Acute Bronchiolitis: Is There a Role for Lung Ultrasound?

Introduction: Viral bronchiolitis is a common cause of lower respiratory tract infection in the first year of life, considered a health burden because of its morbidity and costs. Its diagnosis is based on history and physical examination and the role of radiographic examination is limited to atypical cases. Thus far, Lung Ultrasound (LUS) is not considered in the diagnostic algorithm for bronchiolitis. Methods: PubMed database was searched for trials reporting on lung ultrasound examination and involving infants with a diagnosis of bronchiolitis. Results: Eight studies were suitable. Conclusions: This review analyzed the current evidence about the potential usefulness of LUS in the clinical management of bronchiolitis. Literature supports a peculiar role of LUS in the evaluation of the affected children, considering it as a reliable imaging test that could benefit the clinical management of bronchiolitis.


Introduction
Bronchiolitis is a lower respiratory tract infection (LRTI), resulting from the inhalation of virus-containing droplets. The most common cause is the Respiratory Syncytial Virus (RSV), identified in 50-80% of patients, and responsible for the most severe infection.
Bronchiolitis usually occurs during the first year of life, with a peak incidence between 3 and 6 months of age, and represents the main reason for hospitalization in the young population [1].
The diagnosis of bronchiolitis is based on the history and the physical examination [2]. Viral testing, bacterial cultures, complete blood count and blood gas, pulse oximetry and chest radiography are not routinely included in the diagnostic workup.
In the last decades, Lung Ultrasound (LUS) has been proposed as a diagnostic tool in many diseases in the pediatric and neonatological settings [3][4][5]. In particular, the anatomical conformation of children makes the sonographic evaluation of the lungs possible for the creation of acoustic windows by the relatively unossified thorax of children, the large thymus gland and the thinner subcutaneous tissue [6].
Currently, LUS is not considered in the diagnostic algorithm for bronchiolitis, even though its usefulness has been tested in several studies in the last years as an emerging diagnostic tool. Its main benefits are the absence of the irradiation risk, the possibility of a bedside evaluation, and the short time needed for a report.
In this review, we analyze the current evidence about the potential usefulness of LUS in the clinical management of bronchiolitis.

Methods
An exhaustive search for eligible studies has been performed, using the PubMed database as data source. The following subject MeSH headings were used: "Ultrasonography" [Mesh], "Bronchiolitis"[Mesh], "Bronchiolitis, Viral" [Mesh], "Lung" [Mesh], "Infant" [Mesh]. Furthermore, free text and proper Boolean operators "AND" "OR" were also included to be as comprehensive as possible. Additional studies were sought using references in articles retrieved from searches.
A.D.M. and M.M. screened titles and abstracts from the database, removing duplicates, and analyzed the full-text version of the selected studies to assess their eligibility. All the chosen articles have been published in the last eight years, starting from 2011, when the first study about the use of LUS in acute bronchiolitis was performed. The exclusion criteria were: articles not related to the field of interest, studies without relevant conclusions, specific types of articles (case reports, editorials or letters, comments, review articles). Search limits were set for trial involving only human subjects.
Moreover, this review has been conducted according to the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) Guidelines to ensure a transparent reporting of a systematic review [7].
A systematic approach for the data extraction has been applied to each considered study: main features (language, authors, type of study, year of publication, purpose), methods (study population), clinical and/or radiographic examination, sonographic examination (bronchiolitis-specific LUS findings), results about LUS diagnostic usefulness considering primary and potentially secondary outcomes (including sensitivity, specificity, positive and negative predictive values, accuracy).

LUS in Management of Bronchiolitis
We found eight trials reporting on lung ultrasound examination and involving infants with a diagnosis of bronchiolitis (Table 1).

LUS Findings in Bronchiolitis
The first study investigating the usefulness of LUS as a potential tool for the diagnosis of bronchiolitis was performed by Caiulo et al. in 52 young children under 16 months of age affected by bronchiolitis [8]. They aimed to describe the LUS findings in children with bronchiolitis and to compare the diagnostic accuracy of LUS and CXR.
They underwent an LUS examination on the same day of the CXR, and the LUS follow-up was stopped only after a complete normalization of lung appearance that occurred between day 10 and day 12 in all the affected children. The severity of the disease was assessed by a modified version of Downes' Score, one of the main scoring systems for respiratory distress in newborns, based on five parameters: PO2 (the partial pressure of oxygen usually decreases due to airways obstruction), inspiratory breath sounds, use of accessory muscles, presence of expiratory wheezing and respiratory rate. Each criterion was rated on a scale from 0 to 2, and then, a total score was calculated. LUS findings varied according to the disease severity: a prevalence of few isolate B-lines in the mildest form, a higher frequency of single subpleural consolidation and areas of white lung in the moderate disease, the presence of multiple lung consolidations, pleural line anomalies and pleural effusions in the most severe ones. In all cases, the described signs completely disappeared during the monitoring of patients.
Furthermore, they noted that LUS was able to identify lung abnormalities not revealed by CXR, such as minimal pleural effusion and small pneumothorax. In particular, nine patients with a course of bronchiolitis had normal CXR and abnormal LUS. They sustained a higher reliability of LUS compared to CXR for diagnosis and follow-up of bronchiolitis and a lower time in execution of the test.
Although guidelines limit the use of CXR for the diagnosis of bronchiolitis only to atypical presentations (toxic appearance, bacterial superinfection or pneumothorax as complications) [2] and British Thoracic Society (BTS) guidelines [16] do not consider CXR as a routine test in the management of community-acquired pneumonia, CXR is performed in around 50% of bronchiolitis cases, with the main goal to rule out pneumonia. Apart from the danger related to the exposure of children to ionizing radiations, the use of CXR in bronchiolitis management has been associated with other limitations: the increased prescription of useless antibiotics for the similar radiographic appearance of consolidation and atelectasis, the impossibility to distinguish bacterial and viral pneumonia, variable inter-observer interpretation, high medical costs and longer time [17].
Other authors compared findings at LUS and CXR of children with RSV-associated bronchiolitis through a retrospective study. Jaszczolt et al. [9] underlined the utility of LUS again, considering safety, short time of examination, high sensitivity in finding pleural effusion, small consolidations and signs of interstitial infiltrations. On one hand, the classification of LUS findings included liver-like hypoechoic consolidations (small subpleural < 10 mm in diameter and > 10 mm in diameter with air bronchogram), interstitial syndrome with B7 lines (7 mm between two different lines), alveolar-interstitial syndrome with B3 lines (distance around 3 mm between two different lines), pleural effusion. On the other hand, the considered CXR anomalies were inflammatory infiltrate, hilar enlargement, peribronchial cuffing, pleural effusion and lung hyperexpansion. The most common findings of bronchiolitis at LUS were alveolar-interstitial syndromes and medium-sized subpleural consolidations with air bronchogram with maximum diameter of 30 mm, especially in the inferior and posterior areas of the lungs. The main result was that LUS may be a useful tool in the diagnosis of bronchiolitis for its high sensitivity in finding pleural effusion, small consolidations and signs of interstitial infiltrations. Finally, they concluded that a direct comparison of both methods was not possible, considering the greater ability of CXR in showing bronchi and hilar regions or lung hyperinflation.

LUS and Clinical Progression of Bronchiolitis
Another group of studies demonstrate the possibility to predict the clinical progression of bronchiolitis through different proposed ultrasound scores.
Basile et al. [10] performed an observation cohort study in 106 infants with bronchiolitis, diagnosed with physical examination and lung ultrasound scans. The severity of the disease was described according to a clinical score (tachypnoea, dyspnoea, use of accessory muscles and breath sound) and a sonographic score (presence of B-lines, subpleural lung consolidations, bilateral involvement of intercostal spaces): a high agreement of 90.6% was demonstrated between the two scores, a further demonstration of the accuracy of LUS in the diagnosis and management of bronchiolitis in children. In addition, authors discovered that LUS was able to identify children in need of supplementary oxygen with a very high specificity (98.7%) and sensitivity (96.6%), considering as main indexes subpleural lung consolidation > 1 cm in the posterior area or a severe interstitial syndrome according to the number of involved bilateral intercostal spaces. They also found an excellent inter-observer ultrasound concordance between diagnoses made by different sonographers.
Zoido Garrote et al. [11] confirmed the possibility to predict the clinical progression of bronchiolitis, using different clinical and ultrasound scores. This observational prospective study determined the presence of a moderate correlation between findings at early LUS (at the time of admission) and the severity of acute bronchiolitis, also taking its clinical progression into consideration. The main research areas were the need of Intensive Care Unit admission, the use of Invasive Ventilation, the length of hospitalization and the duration of oxygen therapy. As previously discussed, the main LUS findings associated to bronchiolitis were pleural line anomalies (PAs) with thickening and loss of lung sliding, extent of interstitial syndrome (AIS) as shown by B-lines up to the white lung appearance, the presence of subpleural consolidations (SCs) with a classification according to their number and their dimension. The total LUS score was in the range of 0-50 points, resulting from the sum of the scores for each individual zone (five areas: anterior parasternal, anterior axillary, posterior axillary, posterior paravertebral, posterior linea scapularis). The authors concluded that their results were consistent with the ones of the aforementioned studies [8,12] regarding the correlation with the clinical severity scores, and they confirmed the presence of a higher percentage of pulmonary anomalies in the posterior paravertebral and subscapular area, probably due to gravity or to the obligate supine position of the child.
An additional prospective, multicenter research trial has demonstrated that LUS could be a helpful tool to identify patients at risk for any type of respiratory support (non-invasive or invasive ventilation, except conventional low-flow oxygen through nasal sprongs) in the acute phase of bronchiolitis, starting from the idea that clinical scores have proven to be inaccurate in predicting prognosis. In fact, Bueno-Campaña et al. [12] have analyzed the data of 145 infants < 6 months with the aim to build a score based on LUS findings (presence and localization of B-lines, B-lines confluence and/or consolidation) performed at admission, but also on clinical data (age < or > 1 month and Wood-Downes-Ferres Score, typically used for asthma). Their results were consistent with the ones of the previously described studies, especially regarding the increased association of NIV with posterior consolidations > 1 cm. Moreover, it was the first study to explore LUS usefulness as a predictive tool for the need of respiratory support other than supplemental oxygen in acute bronchiolitis [12].
Another recent paper published in 2018 examined the use of LUS in infants admitted in the Paediatric ICU for severe acute bronchiolitis. This prospective observational single-center study was conducted in France by Taveira et al. [13] to evaluate the correlation between a quantitative LUS score and the length of Non-Invasive Ventilation (NIV) in 47 children under 6 months of age with severe bronchiolitis. The severity of the LUS-diagnosed lesions was considered for each of the six quadrants the lung was divided in, according to a score in the range of 0-2 points (total range 0-24 points): 0 for a healthy lung or a mild interstitial-syndrome with few B-lines, 1 for a severe interstitial syndrome with compact B-lines and with lung appearance, 2 for the presence of consolidations seen as a hypoechoic area with blurred margins and air bronchograms. Moreover, the number of involved intercostal spaces was defined with a horizontal scan. The results of the study did not confirm the primary aim to predict NIV length (CPAP or high-flow nasal cannula), nor an association with length of hospitalization or the clinical score mWCAS. The only statistically significant correlation was between the number of affected intercostal spaces on the right side (white lung appearance, typical of bronchiolitis) and the length on the oxygen therapy (days), which was considered a secondary endpoint [13].

Differential Diagnosis and Complications
Finally, two studies concern the role of LUS in the differential diagnosis of bronchiolitis or in the evaluation of its complications. Varshney et al. [14] conducted a prospective cross-sectional study to describe how LUS could represent a useful tool in a Pediatric Emergency Department (PED) for the assessment of patients with respiratory distress, signs of respiratory tract infection (rhinorrhea and/or cough) and concomitant wheeze. In fact, this clinical presentation can be suggestive of three pathologies that require different types of treatment: bronchiolitis, which needs only supportive care, asthma, which requires bronchodilators, and pneumonia, which improves with antibiotics. Obviously, a CXR can be useful for the diagnosis of pneumonia, but it increases the risk of unnecessary antibiotic administration in children affected by bronchiolitis. A key point of the study was the definition of a positive LUS and the related diagnosis. An LUS was considered pathological if one or more findings occurred in any of the lung zones: more than three B-lines per intercostal space, consolidation, pleural line abnormalities or fluid. On the other hand, a negative LUS was characterized by the absence of aforementioned findings and the presence of A-lines and lung sliding. The important contribution of this study to the literature concerns the possibility to rule out the diagnosis of asthma in case of positive LUS that could direct the clinician toward a diagnosis of pneumonia (100% of patients) or bronchiolitis (46% of patients) [14].
At last, Biagi et al. [15] tried to assess the diagnostic accuracy and reliability of LUS in children with a clinical suspicion of secondary bacterial pneumonia. LUS was performed 12 h after a posteroanterior CXR (lateral projection is not recommended to prevent further radiation according to BTS guidelines), according to the division of each hemithorax in the three different regions. The criteria for a definite diagnosis of pneumonia on LUS were a hypoechoic area with blurred margins and decreased echogenicity of the pleural line, the presence of B-lines and dynamic bronchograms, and the absence of lung sliding. Furthermore, the typical findings of simple bronchiolitis were small subpleural consolidations, absence of air bronchograms, pleural line abnormalities, single or confluent B lines, which could be associated also with viral pneumonia. Two thirds of patients showed consolidation in the posterior lung areas. The authors deduced that LUS has a higher sensitivity for the diagnosis of pneumonia than CXR (100% versus 96%) and its specificity can reach 98.4% when only consolidations >1 cm are considered [15].

Discussion
Traditionally, pulmonary pathologies have been studied with CXR and the first description of LUS as alternative diagnostic imaging test was published in 2012 in the "International evidence-based recommendations for point-of-care lung ultrasound". This report defines LUS application in the diagnosis of pneumothorax, pleural effusion and lung consolidation, according to the literature of the previous 20 years. [18] However, the thought of LUS as a potential tool for the diagnosis of bronchiolitis is much more recent, since the first study about this topic was performed in 2011 by Caiulo et al.; they concluded that LUS could become the routine imaging modality for this group of patients. This data has been confirmed in successive studies, evaluating the association between clinical and US findings, generally through the use of scores, and the agreement between LUS and CXR. The main results were about a high reliability of LUS, considering a strong clinical-sonographic correlation, safety, short time of examination, high sensitivity in finding pleural effusion, small consolidations and signs of interstitial infiltrations. In fact, they underline the role of LUS to diagnose bronchiolitis and rule out primary pneumonia through specific findings, such as B-lines, white lung appearance and pleural line anomalies, or to diagnose secondary pneumonia as a complication, considering the peculiar presence of dynamic bronchograms and the absence of lung sliding.
We acknowledge some limitations of our study. The selected studies demonstrate a heterogeneity regarding the methods, the clinical and US scores and the considered outcomes. For this reason, the comparison does not allow definitive conclusions. However, good results have already been achieved considering the reliability of LUS in the management of bronchiolitis. In the future, as the number of trials regarding this imaging test grow, a single standardized US score could be validated and internationally used.

Conclusions
In summary, LUS can be considered a safe, non-invasive, quick, reproducible, low-cost, real-time imaging test that could benefit the clinical management of bronchiolitis: the interpretation of artifacts generated at the pleural surface in a bronchiolitis-specific pattern (presence of B-lines, subpleural lung consolidations, bilateral involvement of intercostal spaces) strictly correlates with the clinical evaluations in infants with bronchiolitis. A key point is that the routine use of LUS for children with bronchiolitis could reduce the number of performed CXR, the exposure to ionizing radiations with its dangerous consequences in children, i.e., the increased risk of cancer, and could avoid the useless administration of antibiotics. Future prospective studies may elucidate whether LUS can be used to help guide diagnostic and therapeutic decisions for affected children such as need for admission, oxygen therapy, ventilation and possibly to confirm the safety of outpatient management.
Author Contributions: All authors conceived and designed the study protocol. A.D.M. performed the literature search and assessed study details. A.A. wrote the first draft of the paper, which was critically revised by all the other authors. All the authors gave final approval of the version to be submitted and agreed to be accountable for the whole paper.

Conflicts of Interest:
The authors declare no conflict of interest.