High Neutrophil-to-Lymphocyte Ratio Is an Early Predictor of Bronchopulmonary Dysplasia

Background and Objective: Bronchopulmonary dysplasia (BPD) is a common complication in preterm infants; predicting the degree of BPD at an early life stage is difficult. Inflammation is a crucial risk factor for BPD pathogenesis, and the neutrophil-to-lymphocyte ratio (NLR) is a potential systemic inflammatory biomarker. We aimed to assess the predictive value of the NLR for BPD. Methods: We carried out a retrospective, single-center, observational study of neonates with gestational ages (GAs) <32 weeks and assessed the association between the NLR and BPD. Results: The study population included 296 preterm infants with BPD (n = 144) or without BPD (n = 152). Among the infants, 75 (25.3%) had mild BPD, 37 (12.5%) had moderate BPD, and 32 (10.8%) had severe BPD. The BPD group had a higher NLR at birth and at 72 h than the non-BPD group. The NLR cutoff value at 72 h for the prediction of BPD was 3.035 (sensitivity = 0.519, specificity = 0.964), and the area under the curve (AUC) was 0.714. The NLR cutoff value at 72 h for predicting severe BPD was 3.105 (sensitivity = 0.607, specificity = 0.819), with an AUC of 0.756. At the NLR cutoff value at 72 for the prediction of BPD, the AUCs were 0.640 and 0.970 in the preterm infants with EOS and congenital pneumonia, respectively. Conclusions: The NLR is an inexpensive, accessible and convenient tool; an increase in the NLR at 72 h could be an early predictor of BPD, especially severe BPD. Additionally, the NLR at 72 h could be a predictor of BPD in preterm infants with intrauterine infections.


INTRODUCTION
Due to improved neonatal care, the increased use of antenatal steroids, early surfactant and gentle ventilatory support, large numbers of preterm infants survive (1). However, immature lungs, inflammation, infection, hyperoxia, and invasive mechanical ventilation may result in bronchopulmonary dysplasia (BPD), which has become the most frequent respiratory complication among preterm infants (2). BPD is defined as the need for extra oxygen at 4 weeks of age and can be classified as mild, moderate or severe based on the oxygen and ventilatory support at 36 weeks of correct gestational age (3). Recently, BPD has been characterized by immature lung tissue with dysplasia of alveolarization and vascular development (4). Infants with BPD experience prolonged and recurrent hospitalizations, increased rates of other serious prematurity-related complications, possible lifelong cardiopulmonary function alterations, and possible nervous system development disruptions (5,6). These critical complications are often associated with severe BPD. Despite numerous pharmacologic treatment measures, respiratory care practices, and nutritional therapies, few have demonstrated efficacy in resolving BPD. The clinical prediction of BPD at an early stage in life is difficult but necessary.
Proinflammation is an important pathogenic factor of BPD (7). The levels of several proinflammatory cytokines, such as TNF-α, IL-1β, and IL-6, are increased in BPD patients (7)(8)(9), suggesting that an excessive inflammatory response is associated with the onset of BPD. The neutrophil-to-lymphocyte ratio (NLR), could be considered a new potential marker to assess systemic inflammation. A high NLR has been considered to be an independent prognostic indicator for a variety of cancers and cardiovascular diseases, sepsis, infectious conditions and chronic obstructive pulmonary disease (COPD) (10)(11)(12)(13). However, the relationship between the NLR and BPD remains unknown.
As inflammation plays a crucial part in the genesis and development of BPD, in this study, we aimed to evaluate the predictive value of the NLR for BPD.

Patients
All infants born at GA<32 weeks who were hospitalized in the neonatal intensive care unit (NICU) within 6 h after birth between January 2015 and January 2018 were included in the study. BPD was defined and categorized according to the criteria of the National Institute of Child Health and Human Development/National Heart, Lung, and Blood Institute/Office of Rare Diseases (NICHD/NHLBI/ORD) Workshop (3). The study process is presented in Figure 1. Patients with severe congenital dysplasia, including complex congenital heart disease (CHD), congenital gastrointestinal malformations, chromosomal abnormalities, and congenital metabolic diseases, and those with incomplete information or who died were excluded from the study. One with incomplete information because the lack of his mother's medical history. Two dead infants happened on 2 and 3 days after birth, separately. Of the 306 preterm infants initially included in the research, 10 were excluded. Of the remaining 296 infants, 144 infants developed BPD, and 152 infants did not develop BPD. The procedures of this study were conducted in accordance with the World Medical Association Declaration of Helsinki and received ethical approval from the ethical institutional review board of our hospital (the First Abbreviations: BPD, bronchopulmonary dysplasia; SGA, small for gestational age; NRDS, newborn respiratory distress syndrome; CHD, congenital heart disease; EOS, early-onset sepsis; PROM, premature rupture of membranes; CPAP, continuous positive airway pressure; SIMV, synchronized intermittent mandatory ventilation; WBC, white blood cell; N, neutrophil; L, lymphocyte; P, platelet; NLR, neutrophil-to-lymphocyte ratio; RBT, routine blood test; GA, gestational age; AUC, area under the curve; COPD, chronic obstructive pulmonary disease; ROC, receiver operating characteristic; GBS, Streptococcus agalactiae; NETs, neutrophil extracellular traps; MPO, myeloperoxidase; NE, neutrophil elastase; CRP, C reaction protein; PCT, prostate calcitonin. Affiliated Hospital of Wenzhou Medical University, No. 2018-137). Written informed consent was obtained from the parents of each patient.

Statistical Analysis
Continuous variables were expressed as the mean ± standard when normally distributed or the median (interquartile range) when non-normally distributed. The Pearson χ 2 test or Fisher's exact test was applied to analyze categorical variables. Comparisons of continuous variables between two groups were performed with t-tests or Wilcoxon tests. The Kruskal-Wallis test was used to analyze differences in continuous variables among three or more groups. Correlation coefficients of continuous variables were calculated and their significance evaluated using the Pearson correlation test. Receiver operating characteristic (ROC) curves were constructed, and the area under the curve (AUC) was obtained to determine the optimal cutoff value. Statistical significance was evaluated at P < 0.05. SPSS version 25.0 (SPSS Inc., Chicago, IL, USA) was used for all data analyses.

Infant Characteristics
Among the 306 preterm infants initially included in the study, 10 were excluded: 2 presented complex congenital heart disease, 1 presented congenital gastrointestinal malformation, 2 presented chromosomal abnormalities, 2 presented congenital metabolic diseases, and 3 had incomplete information or died in 72 h after birth. Of the remaining 296 infants, 144 developed BPD and 152 did not. Information on the infant characteristics is presented in Table 1. The number of cases of mild, moderate and severe BPD were 75 (25.3%), 37 (12.5%), and 32 (10.8%), respectively. The average gestational age was 29.85 ± 1.57 weeks, the average birth weight was 1397.08 ± 324.13 g, and 56.3% of the patients were males. Infants who developed BPD had a significantly lower gestational age and birth weight than those who did not develop BPD. Patients with Apgar scores <7, NRDS, EOS, or congenital pneumonia and patients born to a mother with a multiple pregnancy, cesarean delivery, or gestational diabetes mellitus or had received antenatal steroid therapy had increased rates of BPD. In the BPD group, the duration of mechanical ventilation, including continuous positive airway pressure (CPAP) therapy and synchronized intermittent mandatory ventilation (SIMV) therapy, was significantly longer than that in the non-BPD group. Relative to infants who did not develop BPD, infants who developed BPD had increased rates of intravenous nutrition therapy, increased lengths of NICU stay, and substantially increased hospital costs. Among the 122 patients with EOS, 17 patients had positive blood cultures for one of the following Escherichia coli (4 patients), Klebsiella pneumoniae (4 patients), Enterobacter cloacae (2 patients), Monilia albicans (2 patients), Staphylococcus epidermidis (4 patients), and Staphylococcus warneri (1 patient).

Comparison of the NLR Between the BPD Group and the Non-BPD Group
The NLR was compared between the BPD group and the non-BDP group for the entire period under study. The results are shown in Table 2. Relative to the N count in the non-BDP group, the N count in the BPD group was significantly increased at birth (7.64 ± 7.37 vs. 5.35 ± 3.72, P = 0.001) and at 72 h (7.95 ± 7.32 vs. 4.26 ± 2.80, P = 0.000), 1 week (7.24 ± 6.76 vs. 4.35 ± 5.12, P = 0.028) and 2 weeks (4.11 ± 2.82 vs. 3.53 ± 1.61, P = 0.036) of age. Relative to the NLR in the non-BDP group, the NLR in the BPD group was significantly increased at birth (3.04 ± 4.68 vs. 2.12 ± 2.99, P = 0.043) and at 72 h of age (2.77 ± 1.65 vs. 1.44 ± 1.09, P = 0.001).

Parameters Predictive of BPD
After analyzing the correlations of the N and L counts and the NLR between the BPD and non-BPD patients, we constructed ROC curves to obtain the sensitivity, specificity and cutoff point of each of the significant parameters for the prediction of BPD. The ROC curves of N count and the NLR for the prediction of

Parameters Predictive of Severe BPD
After analyzing the correlations of the N and L counts and the NLR among the groups of patients with differing severity of BPD, we constructed ROC curves to obtain the sensitivity, specificity and cutoff point of each of the significant parameters for the prediction of severe BPD in Figures 2F-J

Comparison of the NLR Between the BPD Group and Non-BDP Group Among Preterm Infants With or Without EOS
As BPD is associated with intrauterine infection, we divided the preterm infants into an EOS group and a non-EOS group.

Comparison of the NLR Between the BPD Group and Non-BDP Group Among Patients With or Without Congenital Pneumonia
The NLR was compared between the BPD group and non-BDP group among patients with congenital pneumonia and Frontiers in Pediatrics | www.frontiersin.org Continuous variables are expressed as mean ± SD. N, neutrophils; L, lymphocytes; NLR, neutrophil-to-lymphocyte ratio. *P < 0.05. those without. The results are shown in

Parameters Correlated With BPD in Preterm Infants With and Without Intrauterine Infections
After analyzing the correlations of the N and L counts and the NLR between BPD patients and non-BPD patients in preterm  infants with intrauterine infections and those without, we constructed ROC curves to obtain the sensitivity, specificity, and cutoff point of each the significant parameters for the prediction of BPD. The ROC curves of the NLR and N count for the prediction of BPD in preterm infants with EOS are displayed in Figures 3A-D

DISCUSSION
In the present study, we found that an increase in the NLR (at birth and at 72 h) was associated with BPD. Moreover, the NLR at 72 h was found to be the most promising marker for predicting BPD, especially severe BPD, at an early stage of life. Additionally, the NLR at 72 h was identified as a predictor of BPD in preterm infants with congenital infections. BPD is a severe chronic pulmonary disease of preterm infants that involves the disruption of alveolar and pulmonary vascular development. Inflammation is a well-known critical risk factor in the pathogenesis of BPD. Many studies have suggested an association between elevated levels of proinflammatory cytokines and BPD in patients (19)(20)(21). In the course of BPD, Ns and macrophages play crucial roles in lung and systemic inflammation in extrapulmonary tissue. In a previous study, preterm infants with NRDS who subsequently developed BPD had much higher and persistent numbers of Ns and macrophages in their bronchoalveolar lavage fluid than did those who recovered (22). Most likely, Ns are activated during inflammation and adhere to endothelial cells in the pulmonary vascular system, which triggers a series of injurious events. Ns are the first cellular defense of the natural immune system, with a short half-life, whereas Ls are the main cells of the adaptive immune system. N count has been demonstrated as early biomarker of BPD in preterm infants at birth (20,23). In our study, we found that increases in N count at birth, 72 h, 1 week and 2 weeks were associated with BPD; this association may be due to inflammation in the pulmonary and peripheral blood, especially the assembly of Ns. Our ROC analysis also indicated that N counts at 72 h and 1 week predicted BPD.
In recent years, the NLR has become a widely available marker of inflammation. Elevated NLR can be used as a marker similar to C-reactive protein (CRP) in the determination of increased inflammation in acutely exacerbated COPD (24). The NLR has been found to be more accurate than CRP in the prediction of bacteremia or the severity of community-acquired pneumonia (25). In a previous study, patients with septic shock at risk of early death had lower NLR at admission, and late death was associated with higher NLR at the first 5 days (10). We found that the NLR in the BPD group was higher than that in the BPD group at birth and 72 h, whereas the highest NLR was found in severe BPD patients. We confirmed that the NLR at 72 h had the capacity to predict BPD and severe BPD. A high NLR reflects systemic inflammation, which is strongly associated with BPD. We also confirmed that N count is significantly correlated with severe BPD; however, N count is not as relevant as the NLR. It is therefore reasonable to postulate that the NLR reflects the severity of BPD better than N or L count alone.
Intrauterine infections, such as EOS and congenital pneumonia, can lead to pulmonary inflammation, which increases the risk of BPD. We investigated the correlation between the NLR and BPD in preterm infants with and without intrauterine infections. Among the 122 patients with EOS, only 17 patients were positive for blood culture, so it was not possible to clarify the correlation between the NLR and BPD for different pathogens. However, we found that infants with EOS or congenital pneumonia had significantly higher rates of BPD and a higher NLR value at 72 h than did infants without these infections. Furthermore, infants with EOS had significantly higher values of NLR at birth and at 2 weeks, which indicated that the NLR is increased at an early stage in the presence of systemic inflammation, and the NLR was associated with an increased risk of BPD. We also confirmed that the NLR at 72 h was predictive of BPD with EOS or congenital pneumonia; however, this finding should be interpreted with caution as the sample size was low.
This study found that the NLR was a stronger predictor of BPD than was N count, which may provide a new basis for the future prevention and treatment of BPD. An increased NLR was associated with an elevated N count, suggesting that an increased neutrophil count in the early blood circulation is associated with the occurrence of BPD. Interestingly, neutrophil extracellular traps (NETs) formed by neutrophil elastase (NE), DNA, histones, and myeloperoxidase (MPO) (26) have been shown to mediate proinflammatory effects in vivo and in vitro and have served as useful biomarkers for a variety of diseases, including sepsis, acute lung injury, cystic fibrosis and COPD (27)(28)(29). Additionally, previous studies found that the activated neutrophilic polymorphonuclear leukocyte (PMN)derived exosome was related to the pathogenesis of BPD in extracellular matrix (ECM) homeostasis disorders (30). We speculate that a high NLR leads to NET overexpression and activates PMN to release exosomes, thereby playing significant roles in the progression of BPD. This speculation will be investigated in a future study.
The present study has some limitations. First, the study was a single-center, retrospective, observational study; thus, the potential for residual confounding could not be eliminated. The results must be confirmed by studies at other centers. Second, we did not measure conventional proinflammatory markers such as CRP, IL-6, or prostate calcitonin (PCT) and analyze the correlations of these parameters with N count and the NLR.
In summary, we found that the NLR at 72 h might be considered an early significant indicator of both BPD and severe BPD in preterm infants, including those with intrauterine infections. The individualization of therapeutic interventions should be considered based on this marker.

WHAT IS KNOWN
Bronchopulmonary dysplasia is a common complication in preterm infants; predicting the degree of BPD at an early life stage is difficult. Inflammation is a crucial risk factor for BPD pathogenesis, and the NLR is a potential systemic inflammatory biomarker.

WHAT IS NEW
In the present study, we found that the NLR at 72 h might be considered an early significant indicator of both BPD and severe BPD in preterm infants, including those with intrauterine infections.

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/Supplementary Material.

ETHICS STATEMENT
The study was conducted ethically in accordance with the World Medical Association Declaration of Helsinki and received ethical approval from the ethical institutional review board of our hospital (the First Affiliated Hospital of Wenzhou Medical University, No. 2018-137). We obtained written informed consent from all patient guardians in this study.

AUTHOR CONTRIBUTIONS
YS provided the idea of this experiments and managed it, and provided fund. CC managed the experiments, analyzed the results, and was involved in manuscript preparation. XZha and AS were involved in data collection and analysis. XW analyzed the results and was involved in manuscript preparation. YZ was involved in manuscript preparation. SC analyzed the results. XZhe data collection. CL provided the idea of this experiments and managed it.