fphys-12-708890 October 18, 2021 Time: 16:16 # 1
ORIGINAL RESEARCH
published: 22 October 2021
doi: 10.3389/fphys.2021.708890
A 3-Biomarker 2-Point-Based Risk
Stratification Strategy in Acute Heart
Failure
Jesús Álvarez-García1,2* , Álvaro García-Osuna3, Miquel Vives-Borrás1,
Andreu Ferrero-Gregori1, Manuel Martínez-Sellés4, Rafael Vázquez5,
José R. González-Juanatey6, Miguel Rivera7, Javier Segovia8, Domingo Pascual-Figal9,
Ramón Bover10, Ramón Bascompte11, Juan Delgado12, Andrés Grau Sepúlveda13,
Alfredo Bardají14, Félix Pérez-Villa15, José Luis Zamorano2, Marisa Crespo-Leiro16,
Pedro Luis Sánchez17, Jordi Ordoñez-Llanos3† and Juan Cinca1†
Edited by: on behalf of the Investigators of the Spanish Heart Failure Network (REDINSCOR II)
Rui Plácido,
University of Lisbon, Portugal 1 Cardiology Department, Hospital de la Santa Creu i Sant Pau, IIb-SantPau, Centro de Investigación en Red en
Reviewed by: Enfermedades Cardiovasculares (CIBERCV), Barcelona, Spain, 2 Cardiology Department, Hospital Ramón y Cajal, Centro de
Pietro Minuz, Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain, 3 Biochemistry Department, Hospital
University of Verona, Italy de la Santa Creu i Sant Pau, IIb-SantPau, Barcelona, Spain, 4 Cardiology Department, Hospital Universitario Gregorio
Catarina Gomes, Marañón, Centro de Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain, 5 Cardiology
Hospital de Santo António, Portugal Department, Hospital Puerta del Mar, Centro de Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Cádiz,
Spain, 6 Cardiology Department, Hospital Clínico, Centro de Investigación en Red en Enfermedades Cardiovasculares
*Correspondence:
(CIBERCV), Santiago de Compostela, Spain, 7 Cardiology Department, Hospital La Fe, Centro de Investigación en Red en
Jesús Álvarez-García
Enfermedades Cardiovasculares (CIBERCV), Valencia, Spain, 8 Cardiology Department, Hospital Puerta
jalvarezg82@gmail.com
de Hierro-Majadahonda, Centro de Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain,
†These authors have contributed 9 Cardiology Department, Hospital Virgen de la Arrixaca, Centro de Investigación en Red en Enfermedades Cardiovasculares
equally to this work and share senior (CIBERCV), Murcia, Spain, 10 Cardiology Department, Hospital Clínico San Carlos, Centro de Investigación en Red en
authorship Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain, 11 Cardiology Department, Hospital Arnau de Vilanova, Centro
de Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Lleida, Spain, 12 Cardiology Department, Hospital
Specialty section: 12 de Octubre, Centro de Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain, 13 Cardiology
This article was submitted to Department, Hospital Universitario Son Espases, Centro de Investigación en Red en Enfermedades Cardiovasculares
Clinical and Translational Physiology, (CIBERCV), Palma de Mallorca, Spain, 14 Cardiology Department, Hospital Juan XXIII, Centro de Investigación en Red en
a section of the journal Enfermedades Cardiovasculares (CIBERCV), Tarragona, Spain, 15 Cardiology Department, Hospital Clinic, Centro de
Frontiers in Physiology Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), Barcelona, Spain, 16 Cardiology Department, Hospital
Universitario A Coruna, Centro de Investigación en Red en Enfermedades Cardiovasculares (CIBERCV), A Coruna, Spain,
Received: 12 May 2021
17 Cardiology Department, Hospital Clínico Universitario, Centro de Investigación en Red en Enfermedades Cardiovasculares
Accepted: 04 October 2021
(CIBERCV), Salamanca, Spain
Published: 22 October 2021
Citation:
Álvarez-García J, García-Osuna Á, Introduction and Objectives: Most multi-biomarker strategies in acute heart failure
Vives-Borrás M, Ferrero-Gregori A, (HF) have only measured biomarkers in a single-point time. This study aimed to evaluate
Martínez-Sellés M, Vázquez R,
González-Juanatey JR, Rivera M, the prognostic yielding of NT-proBNP, hsTnT, Cys-C, hs-CRP, GDF15, and GAL-3 in HF
Segovia J, Pascual-Figal D, Bover R, patients both at admission and discharge.
Bascompte R, Delgado J,
Grau Sepúlveda A, Bardají A, Methods: We included 830 patients enrolled consecutively in a prospective multicenter
Pérez-Villa F, Zamorano JL, registry. Primary outcome was 12-month mortality. The gain in the C-index, calibration,
Crespo-Leiro M, Sánchez PL,
Ordoñez-Llanos J and Cinca J (2021) net reclassification improvement (NRI), and integrated discrimination improvement (IDI)
A 3-Biomarker 2-Point-Based Risk was calculated after adding each individual biomarker value or their combination on top
Stratification Strategy in Acute Heart
Failure. Front. Physiol. 12:708890. of the best clinical model developed in this study (C-index 0.752, 0.715–0.789) and also
doi: 10.3389/fphys.2021.708890 on top of 4 currently used scores (MAGGIC, GWTG-HF, Redin-SCORE, BCN-bioHF).
Frontiers in Physiology | www.frontiersin.org 1 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 2
Álvarez-García et al. Multimarker in Acute Heart Failure
Results: After 12-month, death occurred in 154 (18.5%) cases. On top of the best
clinical model, the addition of NT-proBNP, hs-CRP, and GDF-15 above the respective
cutoff point at admission and discharge and their delta during compensation improved
the C-index to 0.782 (0.747–0.817), IDI by 5% (p < 0.001), and NRI by 57% (p < 0.001)
for 12-month mortality. A 4-risk grading categories for 12-month mortality (11.7,
19.2, 26.7, and 39.4%, respectively; p < 0.001) were obtained using combination of
these biomarkers.
Conclusion: A model including NT-proBNP, hs-CRP, and GDF-15 measured at
admission and discharge afforded a mortality risk prediction greater than our clinical
model and also better than the most currently used scores. In addition, this 3-biomarker
panel defined 4-risk categories for 12-month mortality.
Keywords: biomarker (BM), panel (C33), acute heart failure (AHF), risk stratification, prognosis
INTRODUCTION However, the data derived from clinical trials might not entirely
reflect the daily real practice.
The prognostic stratification of patients with acute heart failure Therefore, we aimed to analyze the performance of a multi-
(HF) is essential to establish an appropriate personalized follow- biomarker panel covering distinct pathophysiological axes in
up plan. Cardiac biomarkers have improved the predictive HF measured both at admission and at hospital discharge, in a
models of HF patients beyond the already well-established nationwide cohort of patients with acute HF (REDINSCOR II
clinical risk predictors such as functional class or physical registry). We have selected the N-terminal pro-B-type natriuretic
examination (Levy et al., 2006). peptide (NT-proBNP) as marker of neurohormonal activation
More than 10 years ago, Braunwald provided a and myocyte stretch, high-sensitive T troponin (hs-TnT) linked
comprehensive review of the biomarkers related to the different to myocyte injury, cystatin-C (Cys-C) as indicative of renal
pathophysiological substrates involved in HF (Braunwald, dysfunction, growth differentiation factor 15 (GDF-15) and
2008), and remarked the need to identify the biomarkers with galectin-3 (GAL-3) as markers of matrix remodeling, and high
independent predictive value in large prospective cohorts of sensitive C reactive protein (hs-CRP) as marker of inflammation.
patients. Until now, a substantial number of studies have assessed
the prognostic capacity of panels of 3 or more biomarkers in
acute HF (Ishii et al., 2002; van Kimmenade et al., 2006; Januzzi MATERIALS AND METHODS
et al., 2007; Rehman et al., 2008; Manzano-Fernández et al., 2009,
2011; Niizeki et al., 2009; Zairis et al., 2010; Pascual-Figal et al.,
2011; Shah et al., 2012; Bjurman et al., 2013; Lassus et al., 2013; Study Population
Lok et al., 2013; Richter et al., 2013; Srinivas et al., 2014; Demissei This is a subanalysis including 830 patients discharged alive with
et al., 2016, 2017a,b; Herrero-Puente et al., 2017; Tromp et al., available biomarker data both at admission and discharge from
2017), but only in few of them the biomarkers were analyzed at the REDINSCOR II study. This is a multicenter, prospective
both hospital admission and discharge (Demissei et al., 2016, nationwide registry, which enrolled consecutively patients from
2017a,b). A single-point measurement would not allow to 18 secondary and tertiary hospitals since October 2013 to
evaluate the width of the pathophysiological changes occurring December 2014. Inclusion criteria were: (i) age older than
during the clinical compensation and, moreover, might limit the 18 years; (ii) acute HF as the main cause for admission; and (iii)
predictive capacity of the biomarkers. Although the predictive hospitalization ≥ 24 h in the Cardiology Department. Exclusion
capacity of the single-point measurement can be improved criteria were: (i) HF episode secondary to ST-segment elevation
by increasing the number of the biomarkers in the panel, it is acute coronary syndrome; (ii) end-stage disease with a life
theoretically possible that a substantial improvement could be expectancy < 1 year; and (iii) any condition that would prevent
alternatively attained using only few of them, but measured an appropriate follow-up. HF was diagnosed in accordance
at both admission and discharge. The prognostic yielding of with current HF guidelines (McMurray et al., 2012). The study
sequential measurements of a single biomarker (van Vark et al., complies with the Declaration of Helsinki and the protocol was
2017) or a series of them (Demissei et al., 2017a,b) in patients approved by the Ethics Committees of each participating center.
with HF was evaluated in post hoc analysis of clinical trials. All patients gave written informed consent.
Study Variables
Abbreviations: HF, heart failure; NT-proBNP, N-terminal pro-B-type natriuretic Data were collected using specifically designed web forms and
peptide; hs-TnT, high-sensitive T troponin; Cys-C, cystatin-C; GDF-15, growth
differentiation factor 15; GAL-3, galectin-3; hs-CRP, high sensitive C reactive quality controls were done monthly. The following clinical
protein. variables were gathered at study inclusion and before discharge:
Frontiers in Physiology | www.frontiersin.org 2 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 3
Álvarez-García et al. Multimarker in Acute Heart Failure
demographic and previous clinical data, case history and physical was finally composed of variables at admission (number of HF
examination, chest x-rays, ECG, echocardiography, laboratory episodes during the last year, previous stroke, systolic blood
blood tests, and pharmacological and non-pharmacological pressure, presence of right HF signs, significant mitral valve
treatment (Appendix 1). Standard criteria were used to define regurgitation, hyponatremia, and body mass index), and variables
the clinical variables. Left ventricular ejection fraction (LVEF) at discharge (persisting HF signs, heart rate, left bundle branch
was categorized according to the recent HF European guidelines block, eGFR < 60 ml/min/1.73 m2, and length of hospital stay).
(Ponikowski et al., 2016). The discriminative ability of the model for all-cause mortality at
12-month after discharge assessed by the C-statistic index was
Biomarker Panel 0.752 (95% CI 0.715–0.789).
Blood samples were obtained by venipuncture within the On top of this clinical model, we then analyzed sequentially
first 24 h of admission and thereafter at hospital discharge. the added prognostic value of each individual biomarker and
The samples were centrifuged at 2,500 g for 15 min. Serum their combinations using the gain in the C-index, calibration
and plasma aliquots of 0.5 mL were separated and frozen (Grønnesby and Borgan, Brier score, Akaike and Bayesian
at –80◦C until analysis; all samples of the same individual criteria), integrated discrimination improvement (IDI), and net
were processed in the same batch. Biomarker concentrations reclassification improvement (NRI). Moreover, we also analyzed
were measured at a core laboratory (Biochemistry Department, the added prognostic value of each individual biomarker and
Hospital de la Santa Creu i Sant Pau, Barcelona, Spain). We their combinations on top of other well-validated clinical scores
measured the serum levels of NT-proBNP, hs-cTnT, and GDF- usually used in clinical practice as the MAGGIC (Pocock
15 by electrochemiluminescence immunoassays, and cystatin-C et al., 2013), GWTG-HF (Peterson et al., 2010), Redin-SCORE
and hs-CRP by particle-enhanced turbidimetric immunoassays (Álvarez-García et al., 2015), and BCN bio-HF (Lupón et al.,
using reagents from Roche Diagnostics (Basel, Switzerland). 2014). The ROC curve analysis was used to determine the
Galectin-3 was also measured in serum using an enzyme-linked optimal biomarker cut-off value to predict the primary endpoint
fluorescent immunoassay (BioMérieux, Marcy-l’Étoile, France). employing the Youden criteria.
The imprecision of all assays was similar or even lower than that
reported by manufacturers.
RESULTS
Follow-Up and Outcomes
In addition to the specific clinical follow-up needed by the Clinical Characteristics of the Study
patient, the vital status was also checked at 12 months after Population
discharge. We used either telephone interviews or clinical records As shown in Table 1, most patients were elderly, male, Caucasian,
of hospitals, primary care, or institutional death registries. had a previous history of HF (60%), and a high Charlson
The primary outcome was all-cause mortality at 12-month comorbidity index. According to LVEF at admission, 263 patients
after discharge. The secondary outcomes were cardiovascular (32%) were classified as HFrEF, 207 (25%) as HFmrEF and 360
mortality, HF mortality, and readmission for HF at the (43%) as HFpEF.
same time period. The reported events were reviewed by an
ad hoc committee. Biomarker Changes During
Statistical Analysis Compensation of Acute Heart Failure
Continuous variables are expressed as mean (standard deviation) Episode
or as median (interquartile range) whenever appropriate. As summarized in Table 2, the biomarkers linked to myocyte
Differences in continuous variables were tested by the analysis of stress (NT-proBNP), inflammation (hs-CRP), and matrix
variance (ANOVA), Student’s t-test, or Wilcoxon signed rank test remodeling (GDF-15) decreased significantly after the hospital
for independent samples. Categorical variables were presented stay whereas the percentage of change of GAL-3 and hs-TnT was
as frequency and percentage. Differences in the categorical negligible. The increase of biomarker reflecting renal damage
variables were assessed by the χ2 test or by Fisher’s exact (Cys-C) was lower than the expected by the biological variability.
test. A two-sided p-value < 0.05 was considered statistically The Supplementary Table 1 summarizes the best cutoff points
significant. Missing data were imputed using the “MICE” package of each biomarker predicting the primary outcome according to
in R (Multivariate Imputation by Chained Equations) whenever the ROC analysis.
necessary (m = 1). All the analyses were performed using R (v.
3.2) and STATA (v. 13.1). Added Prognostic Value of a
Firstly, we developed the best clinical model to predict the Multi-Biomarker 2-Point Based Strategy
occurrence of the primary endpoint using a multivariable Cox On top of the best clinical model, the addition of elevated
regression analysis. Clinical meaningful variables and those NT-proBNP, hs-CRP, and GDF-15 at admission (>6,319
showing a p-value< 0.1 in the univariate analysis were thereafter ng/L, > 15.8 mg/L, and > 5,452 ng/L, respectively), at discharge
included in the multivariate model. A backward stepwise (>3,239 ng/L, > 12.5 mg/L, and > 4,291 ng/L, respectively),
method was used to identify independent predictors with a and the inclusion of the magnitude of change during the
p-value < 0.05 as inclusion or deletion criteria. This model compensation (–23.3, –21.7, and –15.6%, respectively) gave
Frontiers in Physiology | www.frontiersin.org 3 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 4
Álvarez-García et al. Multimarker in Acute Heart Failure
TABLE 1 | Baseline characteristics of the study population. the biomarkers at discharge. Moreover, the 3-biomarkers model
provided a huge reclassification of patients with and without
Total (N = 830 patients)
increased risk reaching a statistically significant NRI of 56% for
Age, years, median (IQR) 75 (65–82) 12-month mortality. These scores were achieved with a correct
Male, n (%) 471 (57) calibration of the models (Supplementary Figure 1). Similarly,
Caucasian, n (%) 815 (98) the addition of these 3 biomarkers on top of the MAGGIC,
Body mass index, kg/m2, median (IQR) 29 (25–33) GWTG-HF, Redin-SCORE, and BCN bio-HF models was the
Chronic heart failure, n (%) 501 (60) best strategy in terms of gain of C-index and reclassification
Ischemic etiology, n (%) 271 (33) parameters. Table 3 summarizes the discrimination, calibration,
I-II NYHA class 24 h before admission 634 (74) IDI, and NRI parameters for the primary outcome given by
LVEF, %, mean (SD) 46 (18) the clinical models alone, in combination with the 6-biomarker
HFrEF, n (%) 263 (32) model, and the 3-biomarker strategy. The discrimination capacity
HFmEF, n (%) 207 (25) of the models for HF-mortality was also better than that for
HFpEF, n (%) 360 (43) cardiovascular and overall mortality (Supplementary Figure 2).
Previous HF admissions within 1 year, mean (SD) 0.9 (1.5) The 3-biomarker strategy also allowed to identify 4-risk
Newly diagnosed HF, n (%) 329 (40) categories for 12-month all-cause mortality: (1) low-risk group
Hypertension, n (%) 635 (77) (529 patients) presenting either none or 1 elevated biomarker,
Diabetes mellitus, n (%) 387 (47) (2) low-intermediate risk group (78 patients) presenting 2 or 3
Atrial fibrillation, n (%) 354 (43) elevated biomarkers at admission but none or 1 at discharge,
Chronic kidney disease (eGFR < 60 ml/min/1.73 m2), n (%) 249 (30) (3) high-intermediate group (86 patients) presenting either none
Stroke, n (%) 83 (10) or 1 elevated biomarker at admission but 2 or 3 at discharge,
COPD, n (%) 131 (16) and (4) high-risk group (137 patients) presenting 2 or 3 elevated
Charlson comorbidity index, mean (SD) 3.5 (2.7) biomarkers both at admission and discharge. As shown in the
Clinical data at admission Figure 1, the 12-month mortality rates for these four categories
Clinical profile of acute HF, n (%) was, respectively, 11.7, 19.2, 26.7, and 39.4% (p < 0.001 for
Acutely decompensated chronic HF 595 (72) the trend). Considering the low risk category as reference, the
Pulmonary edema 115 (14) mortality risk-ratio was 1.64 (95% CI: 0.98–2.74) for the low-
Right HF 34 (4) intermediate; 2.28 (95% CI: 1.50–3.48) for the high-intermediate;
Others 86 (10) and 3.36 (95% CI: 2.46–4.60) for the highest risk category.
Heart rate, bpm, median (IQR) 85 (72–100) Supplementary Table 2 summarizes the predictive capacity gain
Systolic blood pressure, mmHg, median (IQR) 130 (114–150) of all combinations of the six biomarkers, when added to the
Intravenous therapies, n (%) best clinical model.
Diuretics 803 (97)
Vasodilators, n (%) 117 (14)
Inotropes 47 (6) DISCUSSION
Clinical data at discharge
Heart rate, bpm, median (IQR) 70 (62–80) Main Findings
Systolic blood pressure, mmHg, median (IQR) 117 (105–130) Our study revealed that elevated concentrations of NT-proBNP,
Decrease >3 kg of body weight 199 (24) hs-CRP, and GDF-15 at hospital admission and discharge in
Length of stay, days, median (IQR) 9 (6–13) patients with acute HF predicted 12-month mortality better
ACEI/ARB, n (%) 572 (69) than the best clinical model developed in our population and
Beta-blockers, n (%) 586 (71) permitted to define 4 levels of risk. Moreover, the predictive
MRA, n (%) 370 (46) capacity of this 3-biomarker panel was not increased by adding
Outcomes hs-TnT, cystatin C and Galectin-3 in the model and was also
12-month mortality, n (%) 154 (18.6) superior to the most the currently used scores.
IQR, interquartile range; kg, kilogram; m, meter; NYHA, New York Heart Association;
LVEF, left ventricular ejection fraction; SD, standard deviation; HFrEF, heart failure Predictive Risk Capacity of Biomarker
with reduced ejection fraction; HFmEF, heart failure with mid ejection fraction;
HFpEF, heart failure with preserved ejection fraction; HF, heart failure; eGFR, Strategies
estimated glomerular filtration rate; ml, milliliter; min, minute; COPD, chronic Heart failure encompasses several pathophysiological processes
obstructive pulmonary disease; bpm, beats per minute; mm, millimeter; ACEI, that can be indirectly estimated by analyzing the biomarkers
angiotensin converter enzyme inhibitor; ARB, angiotensin receptor blocker; MRA, specifically related to the underlying mechanisms (Braunwald,
mineraloid receptor antagonist.
2008). Thus, measurement of a set of biomarkers would afford
an integrative knowledge of the complex pathophysiology of
rise to the highest improvement in the C-index for 12-month HF and, ultimately, would permit a better risk assessment
mortality (0.782; 95% CI 0.747–0.817, p < 0.001). Of notice, the and identification of patients requiring a close follow-up plan.
discrimination of this 3-biomarker model was better than that During the last 15 years, at least 20 clinical studies including
including the six biomarkers, and even better than that based only 3 or more biomarkers have been published (Ishii et al., 2002;
Frontiers in Physiology | www.frontiersin.org 4 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 5
Álvarez-García et al. Multimarker in Acute Heart Failure
TABLE 2 | Time course of biomarkers during the clinical compensation of acute HF.
Admission Discharge P-value* Delta**
NT-proBNP, ng/L 3710 (1784/7634) 1814 (874/4220) <0.001 −43.6 (−67.1/−6.7)
Hs-TnT, ng/L 35.2 (20.0/61.9) 34.1 (20.0/60.4) 0.348 −0.9 (−23.5/24.7)
Cystatin C, mg/L 1.5 (1.2/2.0) 1.6 (1.2/2.1) <0.001 4.1 (−5.5/17.1)
Hs-CRP, mg/L 10.2 (4.5/29.5) 7.4 (3.1/18.8) <0.001 −34.7 (−66.7/19.5)
GDF-15, ng/L 3366 (2176/5643) 2882 (1963/4989) <0.001 −11.2 (−30.1/13.0)
GAL-3, mg/L 22.7 (17.1/30.8) 22.1 (16.4/30.8) 0.043 −2.0 (−14.4/14.0)
Median (1st Quartile/3rd Quartile).
*Wilcoxon signed rank test with continuity correction.
**Delta: [(discharge value-admission value)/admission value]*100.
HF, heart failure; NT-proBNP, N-terminal pro-B-type natriuretic peptide; ng, nanograms; L, liter; hsTnT, high sensitivity troponin T; mg, milligrams; hs-CRP, high-sensitivity
C-reactive protein; GDF15, Growth/differentiation factor 15; ml, milliliter; GAL-3, galectin-3.
TABLE 3 | Added prognostic value of a multi-biomarker 2-point-based risk stratification strategy in acute heart failure to predict 12-month all-cause mortality.
C-index P-value vs. G-B Brier AIC BIC IDI P-value NRI P-value
clinical model p-value score for IDI for NRI
Clinical model (CM) 0.752 0.742 0.138 1908 1911
(0.715–0.789)
CM + All biomarker 0.768 <0.001 0.900 0.134 1887 1899 0.031 <0.001 0.434 <0.001
(0.730–0.805)
CM + NT-proBNP, 0.782 <0.001 0.550 0.133 1889 1919 0.050 <0.001 0.566 <0.001
hs-CRP, GDF-15 (0.747–0.817)
MAGGIC 0.639 0.836 0.148 1999 2002
(0.594–0.684)
MAGGIC + All biomarker 0.723 <0.001 0.995 0.140 1941 1953 0.081 <0.001 0.569 <0.001
(0.684–0.762)
MAGGIC + NT-proBNP, 0.745 <0.001 0.563 0.137 1934 1964 0.108 <0.001 0.661 <0.001
hs-CRP, GDF-15 (0.709–0.782)
GWTG 0.646 0.999 0.148 1998 2001
(0.602–0.691)
GWTG + All biomarker 0.722 <0.001 0.998 0.140 1943 1955 0.078 <0.001 0.575 <0.001
(0.682–0.762)
GWTG + NT-proBNP, 0.746 <0.001 0.053 0.137 1934 1965 0.107 <0.001 0.601 <0.001
hs-CRP, GDF-15 (0.709–0.783)
Redin-SCORE 0.636 0.224 0.150 2011 2014
(0.585–0.686)
Redin-SCORE + All 0.720 <0.001 0.561 0.140 1944 1956 0.094 <0.001 0.644 <0.001
biomarker (0.680–0.760)
Redin-SCORE + NT- 0.743 <0.001 0.686 0.137 1936 1967 0.123 <0.001 0.638 <0.001
proBNP, hs-CRP, (0.706–0.780)
GDF-15
BCN-BIO HF 0.617 0.719 0.151 2017 2020
(0.573–0.661)
BCN-BIO HF + All 0.719 <0.001 0.882 0.140 1945 1957 0.099 <0.001 0.693 <0.001
biomarker (0.679–0.759)
BCN-BIO 0.743 <0.001 0.710 0.137 1934 1964 0.132 <0.001 0.694 <0.001
HF + NT-proBNP, (0.706–0.780)
hs-CRP, GDF-15
G-B, Grønnesby and Borgan; AIC, Akaike criteria; BIC, Bayesian criteria; IDI, integrated discrimination improvement; NRI, net reclassification improvement; NT-proBNP,
N-terminal pro-B-type natriuretic peptide; hs-CRP, high-sensitivity C-reactive protein; GDF15, growth/differentiation factor 15; CV, cardiovascular; HF, heart failure.
van Kimmenade et al., 2006; Januzzi et al., 2007; Rehman 2017). As summarized in Supplementary Table 3, half of these
et al., 2008; Manzano-Fernández et al., 2009, 2011; Niizeki reports corresponded to clinical trials (van Kimmenade et al.,
et al., 2009; Zairis et al., 2010; Pascual-Figal et al., 2011; Shah 2006; Januzzi et al., 2007; Rehman et al., 2008; Manzano-
et al., 2012; Bjurman et al., 2013; Lassus et al., 2013; Lok Fernández et al., 2011; Shah et al., 2012; Lok et al., 2013;
et al., 2013; Richter et al., 2013; Srinivas et al., 2014; Demissei Demissei et al., 2016, 2017a,b; Tromp et al., 2017), that recruited
et al., 2016, 2017a,b; Herrero-Puente et al., 2017; Tromp et al., selected groups of patients, and any case the sample size was
Frontiers in Physiology | www.frontiersin.org 5 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 6
Álvarez-García et al. Multimarker in Acute Heart Failure
FIGURE 1 | Risk categories based on the values of NT-proBNP, hs-CRP, and GDF-15 in the study population. Upper panel: our study identified 4-risk categories for
12-month all-cause mortality based on the values of NT-proBNP, hs-CRP, and GDF-15: (1) a low-risk category (blue line) included 529 patients presenting none or 1
biomarker above the cutoff values at admission and discharge, (2) a low-intermediate risk category (green line) included 78 patients presenting 2 or 3 elevated
biomarkers at admission and none or 1 at discharge, (3) a high-intermediate category (orange line) included 86 patients presenting none or 1 elevated biomarker at
admission and 2 or 3 elevated biomarkers at discharge, and (4) a high-risk category (red line) included 137 patients presenting 2 or 3 elevated biomarkers both at
admission and discharge. Bottom panel: The 12-month mortality rate for these 4 categories was, respectively, 11.7, 19.2, 26.7, and 39.4%. Considering the low risk
category as reference, the mortality risk-ratio was 1.64 (95% CI: 0.98–2.74) for the low-intermediate; 2.28 (95% CI: 1.50–3.48) for the high-intermediate; and 3.36
(95% CI: 2.46–4.60) for the highest risk categories.
greater than that of our study (Ishii et al., 2002; Manzano- the linked underlying mechanisms namely myocyte stretch,
Fernández et al., 2009; Niizeki et al., 2009; Zairis et al., 2010; inflammation, and myocardial remodeling had not improved
Pascual-Figal et al., 2011; Bjurman et al., 2013; Lassus et al., upon clinical compensation of the HF episode are at high risk
2013; Richter et al., 2013; Srinivas et al., 2014; Herrero-Puente of mortality. The percentage of change of the other 3 studied
et al., 2017). Of notice, in 17 of 20 studies the biomarker biomarkers hs-cTnT, GAL-3, and Cys-C at discharge was less
was measured either at hospital admission or discharge, and than 5% and these biomarkers did not improve the discriminative
only in 3 cases the biomarkers were measured at both clinical risk capacity beyond that achieved by NT-proBNP, hs-CRP,
circumstances (Demissei et al., 2016, 2017a,b). The external and GDF-15. The lack of risk prediction of hs-cTnT in our
validity of the data reported in these studies might be hampered study could deal with several causes. Elevated hs-TnT values
by limitations inherent to the post hoc analysis in clinical trials, in acute HF could be associated to ischemia, inflammation,
and also to the single-center design in 6 studies (Ishii et al., oxidative stress or impaired renal function. However, all these
2002; Manzano-Fernández et al., 2009; Niizeki et al., 2009; alterations could also increase the 3 biomarkers, particularly
Pascual-Figal et al., 2011; Bjurman et al., 2013; Srinivas et al., GDF-15 and hs-CRP, already included in the model and, the
2014). prognostic role of hs-cTnT could be already covered (Kociol
Our study overcomes some of these limitations and emerges as et al., 2010). In addition, we excluded in our study ST-
the first observational, multicenter registry analyzing the capacity segment elevation acute coronary syndrome as a cause of HF
of a set of biomarkers double measured at hospital admission and hospitalization and hs-TnT is known to be a strong risk predictor
discharge to predict relevant 1-year outcomes in a large group in these patients.
of patients with acute HF. We selected six biomarkers linked
to the main processes involved in HF such as neurohormonal Clinical Implications
activation, myocyte stretch, injury or inflammation, myocardial A decrease in the plasma level of natriuretic peptides during the
remodeling and fibrosis, and impaired renal function involved clinical compensation of a HF episode is associated with lower
in HF development. Among these 6 biomarkers, we found that cardiovascular mortality and lower readmissions at 6 months
the model including the NT-proBNP, hs-CRP, and GDF-15 was (Savarese et al., 2014). However, a systematic recommendation
the best to predict 12-month mortality. Interestingly, these 3 on their use in clinical practice is not reflected in the current
biomarkers presented the largest magnitude of change from guidelines (Yancy et al., 2013; Ponikowski et al., 2016). Recently,
hospital admission to discharge suggesting that patients in whom a consensus document of the American Heart Association stated
Frontiers in Physiology | www.frontiersin.org 6 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 7
Álvarez-García et al. Multimarker in Acute Heart Failure
that the measurement of natriuretic peptides, cardiac troponin, DATA AVAILABILITY STATEMENT
and biomarkers of fibrosis at the time of presentation is useful
and reasonable for establishing prognosis in patients with acutely The raw data supporting the conclusions of this article will be
decompensated HF (Chow et al., 2017). Our study contributes made available by the authors, without undue reservation.
on this important issue by identifying the best combination of
3 out of 6 currently used biomarkers that are the most useful
to predict 1-year mortality of patients after hospitalization for ETHICS STATEMENT
heart failure. Specially, the relevant IDI and NRI values by our 3-
biomarker model reinforce its role improving the ever-complex The protocol was approved by the Ethics Committees of
HF stratification process. each participating center. The patients/participants provided
their written informed consent to participate in this
study.
Study Limitations
This study includes 98% of Caucasian patients, thus our
data might not be fully applicable to other ethnicities or AUTHOR CONTRIBUTIONS
countries. Considering that our study design necessarily required
biomarker measurements available at hospital admission and JÁ-G, JO-L, and JC contributed to conception and design
discharge, we did not include patients lacking the discharge of the study. JÁ-G and AF-G organized the database. AF-G
sample. Moreover, the size of the study sample did not allow performed the statistical analysis. JÁ-G wrote the first draft of
analyzing the performance of the multi-biomarker strategies the manuscript. JÁ-G, ÁG-O, MV-B, AF-G, JO-L, and JC wrote
in subgroups of clinical interest. Therefore, external validation sections of the manuscript. All authors contributed to manuscript
of the clinical model and the full model including biomarkers revision, read, and approved the submitted version.
should be performed.
FUNDING
CONCLUSION This work was supported by grants from Redes Temáticas de
Investigación Cooperativa en Salud del Instituto de Salud Carlos
In a multicenter, prospective registry of patients with acute HF, we III (REDINSCOR), Madrid, Spain (grant no. RD06-0003-0000)
identified 3 out of 6 currently available biomarkers that afforded and Red de Investigación Cardiovascular del Instituto de Salud
the highest discriminative power to predict 12-month mortality Carlos III (RIC), Madrid, Spain (grant no. RD12/0042/0002).
beyond the best clinical model and also above the currently used
MAGGIC, GWTG-HF, Redin-SCORE, and BCN bio-HF scores.
Moreover, this simple 3-biomarker panel permitted to define 4 ACKNOWLEDGMENTS
predictive risk levels for 1-year mortality.
What is Known About the Topic? We are indebted to Roche Diagnostics Intl. and BioMerieux for
freely providing the reagents used for biomarker measurement.
• The prognostic stratification of patients with acute HF is The companies had no further involvement in the different
essential to establish an appropriate personalized follow- steps of the study.
up plan.
• Cardiac biomarkers have improved the predictive models
of HF patients beyond the already well-established clinical SUPPLEMENTARY MATERIAL
risk predictors.
• Most multi-panel strategies in acute HF have only measured The Supplementary Material for this article can be found
biomarkers in a 1-point time. online at: https://www.frontiersin.org/articles/10.3389/fphys.
2021.708890/full#supplementary-material
What Does This Study Add?
Supplementary Figure 1 | Calibration plots of the clinical, 6-biomarker, and
We evaluate the prognostic role of 6 biomarkers at 3-biomarker models for each outcome.
•
admission and discharge after HF admission. Supplementary Figure 2 | Comparison of C-index between clinical, 6-biomarker,
and 3-biomarker models for each outcome. Discrimination of the 3-biomarker
• Our study identifies a simple set of 3 biomarkers to predict
prognosis of HF patients. model (red line) was better than that including only clinical variables (blue line), and
even better than that considering all the biomarkers (green line). In addition,
• This panel permits to define 4 predictive risk levels for C-index for HF-mortality was also better than that for cardiovascular and
12-month mortality. overall mortality.
Frontiers in Physiology | www.frontiersin.org 7 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 8
Álvarez-García et al. Multimarker in Acute Heart Failure
REFERENCES T in patients with acute heart failure. Am. J. Cardiol. 103, 1753–1759. doi:
10.1016/j.amjcard.2009.02.029
Álvarez-García, J., Ferrero-Gregori, A., Puig, T., Vázquez, R., Delgado, J., Pascual- Manzano-Fernández, S., Mueller, T., Pascual-Figal, D., Truong, Q. A., and Januzzi,
Figal, D., et al. (2015). A simple validated method for predicting the risk J. L. (2011). Usefulness of soluble concentrations of interleukin family member
of hospitalization for worsening of heart failure in ambulatory patients: the ST2 as predictor of mortality in patients with acutely decompensated heart
Redin-SCORE. Eur. J. Heart Fail. 17, 818–827. doi: 10.1002/ejhf.287 failure relative to left ventricular ejection fraction. Am. J. Cardiol. 107, 259–267.
Bjurman, C., Jensen, J., Petzold, M., Hammarsten, O., and Fu, M. L. X. (2013). doi: 10.1016/j.amjcard.2010.09.011
Assessment of a multimarker strategy for prediction of mortality in older heart McMurray, J. J. V., Adamopoulos, S., Anker, S. D., Auricchio, A., Böhm, M.,
failure patients: a cohort study. BMJ Open 3:e002254. doi: 10.1136/bmjopen- Dickstein, K., et al. (2012). ESC Guidelines for the diagnosis and treatment
2012-002254 of acute and chronic heart failure 2012: the task force for the diagnosis and
Braunwald, E. (2008). Biomarkers in heart failure. N. Engl. J. Med. 358, 2148–2159. treatment of acute and chronic heart failure 2012 of the european society
doi: 10.1056/NEJMra0800239 of cardiology. developed in collaboration with the heart. Eur. Heart J. 33,
Chow, S. L., Maisel, A. S., Anand, I., Bozkurt, B., de Boer, R. A., Felker, G. M., et al. 1787–1847. doi: 10.1093/eurheartj/ehs104
(2017). Role of biomarkers for the prevention, assessment, and management Niizeki, T., Takeishi, Y., Kitahara, T., Suzuki, S., Sasaki, T., Ishino, M., et al. (2009).
of heart failure: a scientific statement from the american heart association. Combination of conventional biomarkers for risk stratification in chronic heart
Circulation 135, 1054–1091. doi: 10.1161/CIR.0000000000000490 failure. J. Cardiol. 53, 179–187. doi: 10.1016/j.jjcc.2008.10.003
Demissei, B. G., Cotter, G., Prescott, M. F., Felker, G. M., Filippatos, Pascual-Figal, D. A., Manzano-Fernández, S., Boronat, M., Casas, T., Garrido,
G., Greenberg, B. H., et al. (2017a). A multimarker multi-time point- I. P., Bonaque, J. C., et al. (2011). Soluble ST2, high-sensitivity troponin T-
based risk stratification strategy in acute heart failure: results from and N-terminal pro-B-type natriuretic peptide: complementary role for risk
the RELAX-AHF trial. Eur. J. Heart Fail. 19, 1001–1010. doi: 10.1002/ stratification in acutely decompensated heart failure. Eur. J. Heart Fail. 13,
ejhf.749 718–725. doi: 10.1093/eurjhf/hfr047
Demissei, B. G., Postmus, D., Cleland, J. G., O’Connor, C. M., Metra, M., Peterson, P. N., Rumsfeld, J. S., Liang, L., Albert, N. M., Hernandez, A. F.,
Ponikowski, P., et al. (2017b). Plasma biomarkers to predict or rule out early Peterson, E. D., et al. (2010). A validated risk score for in-hospital mortality
post-discharge events after hospitalization for acute heart failure. Eur. J. Heart in patients with heart failure from the american heart association get with the
Fail. 19, 728–738. doi: 10.1002/ejhf.766 guidelines program. Circ. Cardiovasc. Qual. Outcomes 3, 25–32. doi: 10.1161/
Demissei, B. G., Valente, M. A. E., Cleland, J. G., O’Connor, C. M., Metra, M., CIRCOUTCOMES.109.854877
Ponikowski, P., et al. (2016). Optimizing clinical use of biomarkers in high-risk Pocock, S. J., Ariti, C. A., McMurray, J. J. V., Maggioni, A., Køber, L., Squire, I. B.,
acute heart failure patients. Eur. J. Heart Fail. 18, 269–280. doi: 10.1002/ejhf.443 et al. (2013). Predicting survival in heart failure: a risk score based on 39 372
Herrero-Puente, P., Prieto-García, B., García-García, M., Jacob, J., Martín-Sánchez, patients from 30 studies. Eur. Heart J. 34, 1404–1413. doi: 10.1093/eurheartj/
F. J., Pascual-Figal, D., et al. (2017). Predictive capacity of a multimarker ehs337
strategy to determine short-term mortality in patients attending a hospital Ponikowski, P., Voors, A. A., Anker, S. D., Bueno, H., Cleland, J. G. F., Coats,
emergency Department for acute heart failure. BIO-EAHFE study. Clin. Chim. A. J. S., et al. (2016). 2016 ESC Guidelines for the diagnosis and treatment of
Acta 466, 22–30. doi: 10.1016/j.cca.2017.01.003 acute and chronic heart failure: the task force for the diagnosis and treatment
Ishii, J., Nomura, M., Nakamura, Y., Naruse, H., Mori, Y., Ishikawa, T., et al. of acute and chronic heart failure of the European society of cardiology
(2002). Risk stratification using a combination of cardiac troponin T and brain (ESC)Developed with the special contribution of the heart failure association
natriuretic peptide in patients hospitalized for worsening chronic heart failure. (HFA) of the ESC. Eur. Heart J. 37, 2129–2200. doi: 10.1093/eurheartj/ehw128
Am. J. Cardiol. 89, 691–695. Rehman, S. U., Mueller, T., and Januzzi, J. L. (2008). Characteristics of the novel
Januzzi, J. L., Peacock, W. F., Maisel, A. S., Chae, C. U., Jesse, R. L., Baggish, interleukin family biomarker ST2 in patients with acute heart failure. J. Am.
A. L., et al. (2007). Measurement of the interleukin family member ST2 in Coll. Cardiol. 52, 1458–1465. doi: 10.1016/j.jacc.2008.07.042
patients with acute dyspnea: results from the PRIDE (Pro-brain natriuretic Richter, B., Koller, L., Hohensinner, P. J., Zorn, G., Brekalo, M., Berger, R.,
peptide investigation of dyspnea in the emergency department) study. J. Am. et al. (2013). A multi-biomarker risk score improves prediction of long-term
Coll. Cardiol. 50, 607–613. doi: 10.1016/j.jacc.2007.05.014 mortality in patients with advanced heart failure. Int. J. Cardiol. 168, 1251–1257.
Kociol, R. D., Pang, P. S., Gheorghiade, M., Fonarow, G. C., O’Connor, C. M., doi: 10.1016/j.ijcard.2012.11.052
and Felker, G. M. (2010). Troponin elevation in heart failure prevalence, Savarese, G., Musella, F., D’Amore, C., Vassallo, E., Losco, T., Gambardella, F.,
mechanisms, and clinical implications. J. Am. Coll. Cardiol. 56, 1071–1078. et al. (2014). Changes of natriuretic peptides predict hospital admissions in
doi: 10.1016/j.jacc.2010.06.016 patients with chronic heart failure: a meta-analysis. JACCHeart Fail. 2, 148–158.
Lassus, J., Gayat, E., Mueller, C., Peacock, W. F., Spinar, J., Harjola, V. P., doi: 10.1016/j.jchf.2013.11.007
et al. (2013). Incremental value of biomarkers to clinical variables for Shah, R. V., Truong, Q. A., Gaggin, H. K., Pfannkuche, J., Hartmann, O., and
mortality prediction in acutely decompensated heart failure: the multinational Januzzi, J. L. (2012). Mid-regional pro-atrial natriuretic peptide and pro-
observational cohort on acute heart failure (MOCA) study. Int. J. Cardiol. 168, adrenomedullin testing for the diagnostic and prognostic evaluation of patients
2186–2194. doi: 10.1016/j.ijcard.2013.01.228 with acute dyspnoea. Eur. Heart J. 33, 2197–2205. doi: 10.1093/eurheartj/
Levy, W. C., Mozaffarian, D., Linker, D. T., Sutradhar, S. C., Anker, S. D., Cropp, ehs136
A. B., et al. (2006). The seattle heart failure model: prediction of survival in Srinivas, P., Manjunath, C. N., Banu, S., and Ravindranath, K. S. (2014). Prognostic
heart failure. Circulation 113, 1424–1433. doi: 10.1161/CIRCULATIONAHA. significance of a multimarker strategy of biomarkers in acute heart failure.
105.584102 J. Clin. Diagn. Res. 8, MC01–MC06. doi: 10.7860/JCDR/2014/9289.4783
Lok, D. J., Klip, I. T., Lok, S. I., Bruggink-André de la Porte, P. W., Badings, Tromp, J., Khan, M. A. F., Klip, I. T., Meyer, S., de Boer, R. A., Jaarsma, T., et al.
E., van Wijngaarden, J., et al. (2013). Incremental prognostic power of (2017). Biomarker profiles in heart failure patients with preserved and reduced
novel biomarkers (growth-differentiation factor-15, high-sensitivity C-reactive ejection fraction. J. Am. Heart Assoc. 6:e003989. doi: 10.1161/JAHA.116.003989
protein, galectin-3, and high-sensitivity troponin-T) in patients with advanced van Kimmenade, R. R., Januzzi, J. L., Ellinor, P. T., Sharma, U. C., Bakker, J. A.,
chronic heart failure. Am. J. Cardiol. 112, 831–837. doi: 10.1016/j.amjcard.2013. Low, A. F., et al. (2006). Utility of amino-terminal pro-brain natriuretic peptide.
05.013 J. Am. Coll. Cardiol. 48, 1217–1224. doi: 10.1016/J.JACC.2006.03.061
Lupón, J., de Antonio, M., Vila, J., Peñafiel, J., Galán, A., Zamora, E., et al. (2014). van Vark, L. C., Lesman-Leegte, I., Baart, S. J., Postmus, D., Pinto, Y. M., Orsel,
Development of a novel heart failure risk tool: the barcelona bio-heart failure J. G., et al. (2017). Prognostic value of serial ST2 measurements in patients with
risk calculator (BCN bio-HF calculator). PLoS One 9:e85466. doi: 10.1371/ acute heart failure. J. Am. Coll. Cardiol. 70, 2378–2388. doi: 10.1016/j.jacc.2017.
journal.pone.0085466 09.026
Manzano-Fernández, S., Boronat-Garcia, M., Albaladejo-Otón, M. D., Pastor, P., Yancy, C. W., Jessup, M., Bozkurt, B., Butler, J., Casey, D. E. Jr., Drazner, M. H.,
Garrido, I. P., and Pastor-Pérez, F. J. (2009). Complementary prognostic value et al. (2013). 2013 ACCF/AHA guideline for the management of heart failure.
of cystatin C, N-terminal pro-B-type natriuretic Peptide and cardiac troponin Circulation 128, e240–e327. doi: 10.1161/CIR.0b013e31829e8776
Frontiers in Physiology | www.frontiersin.org 8 October 2021 | Volume 12 | Article 708890
fphys-12-708890 October 18, 2021 Time: 16:16 # 9
Álvarez-García et al. Multimarker in Acute Heart Failure
Zairis, M. N., Tsiaousis, G. Z., Georgilas, A. T., Makrygiannis, S. S., Adamopoulou, this article, or claim that may be made by its manufacturer, is not guaranteed or
E. N., Handanis, S. M., et al. (2010). Multimarker strategy for the prediction endorsed by the publisher.
of 31 days cardiac death in patients with acutely decompensated chronic heart
failure. Int. J. Cardiol. 141, 284–290. doi: 10.1016/j.ijcard.2008.12.017 Copyright © 2021 Álvarez-García, García-Osuna, Vives-Borrás, Ferrero-Gregori,
Martínez-Sellés, Vázquez, González-Juanatey, Rivera, Segovia, Pascual-Figal, Bover,
Conflict of Interest: The authors declare that the research was conducted in the Bascompte, Delgado, Grau Sepúlveda, Bardají, Pérez-Villa, Zamorano, Crespo-Leiro,
absence of any commercial or financial relationships that could be construed as a Sánchez, Ordoñez-Llanos and Cinca. This is an open-access article distributed
potential conflict of interest. under the terms of the Creative Commons Attribution License (CC BY). The use,
distribution or reproduction in other forums is permitted, provided the original
Publisher’s Note: All claims expressed in this article are solely those of the authors author(s) and the copyright owner(s) are credited and that the original publication
and do not necessarily represent those of their affiliated organizations, or those of in this journal is cited, in accordance with accepted academic practice. No use,
the publisher, the editors and the reviewers. Any product that may be evaluated in distribution or reproduction is permitted which does not comply with these terms.
Frontiers in Physiology | www.frontiersin.org 9 October 2021 | Volume 12 | Article 708890