© 2020 The Authors. This article is published by AboutScience and licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0). Any commercial use is not permitted and is subject to Publisher’s permissions. Full information is available at www.aboutscience.eu JCB ISSN 1849-4544 DOI: 10.33393/jcb.2020.2164 J Circ Biomark 2020; 9(1): 8-12 ORIGINAL RESEARCH ARTICLE Complementary tools such as lung ultrasound (LUS) and car- diac biomarkers might aid in the diagnostic approach. LUS is highly sensitive for pulmonary congestion asses- sment in HF (1), since the number and distribution of B-lines denote the amount of extravascular fluid in the lung. Current guidelines included natriuretic peptides to mini- mize HF diagnosis complexity, especially in the nonacute setting when echocardiography is not immediately available. In recent years, several cardiac biomarkers have been descri- bed, reflecting different active pathogenic pathways in HF (2). The association of B-lines and N-terminal pro-brain natri- uretic peptide (NT-proBNP) has been characterized in decom- pensated acute HF patients. Nevertheless, there are few data on outpatients, and no data with other cardiac biomarkers. Accordingly, we investigated the correlation between B-lines and different biomarkers in outpatients with suspicion of HF Lung ultrasound and biomarkers in primary care: Partners for a better management of patients with heart failure? Mar Domingo1-3, Laura Conangla1-3, Josep Lupón1,4,5, Asunción Wilke2,4, Gladys Juncà1, Elena Revuelta-López1,5,6, Xavier Tejedor7, Antoni Bayes-Genis1,4,5 1Heart Failure Unit and Cardiology Department, Hospital Universitari Germans Trias i Pujol, Badalona - Spain 2Primary Care Service Barcelonès Nord i Maresme, Catalan Health Institute, Badalona - Spain 3Institut Universitari d’Investigació en Atenció Primària Jordi Gol (IDIAP Jordi Gol), Barcelona - Spain 4Department of Medicine, Universitat Autònoma de Barcelona, Barcelona - Spain 5CIBERCV, Instituto de Salud Carlos III, Madrid - Spain 6ICREC Research Program, Germans Trias i Pujol Health Science Research Institute, Can Ruti Campus, Badalona - Spain 7Biochemistry Service, Hospital Universitari Germans Trias i Pujol, Badalona - Spain ABSTRACT Introduction: The association of pulmonary congestion assessed by lung ultrasound (LUS) and biomarkers—other than N-terminal pro-brain natriuretic peptide (NT-proBNP)—is uncertain. Methods: We investigated the relationship between total B-line count by LUS and several biomarkers in outpa- tients with suspicion of heart failure (HF). Primary care patients with suspected new-onset nonacute HF were evaluated both with a 12-scan LUS protocol (8 anterolateral areas plus 4 lower posterior thoracic areas) and 11 inflammatory and cardiovascular biomarkers. A cardiologist blinded to LUS and biomarkers except NT-proBNP confirmed HF diagnosis. After log-transformation of biomarkers’ concentrations, unadjusted and adjusted cor- relations were performed. Results: A total of 170 patients were included (age 76 ± 10 years, 67.6% women). HF diagnosis was confirmed in 38 (22.4%) patients. After adjustment by age, sex, body mass index, and renal function, total B-line sum sig- nificantly correlated with NT-proBNP (R = 0.29, p < 0.001), growth/differentiation factor-15 (GDF-15; R = 0.23, p = 0.003), high-sensitive Troponin T (hsTnT; R = 0.36, p < 0.001), soluble interleukin-1 receptor-like 1 (sST2; R = 0.29, p < 0.001), cancer antigen 125 (CA-125; R = 0.17, p = 0.03), high-sensitivity C-reactive protein (hsCRP; R = 0.20, p = 0.009), and interleukin (IL)-6 (R = 0.23, p = 0.003). In contrast, IL-33 (R = −0.01, p = 0.93), IL-1β (R = −0.10, p = 0.20), soluble neprilysin (sNEP; R = 0.09, p = 0.24), tumor necrosis factor-alpha (TNF-α; R = 0.07, p = 0.39), and TNF-α receptor superfamily member 1A (TNFRSF1A; R = 0.14, p = 0.07) did not. Conclusions: Total B-line sum correlated significantly, although moderately, with congestion and several inflam- mation biomarkers. Unexpectedly, the highest correlation found was with hsTnT. Keywords: Biomarkers, Congestion, Diagnosis, Heart failure, Lung ultrasound, Primary care Introduction Heart failure (HF) diagnosis is challenging in ambula- tory patients, since signs and symptoms are mild and can be related with other diseases, and even to natural aging. Received: June 11, 2020 Accepted: September 4, 2020 Published online: October 16, 2020 Corresponding author: Antoni Bayes-Genis Heart Institute, Hospital Universitari Germans Trias i Pujol Department of Medicine, Universitat Autònoma de Barcelona Carretera del Canyet s/n 08916, Badalona - Spain abayesgenis@gmail.com Domingo et al 9 © 2020 The Authors. Published by AboutScience in the primary care setting. We hypothesized that biomarkers with multiple bio-profiling other than NT-proBNP might be associated with pulmonary congestion by LUS. Methods Study design and patients The present study is a biomarker subanalysis of a prospec- tive cohort of ambulatory patients >50 years old, referred by their primary care physician to NT-proBNP test for suspected new-onset non-acute HF (July 2015 to January 2018) (3). We excluded patients with established HF diagnosis, pulmonary fibrosis, or radiological pachypleuritis. The study was perfor- med in accordance with the Declaration of Helsinki; the local ethics committee approved the research protocol and infor- med consent was obtained from all subjects. Procedures All inclusion visits were scheduled in a centralized setting, where the primary care physician investigator (LC, MD, AW) evaluated the patients, focusing on Framingham criteria, and performed LUS. Blood samples were collected for NT-proBNP measurement (XT), and serum aliquots were stored at −80°C prior to assay (ER-L). At a subsequent visit, a cardiologist investigator (GJ) assessed all participants and performed a transthoracic Doppler echocardiogram. This physician confir- med HF diagnosis, following the European Society of Cardio- logy guidelines. The cardiologist had access to the patients’ electronic records, including the primary care investigator visit and NT-proBNP, but was blinded to LUS and other bio- markers. Assays Biomarker panel NT-proBNP, high-sensitive Troponin T (hsTnT) and growth/ differentiation factor-15 (GDF-15) were measured by Cobas Elecsys® kits (Roche Diagnostics). Cancer antigen 125 (CA- 125) was tested by ARCHITECT CA 125 II assay (Abbott Dia- gnostic). High-sensitivity C-reactive protein (hsCRP) was measured by hsCRP reagent (Beckman Coulter). Human solu- ble neprilysin (NEP) and soluble interleukin-1 receptor-like 1 (sST2) were measured by Human Soluble neprilysin/CD10 ELISA kit (Aviscera Bioscience) and Presage® ST2 (Critical Dia- gnostics) assays, respectively. Interleukin (IL)-1β, IL-33, IL-6, tumor necrosis factor-alpha (TNF-α), and TNF-α receptor superfamily member 1A (TNFRSF1A) were tested by Quanti- kine® immunoassay kits (R&D Systems). NT-proBNP was analyzed after collection. The rest of the biomarkers were analyzed in the first or second freeze-thaw cycle. Lung ultrasound LUS was performed with a pocket device (V-scan simple model with a sectorial phased array transducer; General Electric®) and interpreted bench side. LUS was performed with patient in a seated position; 8 anterolateral thoracic areas plus 4 posterior lower areas were examined. Each of the 12 areas was classified according to the number of B-lines in the sagittal scan. A thoracic area was considered positive if ≥3 B-lines were observed. Pleural effusion was considered as 10 B-lines. LUS congestion was defined as 2 out of 6 positive scans in each hemithorax. Transthoracic Doppler echocardiography Echocardiographic study was performed using an iE33 ultrasound system (Philips Medical Systems; Andover, Massa- chusetts) with a S5-1 sector transducer (5.1 MHz bandwidth), and analyses were performed with an EchoPAC. Statistical analysis Categorical values are described as absolute numbers (percentages) and continuous variables as means (standard deviations) or medians [interquartile ranges], depending on whether data distribution was normal as assessed by nor- mal Q-Q plots. To assess the relationship of total B-line sum acquired by LUS with biomarkers’ concentrations, Pearson correlation was used after logarithmic transformation of biomarker levels; afterward, partial correlations adjusted by age and sex, and finally by age, sex, body mass index, and estimated glomerular filtration rate (eGFR) were performed. Analyses were performed using Statistical Package for the Social Sciences (SPSS) 24. A two-sided p < 0.05 was conside- red significant. Results Table I shows baseline characteristic and biomarker values of the 170 patients included. They were elderly, pre- dominantly women, obese or overweight, and mainly in New York Heart Association (NYHA) class II. HF diagnosis was con- firmed in 38 (22.4%) patients, and only one had left ventricu- lar ejection fraction (LVEF) < 40%. Patients with HF diagnosis had higher levels of all biomarkers except IL-33, IL-1β, and soluble neprilysin (sNEP). They also had a higher number of total B-line count (p < 0.001). Although 85% of patients had exertional dyspnea, only 17.1% had crackles, 9.4% orthop- nea, and 3.5% paroxysmal nocturnal dyspnea. Correlations between total B-line sum and studied bio- markers are shown in Table II. Unadjusted analyses sho- wed that total B-line sum was significantly associated with NT-proBNP, GDF-15, hsTnT, sST2, CA-125, hsCRP, IL-6, and TNFRSF1A (R range 0.18-0.35), while IL-33, IL-1β, sNEP, and TNF-α levels were not associated with total B-line sum. After the adjustments for the four covariates, R values tended to slightly decrease except for hsTnT and TNFRSF1A that lost statistical significance. Discussion Bedside LUS has appeared as a step forward for HF dia- gnosis, and biomarkers other than NT-proBNP are currently Lung ultrasound and biomarkers for heart failure diagnosis in primary care10 © 2020 The Authors. Published by AboutScience TABLE I - Demographic, clinical characteristics and biomarker levels of patients Total n = 170 HF diagnosis n = 38 No HF diagnosis n = 132 p-value Age, years 76 ± 10.4 81.2 ± 8.3 74.4 ± 10 <0.001 Female sex, n (%) 115 (67.6) 23 (60.5) 92 (69.7) 0.29 LVEF, % 63 ± 5.8 59.9 ± 7.2 63.8 ± 5 <0.001 Comorbidities, n (%)  Hypertension  Diabetes mellitus  COPD  Valvular heart disease  Myocardial infarction  Atrial fibrillation  Obesity (BMI >30 kg/m2)  eGFR < 60 mL/min/1.72 m2 132 (77.6) 43 (25.3) 19 (11.2) 6 (3.5) 15 (8.8) 19 (11.2) 84 (49.4) 48 (28.2) 36 (94.7) 12 (31.6) 7 (18.4) 3 (7.9) 7 (18.4) 16 (42.1) 20 (52.6) 18 (47.4) 96 (72.7) 31 (23.5) 12 (9.1) 3 (2.3) 8 (6.1) 3 (2.3) 64 (48.5) 30 (22.7) 0.004 0.31 0.11 0.10 0.02 <0.001 0.68 0.003 Functional class, n (%)  I  II  III 20 (11.8) 116 (68.2) 34 (20.0) 2 (6.1) 18 (54.5) 18 (39.4) 18 (13.6) 98 (74.2) 16 (12.1) <0.001 Exertion dyspnea 145 (85.3) 36 (94.7) 109 (82.6) 0.06 Orthopnea 16 (9.4) 8 (21.1) 8 (6.1) 0.005 Paroxysmal nocturnal dyspnea 6 (3.5) 2 (5.3) 4 (3.0) 0.51 Lung crackles 29 (17.1) 11 (28.9) 18 (13.6) 0.03 Total B-line sum 5.6 ± 10.1 14.1 ± 15.0 3.2 ± 6.4 <0.001 Biomarkers  NT-proBNP, ng/L  GDF-15, ng/L  hsTnT, ng/L  sST2, ng/mL  CA-125, U/mL  hsCRP, mg/L  IL-33, pg/mL  IL-1β, ng/mL  IL-6, pg/mL  sNEP, ng/mL  TNF-α, pg/mL  TNFRSF1A, ng/mL 202 (104-640) 1708 (1175-2511) 11.9 (6.7-21.7) 27.5 (21.8-37.6) 13.7 (9.7-22.7) 3.4 (1.9-7.3) 93.7 (93.7-601.2) 0.34 (0.27-0.44) 4.4 (2.9-7.2) 0.209 (0.062-0.605) 56.3 (49.9-67.5) 1.88 (1.45-2.39) 1350 (666-3551) 3133 (2040-4075) 24.8 (13.8-39.9) 39.6 (32-56.6) 21.4 (11.4-55.4) 5.3 (2.3-19.2) 93.7 (93.7-548) 0.33 (0.25-0.45) 6.3 (4.3-14.3) 0.206 (0.062-0.465) 60.8 (52.7-71.1) 2.39 (1.72-3.56) 148 (88-289) 1470 (1113-2117) 9.6 (5.9-15.9) 25.2 (20.7-33.8) 13 (8.8-20) 3.1 (1.7-5.8) 93.7 (93.7-632.3) 0.35 (0.28-0.43) 3.9 (2.8-5.9) 0.208 (0.062-0.630) 55.1 (49.1-66.1) 1.75 (1.42-2.21) <0.001 <0.001 <0.001 <0.001 <0.001 0.005 0.78 0.60 <0.001 0.62 0.02 <0.001 Data are expressed as mean (standard deviation), median (percentiles 25th-75th), or absolute numbers (percentages). BMI = body mass index; CA-125 = cancer antigen 125; COPD = chronic obstructive pulmonary disease; eGFR = estimated glomerular filtration rate; GDF-15 = growth differentiation factor 15; HF = heart failure; hsCRP = high-sensitivity C-reactive protein; hsTnT = high-sensitivity troponin T; IL = interleukin; LVEF = left ventricular ejection fraction; NT-proBNP = N-terminal pro-brain natriuretic peptide; NYHA = New York Heart Association; sNEP = soluble neprilysin; sST2 = soluble interleukin-1 receptor-like 1; TNF-α = tumor necrosis factor α; TNFRSF1A = TNF receptor superfamily member 1A. under investigation. Although there is a growing interest on both, added value for better patient diagnosis and mana- gement has scarcely been studied. In our study, we asses- sed the correlation between B-lines and a biomarker panel in primary care outpatients with new-onset nonacute HF suspicion. Our results showed that (i) total B-line sum observed by LUS was significantly—although moderately— associated with several biomarkers of active pathogenic pathways in HF, especially with those related to congestion and inflammation; and (ii) hsTnT, a biomarker related to Domingo et al 11 © 2020 The Authors. Published by AboutScience myocardial injury, was mainly and unexpectedly associated with total B-line sum. HF diagnosis can be difficult at early stages, especially in women and old patients with comorbidities. Our patients were elderly, mainly women, and not very symptomatic. Less than 10% of patients had lung congestion symptoms and only 17% crackles. In this context, HF was only confirmed in 22.4% of patients since Framingham criteria, despite being highly specific, have a poor sensitivity for HF diagnosis in ambula- tory patients. Hand-held devices could be easily incorporated in pri- mary care and have shown a good correlation with standard ultrasound equipment for B-line detection. In our study, we used the 8-zone technique adding 4 posterior zones since these areas are the first that show signs of congestion and could add accuracy in outpatients. As expected and according to previous studies in acute HF, total B-line sum was significantly associated with NT- proBNP levels. However, we projected a highest correlation, since increased intracardiac filling pressures often precede lung congestion. Nevertheless, in mildly symptomatic pri- mary care patients, pulmonary congestion is not always present unlike hemodynamic dysfunction. sST2 levels also correlated with B-line count in a similar level that NT-proBNP. ST2 is a member of the IL-1 receptor family linked to myo- cardial fibrosis and adverse remodeling, both related to diastolic dysfunction and increased end-diastolic pressures that can contribute to pulmonary congestion. sST2 has been described as a 3-in-1 biomarker and provides insight into the hemodynamic, inflammatory, and pro-fibrotic/remodeling burden of the myocardium (4). Total B-line sum also corre- lated with GDF-15, a marker of cell injury inflammation, oxi- dative stress, and hypoxia. These results are consistent with previous studies where GDF-15 may indicate a greater syste- mic inflammatory response in old patients and those with HF and preserved EF (5), as was the population of our study. It is remarkable that these two biomarkers, being in part inflam- matory biomarkers but also associated with other several pathogenic pathways in HF, correlated with total B-line sum in a greater degree than the more “pure” inflammatory ones (hsCRP, IL-1β, IL-33, IL-6, TNF-α, and TNFRSF1A). Maybe the more remarkable finding was the high corre- lation of total B-line sum with hsTnT, a biomarker of myocar- dial injury frequently elevated in patients with HF without coronary ischemia. Unexpectedly, hsTnT correlation was even higher than that observed with NT-proBNP and sST2. Recently, Myhre et al (6) showed that high-sensitive car- diac Troponin T (hs-cTnT) concentrations were associated with worse diastolic function, suggesting that high levels of hs-cTnT may serve as an early marker of subclinical altera- tions in diastolic function that may lead to a predisposition to HF. Finally, although there was a statistically significantly cor- relation between total B-line sum and CA-125, we anticipated a higher correlation, since both are surrogates of pulmonary and systemic congestion (7), respectively. Congestion plays a major role in acute HF syndromes; however, it is known that severity and organ distribution are largely heterogeneous. In fact, our primary care patients showed low percentages of congestion signs or symptoms. Limitations of our study include the limited sample size and the low incidence of HF since our target was primary care patients with mild symptoms and suspicion of HF. These facts might have an impact on the external validity of the study. Also HF diagnosis was performed by a single cardiologist. Although larger studies in diverse populations are needed, our data are hypotheses generating correlations between B-lines, a surrogate of pulmonary congestion, and of biomar- kers in HF patients. Conclusion In primary care outpatients with new-onset nonacute HF suspicion, total B-line sum is significantly—although moderately—associated with several biomarkers of conge- stion and inflammation, and remarkably with hsTnT. TABLE II - Correlations between total B-line sum and studied biomarkers NT-proBNP GDF-15 hsTnT sST2 CA-125 hsCRP IL-6 IL-33 IL-1β sNEP TNF-α TNFRSF1A Unadjusted R 0.32 0.27 0.35 0.32 0.21 0.22 0.27 –0.02 –0.08 0.10 0.09 0.18 p-value <0.001 <0.001 <0.001 <0.001 0.007 0.004 <0.001 0.81 0.28 0.21 0.26 0.02 Adjusted by age and sex R 0.29 0.23 0.34 0.30 0.19 0.21 0.24 –0.01 –0.11 0.09 0.08 0.15 p-value <0.001 0.003 <0.001 <0.001 0.02 0.007 0.002 0.93 0.18 0.24 0.32 0.06 Adjusted by age, sex, BMI, and eGFR* R 0.29 0.23 0.36 0.29 0.17 0.20 0.23 –0.01 –0.10 0.09 0.07 0.14 p-value <0.001 0.003 <0.001 <0.001 0.03 0.009 0.003 0.93 0.20 0.24 0.39 0.07 *estimated by CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). BMI = body mass index; CA-125 = cancer antigen 125; eGFR = estimated glomerular filtration rate; GDF-15 = growth differentiation factor 15; hsCRP = high- sensitivity C-reactive protein; hsTnT = high-sensitivity troponin T; IL = interleukin; NT-proBNP = N-terminal pro-brain natriuretic peptide; sNEP = soluble neprilysin; sST2 = soluble interleukin-1 receptor-like 1; TNF-α = tumor necrosis factor α; TNFRSF1A = TNF receptor superfamily member 1A. Lung ultrasound and biomarkers for heart failure diagnosis in primary care12 © 2020 The Authors. Published by AboutScience Acknowledgments The authors thank Jaume Barallat and Adriana Cserkóová for their valuable technical support in sample processing. Disclosures Conflict of interest: The authors declare no conflict of interest. Financial support: This work was supported by La Marató de TV3 [PI 201510.10], the Primary Healthcare University Research Institute IDIAP-Jordi Gol, and the Catalan Society of Family Physicians (CAMFiC). References 1. Platz E, Jhund PS, Campbell RT, McMurray JJ. Assessment and prevalence of pulmonary oedema in contemporary acute heart failure trials: a systematic review. Eur J Heart Fail. 2015;17:906-916. 2. Berezin AE. Prognostication in different heart failure pheno types: the role of circulating biomarkers. J Circ Biomark. 2016;5:6. 3. Conangla L, Domingo M, Lupón J, et al. Lung ultrasound for heart failure diagnosis in primary care. J Card Fail. 2020 Jun 6:S1071- 9164(19)31822-6. doi: 10.1016/j.cardfail.2020.04.019. Online ahead of print. 4. Pascual-Figal DA, Bayes-Genis A, Asensio-Lopez MC, et al. The interleukin-1 axis and risk of death in patients with acutely decompensated heart failure. J Am Coll Cardiol. 2019;73:1016-1025. 5. Wollert KC, Kempf T, Wallentin L. Growth differentiation fac- tor 15 as a biomarker in cardiovascular disease. Clin Chem. 2017;63:140-151. 6. Myhre PL, Claggett B, Ballantyne CM, et al. Association between circulating troponin concentrations, left ventricular systolic and diastolic functions, and incident heart failure in older adults. JAMA Cardiol. 2019;4:997-1006. 7. Núñez J, Bayés-Genís A, Revuelta-López E, et al. Clinical role of CA125 in worsening heart failure: a BIOSTAT-CHF study subanalysis. JACC Heart Fail. 2020;8:386-397.