Article

Title: Left ventricular myocardial dysfunction secondary to adverse ventricular-ventricular interactions in previously healthy infants with Respiratory Syncytial Virus Bronchiolitis…

Moises Rodriguez-Gonzalez 1,4*, Alvaro Antonio Perez-Reviriego1,4, Ana Castellano-Martinez2,4, Simon Lubian-Lopez3,4 and Isabel Benavente-Fernandez3,4

1 Pediatric Cardiology Division, Puerta del Mar University Hospital, Cadiz, Spain;doctormoisesrodriguez@gmail.com;alvaro.apr@hotmail.com

2 Pediatric Nephrology Division, Puerta del Mar University Hospital, Cadiz, Spain;anacastellanomart@gmail.com

3 Neonatology Division, Puerta del Mar University Hospital, Cadiz, Spain;isabenavente@gmail.com;slubian@yahoo.es

4 Biomedical Research and Innovation Institute of Cadiz (INiBICA), Research Unit, Puerta del Mar University Hospital

Received: date; Accepted: date; Published: date

Abstract: Aim: To investigate if the presence of left ventricular myocardial dysfunction (LVMD) assessed by Tei index (LVTX) may have a direct impact on the outcomes in Respiratory Syncytial Virus bronchiolitis (RSVB), and if NT-proBNP will increase the accuracy of traditional clinical and laboratory markers in predicting the severity of the disease. Methods: A single-centre, prospective, cohort study including healthy infants aged 1-12 month-old admitted due to RSVB between October 1, 2016 and April 1, 2017. All patients underwent clinical, laboratory and echocardiographic evaluation within 24 hours of admission. PICU admission was defined as severe disease. Results: We enrolled 50 cases of RSVB (median age of 2 (1-6.5) months; 40% female) and 50 age-matched controls. We observed higher values of LVTX in infants with RSVB than in controls (0.42 vs 0.36; p=0.008). A total of 9 (18%) cases presented LVMD (LVTX>0.5), with higher incidence of PICU admission (89% vs 5%; p<0.001). The diagnostic performance of NTproBNP to predict LVMD in infants with RSVB resulted high (AUC 0.91 (CI95% 0.79-0.98). The diagnostic yield of the predictive model for PICU admission that included NTproBNP was excellent (AUC 0.945, CI95% 0.880-1), and significantly higher than the yields for models without NTproBNP. Conclusions: LVMD could be present in healthy infants with RSVB, negatively impacting the outcome. NTproBNP seems to be an adequate biomarker for LVMD and subsequently outcome.

Keywords: Respiratory Syncytial Virus; NT-proBNP; Echocardiography; Pulmonary hypertension; Myocardial dysfunction; Tissue Doppler Imaging; Tei Index; Biomarkers; Infants.

1. Introduction

Respiratory Syncytial Virus bronchiolitis (RSVB) is the leading cause of lower respiratory infection and hospital admission among children up to 2 years of age worldwide [1]. Approximately 2-6% cases of RSVB will develop a severe form of disease, requiring ad mission at the pediatric intensive care unit (PICU) and mechanical ventilation (MV) [1,2]. RSVB constitutes approximately 13% of all PICU admissions [2]. Current guidelines recognise identification of specific risk factors (congenital heart disease (CHD), chronic lung disease (CLD), prematurity, etc.) and clinical evaluation as the best tools to asses severity, predict evolution and tailor management [3].

Cardiovascular involvement seems to be a relevant prognostic factor in RSVB. Cardiovascular complications, are present in up to 9% of cases of RSVB, and constitute the second most common extra pulmonary manifestations after infectious complications [4]. These events present usually in an abrupt and unexpected manner in those children with severe RSVB, and infants with CHD are particularly susceptible to have these complications and adverse outcomes [5]. Interestingly, nearly half of children admitted to PICU with RSVB are healthy prior to the clinical event [2]. In these patients the presence of acute lung injury secondary to RSVB can also lead to important cardiovascular effects, especially raising pulmonary vascular resistance and overloading the right ventricle (RV) [6-8]. Moreover, previous studies assessing plasmatic levels of cardiac troponin in RSVB suggest an underrecognized but yet clinically significant incidence of myocardial damage in this population [9-11]. Furthermore, RV global dysfunction in ventilated healthy infants has been reported [12]. Recently, we found that mild to moderate forms of PH could impact the outcome of healthy infants with RSVB [13].

Adverse RV-LV interactions and left ventricle (LV) myocardial dysfunction (LVMD) are emerging as important determinants of PH outcomes. PH can induce complex changes in LV geometry and causes an abnormal relaxation and a non-uniform contraction pattern in the LV wall, leading to LVMD [14-16]. However, most studies in healthy infants with RSVB found no abnormalities when assessing LVMD through conventional echocardiographic parameters [12,13,17-19]. Remarkably, there are no studies to date assessing LVMD in RSVB by more sensitive methods such as tissue Doppler imaging (TDI) echocardiography.

N-terminal pro-B-type natriuretic peptide (NT-proBNP) is a hormone synthesized and released into the circulation by ventricular myocytes in response to pressure/volume overload and an increase in myocardial wall stress [20]. Elevated serum NT-proBNP levels have been defined as a powerful biomarker in the diagnosis of PH, and both LVMD and RV myocardial dysfunction (RVMD) secondary to pulmonary diseases [21-25]. Of note, we recently showed how NTproBNP could be considered an adequate biomarker for PH in previously healthy infants with RSVB [13].

In this study we aimed to investigate the adverse RV-LV interactions and LVMD through TDI-echocardiography in previously healthy infants with RSVB. We hypothesized that acute PH with RV pressure overload may indeed have a direct impact on LV performance. We also hypothesized that those infants with LVMD are prone to develop a more severe form of disease. Finally, we sought to test NT-proBNP as a biomarker for LVMD, and to explore if NT-proBNP will increase the accuracy of traditional clinical and laboratory markers in predicting the severity of the disease.

2. Materials and Methods

2.1. Design, settings and study population:

This was a single-centre, prospective, cohort study including infants aged 1-12 month-old admitted to the Pediatric Department of our institution (a tertiary university-affiliated hospital in Spain) due to RSVB (determined by a confirmed RSV antigen testing) between October 1, 2016 and April 1, 2017. All patients underwent clinical, laboratory and echocardiographic evaluation within 24 hours of admission. We excluded patients with co-existing CHD or CLD, prematurity, those that received MV or intravenous fluid before assessment, and those with poor quality echocardiographic images or incomplete medical records. Severe cases were screened for coinfection and if existed they were also excluded. The control group consisted of age-matched healthy infants who underwent evaluation for heart murmur at our Pediatric cardiology outpatient clinic during the study period. These controls followed the same echocardiographic protocol as study patients. Our Institutional Review Board approved the study. Informed consent was obtained for all patients.

2.2. Clinical and laboratory assessment and outcomes:

The bronchiolitis score of Sant Joan de Déu (BROSJOD) [26]was used to assess severity at admission clinically. A BROSJOD score greater than 10 points is indicative of severe clinical state. Venous pH and pCO2 were determined, and respiratory acidosis (RA) was considered when pH<7.35 and pCO2>45 mmHg coexisted in the same patient. Plasma NT-proBNP levels at admission were determined using a commercially available electrochemiluminescent immunoassay kit (ElecSys 2010, Roche Diagnostics). The primary outcome was PICU admission during the hospitalization. PICU admission criteria for RSVB at our institution rely on the presence of: apnea, extreme bradycardia, need of respiratory support greater than high-flow nasal cannula oxygen therapy, or inotropic support.