariate factors with a value of P < 0.05. A final model was derived by using backward stepwise selection. The linearity assumption for all continuous variables was simultaneously tested, and the variable transformed, if appropriate, with fractional polynomials. Covariates included in the final multivariate model were blood urea nitrogen, systolic blood pressure, pedal oedema, IVC diameter, and CA125. The covariates' contribution to the model's predictability was assessed by the coefficient of determination (R²). Intra- and interobserver reproducibility was assessed as a percentage of agreeability and Cohen's kappa statistic. All analyses were performed using STATA 15.1 (StataCorp, 2017, Stata Statistical Software: Release 15; StataCorp, LLC, College Station, TX, USA).

Results

Baseline characteristics

Of the 73 eligible patients, 3 patients (4.1%) were excluded because of inadequate intrarenal venous Doppler images. Finally, 70 patients were enrolled and included in the analysis. The mean age of the sample was 74 ± 12 years, 33 (47.1%) were female, 40 (57.1%) had an LVEF <50%, and 30 (42.9%) had a prior history of chronic kidney disease (eGFR <60 mL/min/1.73 m²). The median (IQR) of eGFR, NT-proBNP, and CA125 at admission were 50 mL/min/1.73 m² (33–72), 6149 pg/mL (3604–12 330), and 64 U/mL (37–122), respectively. A total of 43 subjects (61.4%) had a discontinuous IRVF pattern at baseline (22 biphasic and 21 monophasic). Baseline characteristics across congestive and non-congestive IRVF patterns are summarized in Table 1. Overall, the group of patients with congestive IRVF patterns showed greater evidence of clinical congestion and echocardiographic features of right-sided HF. Conversely, those with non-congestive IRVF had higher blood pressure at admission. Of note, even though the presence of haemodynamically significant TR did not reach statistical significance between patients with congestive and non-congestive IRVF patterns, the proportion of patients with more than moderate TR was numerically higher in the subgroup of patients with congestive IRVF patterns (especially those with monophasic IRVF, Supplementary material online, Table S1).

Table 1.

Baseline characteristics according to intrarenal venous flow pattern

Variables	All patients (n = 70)	Non-congestive IRVF (n = 27)	Congestive IRVF (n = 43)	P-Value
Demographics and medical history
Age, years	77 [67-83]	77 [67-83]	78 [67-83]	0.772
Female sex	33 (47.1)	15 (55.6)	18 (41.9)	0.264
HTA	57 (81.4)	24 (88.9)	33 (76.7)	0.203
Dyslipidaemia	45 (64.3)	17 (63.0)	28 (65.1)	0.855
DM	46 (65.7)	20 (74.1)	26 (60.5)	0.243
Obesity	26 (37.7)	11 (40.7)	15 (35.7)	0.674
Smoker	8 (11.4)	3 (11.1)	5 (11.6)	0.947
Former smoker	26 (37.7)	11 (40.7)	15 (35.7)	0.674
Chronic kidney diseasea	30 (42.9)	12 (44.4)	18 (41.9)	0.832
CAD	18 (26)	6 (22)	12 (28)	0.596
De novo HF	34 (48.57)	16 (59.26)	18 (41.86)	0.156
HFpEF	30 (43)	13 (48)	17 (40)	0.478
NYHA previous to admission
NYHA I	32 (46)	14 (52)	18 (42)	0.414
NYHA II	32 (46)	12 (44)	20 (47)	0.866
NYHA III	6 (9)	1 (4)	5 (12)	0.249
Vital signs
SBP, mmHg	140 [122–163]	152 [132–175]	133 [118–158]	0.037
DBP, mmHg	80 [69–91]	89 [75–92]	73 [65–90]	0.024
HR, b.p.m.	82 [70–104]	83 [68–105]	82 [70–104]	1.00
Clinical presentation
NYHA on admission
NYHA II	2 (2.9)	0 (0.0)	2 (4.7)	0.256
NYHA III	34 (48.6)	11 (40.7)	23 (53.5)	0.299
NYHA IV	34 (48.6)	16 (59.3)	18 (41.9)	0.156
Rales	69 (98.6)	26 (96.3)	43 (100)	0.204
Oedema	42 (60.0)	9 (33.3)	33 (76.7)	<0.001
JVP ≥10 cmH2O	25 (35.7)	5 (18.5)	20 (46.5)	0.017
CCS on admission	5 [4–6]	4 [4–5]	6 [4–6]	0.082
Electrocardiogram
Atrial fibrillation	33 (47)	9 (33)	24 (56)	0.067
LBBB	14 (20)	6 (22)	8 (19)	0.713
Echocardiography
LVEF, %	42 [29–61]	46 [33–59]	35 [28–63]	0.282
Mitral E/e′	18.2 [13.9–24]	18.6 [15.8–22.2]	17.3 [13.3–25]	0.339
MR > moderate	19 (27.1)	6 (22.2)	13 (30.2)	0.463
TR > moderate	16 (22.9)	4 (14.8)	12 (27.9)	0.204
Estimated SPAP, mmHg	50 [40–62]	46 [40–50]	55 [42–66]	0.049
Basal RV diameter	39 [36–45]	37 [35–39]	42 [37–46]	0.007
TAPSE, mm	17 [15–20]	19 [16–20]	16 [14–21]	0.118
IVC diameter, mm	24 [20–27]	22 [18–24]	25 [22–28]	0.001
Collapsibility of IVC <50%	55 (78.6)	16 (59.3)	39 (90.7)	0.002
Estimated RAP
Normal (3 mmhg)	13 (18.6)	9 (33.3)	4 (9.3)	0.012
Mildly elevated (8 mmHg)	12 (17.1)	6 (22.2)	6 (13.9)	0.372
Severely elevated (15 mmHg)	45 (64.3)	12 (44.4)	33 (76.7)	0.006
Laboratory
Creatinine, mg/dL	1.28 [0.94–1.72]	1.2 [0.82–1.64]	1.31 [0.99–1.76]	0.320
eGFR, mL/min/1,73 m2	51 [33–72]	54 [33–73]	48 [33–72]	0.608
Urea, mg/dL	54 [40–82]	48 [39–61]	61[40–93]	0.115
Sodium, mEq	141 [138–143]	141 [138–144]	141 [138–143]	0.442
Potassium, mEq	4.0 [3.5–4.4]	3.9 [3.4–4.4]	4.0 [3.5–4.4]	0.515
Haemoglobin, mg/dL	12.3 [10.8–13.3]	12.6 [11.3–13.5]	12.1 [10.5–13.2]	0.294
Haematocrit, %	39 [34–43]	40 [34–44]	39 [33–43]	0.359
NTproBNP, pg/mL	5018 [3160–10673]	4582 [2788–10194]	5755 [3577–12122]	0.264
CA125, U/mL	64 [37–122]	39 [21–78]	107 [46–140]	<0.001

Data are expressed as number (%) or median [interquartile range] as appropriate.

CA125, antigen carbohydrate 125; CAD, coronary artery disease; CCS, composite congestion score; DBP, diastolic blood pressure, DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HF, heart failure; HFpEF, heart failure with preserved ejection fraction; HR, heart rate; HTA, hypertension; IRVF: intrarenal venous flow; LBBB, left bundle branch block, LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; MR, mitral regurgitation; RAP, right atrial pressure; RV, right ventricle; SBP, systolic blood pressure; SPAP, systolic pulmonary arterial pressure; TAPSE, tricuspid annular plane systolic excursion; TR, tricuspid regurgitation.

^aChronic kidney disease was defined as eGFR <60 mL/min/1.73 m² using creatinine obtained at the latest available outpatient visit prior to admission and assessed during a stable phase of the disease.

Association between CA125, NT-proBNP, and other surrogates of increased CVP with congestive IRVF

In univariate analysis, the group of patients with congestive IRVF patterns showed significantly higher median values of CA125 (107 U/mL [46–140] vs. 39 U/mL [21–78], P < 0.001), but not NTproBNP (5755 pg/mL [3577–12 122] vs. 4582 pg/mL [2788–10 194], P = 0.264). In fact, we found a stepwise increase in CA125 values when moving from normal to greater severity of renal congestion (continuous, biphasic, and monophasic IVFR patterns), as is shown in Figure 1. Similarly, a positive association between CA125 and other echo metrics of right-sided HF was also observed (Supplementary material online, Figure S1). The ROC curves of CA125 and NTproBNP for identifying a congestive IRVF were 0.746 vs. 0.580, respectively (P-value = 0.071). The cut-off for CA125 of 63.5 U/mL showed a sensitivity of 67%, a specificity of 74%, and an area under the ROC curve of 0.71 (Supplementary material online, Figure S1). After multivariate adjustment, CA125 remained non-linearly and positively associated with a congestive IRVF pattern (Figure 2) and emerged as the most important covariate explaining the variability of the model (R²: 47.5%), followed by pedal oedema (positively associated, R²: 26.6%) and systolic blood pressure (inversely associated, R²: 12.1%). Table 2 shows the final multivariate model. Plots depicting the association between covariates and congestive IRVF are shown in Figure 3. Of note, IVC diameter was not an independent determinant of congestive IRVF (P = 0.103). The area under the ROC curve of the multivariate model was 0.893. Under the same multivariate setting, NT-proBNP was not associated with a congestive IRVF (Figure 4).

Figure 1.

Antigen carbohydrate 125 values according to intrarenal venous flow patterns. Values are expressed as median (interquartile range).

Figure 2.

Antigen carbohydrate 125 and congestive intrarenal venous flow. Functional form.

Table 2.

Multivariate logistic regression analyses for predictors of congestive intrarenal venous flow pattern patterns

Variables	odds ratio (OR) (95% confidence interval (CI))	P-Value
BUN, mg/dL	3.21 (0.71–14.39)	0.127
SBP, mmHg	0.97 (0.94–0.99)	0.019
Pedal oedema	5.31 (1.26–22.34)	0.022
IVC diameter, mm	1.00 (0.99–1.00)	0.103
CA125, U/mL	1.39 (1.08–1.77)	0.008

BUN, blood urea nitrogen; CA125, antigen carbohydrate 125; IVC, inferior vena cava; SBP, systolic blood pressure.

Figure 3.

Functional form of the association between (a) inferior vena cava diameter, (b) systolic blood pressure, and (c) blood urea nitrogen and congestive intrarenal venous flow.

Figure 4.

Amino-terminal pro-B-type natriuretic peptide and congestive intrarenal venous flow. Functional form.

Reproducibility

Classifications of IRVF patterns were consistent between intraobserver (% of agreement: 100%; kappa = 1.00, P < 0.001) and interobserver (% of agreement: 93.3%; kappa = 0.90, P < 0.001) assessments.

Discussion

In this nonselected AHF population, we observed three major findings. First, CA125 emerged as an independent predictor of congestive IRVF at admission. The relationship was positive and non-linear, with a stronger association in patients with CA125 plasma levels >63.5 U/mL. Second, this Doppler-derived intrarenal haemodynamic phenotype was a prevalent finding encountered in about two-third of patients during the first 24 h of hospital admission. Third, the contribution of established surrogates of increased CVP and myocardial stretch, such as IVC diameter and NT-proBNP, respectively, were marginal or not independently associated with congestive IRVF.

IRVF, NTproBNP, and echo-surrogates of increased CVP

The potential usefulness of renal Doppler ultrasonography as a surrogate of renal congestion in HF is based on the assumption that increased renal interstitial pressure caused by elevated CVP may reduce renal parenchymal compliance around intrarenal venous vessels, dampening the normal continuous flow to discontinuous patterns.^4,5 However, there is limited available data to support this notion in the AHF setting. To the best of our knowledge, this is the first study evaluating the association between commonly used metrics of congestion with IRVF during the early phase of admission for AHF. In this cohort, and consistent with previous studies,^7,10 congestive IRVF patterns were more frequently observed in patients with clinical and echocardiographic proxies of right-sided HF. However, IVC diameter and NTproBNP as metrics of increased CVP and cardiac filling pressures, respectively, were not independently associated with congestive IRVF patterns. Several reasons may explain this lack of association. First, because the kidneys are encapsulated organs, renal venous outflow does not depend exclusively on CVP but also on extrinsic factors that may exert extrarenal compression (e.g. ascites, visceral oedema, engorgement of splanchnic circulation).^21,22 Second, the traditional echo-derived surrogates of right-sided filling pressures are influenced by the severity of tricuspid regurgitation, atrial fibrillation, and myocardial compliance.¹⁷ Therefore, its presence does not necessarily indicate a state of systemic venous congestion. Third, IRVF might become disrupted because of sympathetically mediated reductions in systemic and intrarenal venous capacitance irrespective of other metrics indicative of increased cardiac filling pressures.^23,24 Thus, discontinuous IRVF might not be an exact representation of elevated CVP, especially in the acute setting. Fourth, some patients presenting with the phenotype characterized by acute redistribution of blood to the lung vasculature in the context of increased left-sided filling pressures may have elevated NT-proBNP values without being truly volume overloaded. Fifth, in addition to its association with intravascular and intracardiac pressures, NT-proBNP plasma levels are influenced by several other non-haemodynamic factors such as age, obesity, and renal function.²⁵

IRVF and CA125

In contrast to NTproBNP and IVC diameter, CA125 was independently associated with congestive IRVF. Although the exact mechanism behind CA125 up-regulation in AHF is not well characterized, one hypothesis is that elevated hydrostatic pressure and inflammatory cytokines in the setting of congestion activate mesothelial cells in serosal surfaces leading to CA125 overproduction and release.²⁶ In fact, plasma levels of this glycoprotein are elevated in almost two-third of patients with AHF²⁷ and are positively correlated to proxies of right-sided HF²⁵ and to the intensity of congestion.²⁸ The present study adds new evidence to this field by showing that CA125 emerged as the most important predictor of congestive IRVF, accounting for roughly 47.5% of the model variability. Our clinical observations raise the hypothesis that CA125 might be capturing two crucial mechanisms involved in renal venous hypertension and, thus, in the development of discontinuous IRVF patterns (Figure 5): (i) increased intravascular hydrostatic pressures in the setting of systemic venous congestion and (ii) elevated intra-abdominal pressure secondary to visceral oedema and subclinical ascites.^21,29

Figure 5.

Graphical scheme of the proposed mechanisms behind the association between antigen carbohydrate 125 and congestive intrarenal venous flow.

Clinical implications

Beyond its biological plausibility, present findings have relevant clinical implications. First, elevated CA125 levels at admission may help identify a subgroup of patients with a higher probability of renal congestion. Therefore, and considering that renal venous pressure (RVP) is a major determinant of renal blood flow (renal perfusion pressure [mean arterial pressure − RVP]/renal vascular resistance), higher CA125 may be useful to identify patients who benefit from a more aggressive diuretic strategy.²⁸ Indeed, current findings may be interpreted as proof of concept explaining the results of a recent randomized clinical trial endorsing the utility of CA125 for tailoring the intensity of diuretic therapy in patients with AHF and renal dysfunction on admission.²⁸ Second, the observed cut-off point of 63.5 U/mL for identifying congestive IRVF patterns is closely related to the cut-point associated with an increased risk of 6-month mortality.³⁰ Third, although point-of-care ultrasound applications are emerging as promising tools for phenotyping and monitoring HF patients, there are still logistic limitations that preclude its widespread use in daily clinical practice. Conversely, CA125 is widely available and low cost and has an independent predictive ability beyond clinical and biochemical markers of congestion (including natriuretic peptides).²⁶

Limitations

Several limitations need to be acknowledged. First, this study has the inherent limitations of being a single-center observational study. Second, due to the limited sample size, some of the negative results could be explained by type II error (insufficient statistical power). This issue also precludes confidently evaluating the contribution of other clinical/echocardiographic proxies of right-sided HF to predict a congestive IRVF pattern. Third, we included only patients with AHF, so our conclusions do not apply to patients with stable chronic HF. Fourth, we did not perform an invasive haemodynamic assessment, so we cannot establish direct correlations between IRVF patterns and invasive right-sided filling pressures. Fifth, we did not measure intra-abdominal pressure directly, so its potential contribution is merely speculative. Sixth, although experienced sonographers performed all the IRD studies and echocardiograms, they did not evaluate the same patient twice at the same time point. Thus, imaging acquisition reliability was not assessed. Finally, IRD studies were not reviewed by an independent core laboratory.

Conclusions

In patients with AHF, CA125 rather than NTproBNP is a useful biomarker for identifying congestive IRVF patterns. Further studies should focus on establishing well-validated risk factors (and potentially risk scores) for identifying patients at risk of ‘congestive nephropathy’ in whom IRVF assessment might be recommended.

Supplementary material

Supplementary material is available at European Heart Journal: Acute Cardiovascular Care online.

Funding

This work was supported by grants from the Ministry of Economy and Competitiveness, Instituto Carlos III (PI20/00392), CIBER Cardiovascular (16/11/00420 and 16/11/00403), and Heart Failure Association of the Spanish Society of Cardiology (2019).

Conflict of interest: none declared.