DISCUSSION
Our prospective observational study is the first study that evaluates the LUS findings of cardiopulmonary interactions in acutely ill children with elevated pro-BNP levels, with the aim of establishing the specific LUS pattern in this category of patients without primary lung diseases.
We found that LUS findings in these acutely ill children are different from the ultrasound pattern of other categories of children and in particular 1) children with acute lower respiratory tract infections and 2) healthy infants (Figures 1,2 and 3). In particular, we found that no children belonging to the category ofacutely ill children had short vertical artifacts but all had multiple long artifacts that were well delineated and non-blurred, non-confluent, with localization mainly at the bases and bilateral distribution. At the origin of these artifacts, the pleural line always appeared regular and none had subpleural consolidations. Many had anechoic pleural effusions, most of them unilateral.
We also found that the distribution pattern of vertical artifacts (with or without spared areas) is influenced not only by the cause, severity and timing of onset of pulmonary congestion - as already described in the literature7,8,10 - but also by the level of pro-BNP and the type of pathology (cardiac disease; systemic inflammatory disease/sepsis; heart disease and systemic inflammatory disease/sepsis) that causes the increase in pro-BNP (Tables 5 and 6). However, due to the scarcity of the comparison population, we can consider this last detection as a preliminary result pending the execution of a study with a larger sample of patients, in order to understand if there is a correlation between the ultrasound results and their distribution with pro-BNP levels and / or specific diagnosis.
From a theoretical point of view, our results reinforce the need for a better understanding of the physical mechanisms that generate artifacts27 even in children.
In the presence of cardiogenic edema, ARDS, interstitial lung diseases, bronchiolitis and non-consolidating pneumonia and contusions, part of the lung volume originally occupied by air, can be replaced with water, connective, cells, hyaline membrane or edematous tissue, eventually creating acoustic traps for the US beam containing a medium that is physically (in terms of acoustic impedance) very different from the surrounding environment (air).7,8,10,25
Therefore, the B lines, in their variable appearances, indicated a loss of peripheral lung aeration (without tissue consolidation) due to interstitial disease or simply to lung deflation without histologic changes.7,8,10,25However, the B lines, in the absence of an analysis of their appearance, cannot easily differentiate the cause.7,8,10,12,25 Studies performed on physical models and in adult patients7-12,21,25,have shown that the vertical artifacts generated by an inflammatory lung have a different appearance from those generated by cardiogenic edema and / or pulmonary congestion not primarily due to lung diseases.7-12,21,25
In the first case, as for example in the context of lung inflammation or in ARDS, the random peri and intralobular distortion can explain the appearance of acoustically permissive irregular channels and consequently the appearance of different arrangements of the B lines, pleural irregularities, consolidations, inhomogeneous isolated air spaces and inhomogeneous edema.7,8,10,25 Inhomogeneous edema, alveolar collapses, fibrosis and lung inflammation contribute to the development of pleural irregularities and the reduction of lung compliance and pleural flow in ARDS and other lung inflammatory diseases.
In the second case, the presence of homogeneously distributed edema due to the absence of inflammation and / or fibrosis causes the secondary interlobular septa to be thickened but anatomically intact and therefore can act as acoustic traps in which specific frequencies give rise to separate B lines , uniform, bright and long, associated with regular pleural line, with no spared areas especially in early cardiogenic edema.7,8,10,25
When cardiogenic pulmonary edema is severe or in an advanced stage, some LUS findings can be superimposed on the LUS findings of ARDS such as: - irregularities of the pleural line with small supleural consolidations or - the presence of ”spared areas”.8,9
None of the acutely ill patients that we enrolled for the study had a clinical and functional picture of ARDS: in fact, these are patients with isolated heart disease with early cardiogenic pulmonary edema and patients with systemic inflammatory syndrome and / or sepsis who did not have such a severe picture as to develop ARDS. In agreement with the data in the literature8,9, in none of the acutely ill patients we found short vertical artifacts, pleuric line irregularities and sub pleural consolidations.
Furthermore, a recently published2 study showed that healthy young children can also generate vertical artifacts and this discovery has provided us with new information and a new group of patients to study the B lines. We have in fact applied these new results in the design. and in the execution of our study.
From a practical point of view, our study confirms what we have learned so far from these preclinical and clinical studies.7-12,21,25 It underlines the idea that vertical artifacts must be interpreted according to the clinical scenario case by case and how through this interpretation it is possible to find specific patterns for some categories of patients different from each other in order to use LUS as an aid to orient ourselves more easily in the differential diagnosis of the pathologies that are accumulated by the presence of SIS.
In particular, in consideration of the finding of statistically significant differences (Table 4) for all the ultrasound findings studied between the group of acutely ill children compared to that of bronchiolitis but also compared to that of healthy patients , it was possible to find a specific ultrasound pattern of cardio-pulmonary interactions.
This means that the finding in a child of a SIS with multiple, bright, long, separate and non-confluent B-lines/long vertical artifacts deriving from a normal and regular pleural line, in the absence of subpleural consolidations, is strongly predictive of cardiogenic pulmonary edema or pulmonary congestion in the course of systemic inflammatory disease/sepsis (Figure 1).
Regarding the latter subcategory of acutely ill of patients, in our clinical practice we have observed elevated levels of pro-BNP and lung ultrasound findings suggesting a pulmonary imbibition picture even in patients with systemic inflammatory pathology / sepsis in the absence of cardiac dysfunction. Therefore we have decided to include in the category of acutely ill children not only patients with heart disease but also those with systemic inflammatory pathology with and without cardiac dysfunction (such as rheumatic carditis, Kawasaki disease, PIMS-ST, sepsis) with high levels of pro-BNP.
BNP and NT-proBNP are released into the bloodstream by cardiomyocytes in response to stress and pressure placed on the atria and ventricles. pro-BNP is a biomarker of cardiac overload associated with the severity of cardiovascular disease14and the severity of sepsis.15,16,17 It has long been proven that patients show signs of heart failure during sepsis.14-17Indeed, the destructive systemic inflammatory response and intensive resuscitation occurring during sepsis result in widespread organ overload and damage, including myocardial damage.15-17
However, today we know that in the course of sepsis and / or systemic inflammatory disease, the increase in plasma levels of pro-BNP can be present even in the absence of damage or dysfunction on the cardiomyocytes.16,17 In particular, Meader and colleagues and by Renana and colleagues16,28, have shown that in patients with sepsis, the level of BNP in the blood increases even without heart failure and that it is positively correlated with the level of CRP, suggesting the role of inflammation in the increase of BNP.16,28 Furthermore, experimental studies17,29 have shown that endotoxins and pro-inflammatory cytokines released in conspicuous quantities during sepsis or systemic inflammation, as well as playing a direct role on cardiac toxicity, directly increase the gene expression of the BNP in cardiomyocytes. This would explain the higher levels of this biomarker (positively correlated with the level of CRP and IL-6), in systemic inflammatory processes even in the absence of significant organic or functional cardiac alterations. 16,17,28,29
In this sub-category of acutely ill of patients with systemic inflammatory disease / sepsis, we found a lung ultrasound picture of pulmonary congestion. But, although none of our enrolled acutely ill children developed ARDS during sepsis and / or systemic inflammatory disease, we know that these patients can also develop an inflammatory of ARDS pulmonary picture9,10,30 especially in the most severe cases. In these cases, according to what emerged from our study, LUS would support the clinic and laboratory data in distinguishing the two conditions with different ultrasound patterns - congestion and inflammation - with clinical implications, important diagnostics and therapeutics.
We have also found that the distribution pattern of long vertical artifacts (“with or without spared areas”) is influenced not only by the severity and time of onset of pulmonary congestion8, but also by the level of pro-BNP and the type of pathology (cardiac disease, systemic inflammatory disease / sepsis, cardiac disease and systemic inflammatory disease / sepsis) which causes the increase in pro-BNP (Tables 5 and 6).
In fact, in the literature there is discussion on homogeneous septal distribution “without spared areas” in the case of early cardiogenic edema. In the later stages of pulmonary edema, the thickening of the interlobular septa becomes excessive and goes beyond their capacity, therefore alveolar flooding occurs when the secondary interlobular septae are maximally distended by the transudative fluid. Flooding and alveolar instability, together with the collapse of peripheral airspace, generate many different types of acustic channels and consequently different artifacts with many arrangements. Therefore, pulmonary cardiogenic edema, in the later stages of the pathology, can lose the ultrasound characteristic of the septal and homogeneous B lines and increasingly tend to resemble the edema of early ARDS without consolidation “with spared areas”.8,9,10
We found that in 85.71% of cases of children belonging to the group with proBNP> 1000 pg / ml (most of these belong to the group of heart disease with or without systemic inflammation / sepsis as shown in Table 1 and Table 2 ) vertical artifacts are homogeneously distributed “without spared areas” (as from early cardiogenic edema). On the other hand, 100% of acutely ill patients with pro-BNP <1000 pg / ml have destruction of long vertical artifacts ”with spared areas” (Table 5). However, the majority of children with pro-BNP <1000 belong to the group of acutely ill children with systemic inflammatory disease / sepsis in the absence of organic and / or functional damage in the myocardium (Table 1, Table 3). Table 6 also shows that 87.5% of patients with inflammatory pathology and / or sepsis in the absence of cardiac dysfunction have a distribution of artifacts ”with spared areas”. We therefore hypothesized that in this subcategory of patients the distribution of artifacts can also be influenced by the underlying inflammatory process. Due to the scarcity of the comparison population, we can consider these results as preliminary pending the execution of a study on a larger sample of patients, in order to understand if there is a correlation between the ultrasound results, their distribution and BNP levels and the type of cardiac and / or inflammatory disease (Tables 5 and 6).
Recent studies and systematic reviews performed in adult patients have shown that LUS can identify acute heart failure and pulmonary edema with high sensitivity and specificity, demonstrating greater sensitivity and specificity than chest x-ray.11,12,13
With our study we could not compare the ultrasound results with the X-ray results because not all acutely ill children performed chest X-rays. It has not been performed for example in children with a diagnosis of heart disease already known or for those children for whom it was not necessary to perform it by virtue of the diagnostic suspicion already clear from the initial symptoms as in some cases of systemic inflammatory disease/sepsis. Others, however, in particular children with a diagnosis of not known heart disease in the acute phase (such as acute heart failure in the course of unknown heart disease or acute rheumatic carditis not known with atrioventricular overload) who presented in the emergency room with cough and dyspnea or those presenting with persistent fever (with subsequent diagnosis of systemic inflammatory disease/sepsis) underwent chest x-rays to rule out infectious respiratory disease.
As has happened in our clinical practice, -in some cases (as in the case of mild and early pulmonary edema), the x-ray can be completely negative; - in other cases (such as in the presence of pleural effusion with consensual atelectasis), chest x-ray may be a confounding factor in suspecting an quadro of pleuropneumonia; - and in still other cases (in children who present a febrile clinical picture in the context of a systemic inflammatory disease / sepsis) chest X-ray can once again erroneously suspect a suspected viral infection and / or bacterial of the lower airways as responsible for fever peaks.3,4,12
All this obviously has implications in clinical practice: using only the chest X-ray in the initial phase of the diagnostic process, there is a risk of making an incorrect diagnosis with the risk of delaying the correct diagnosis on one side and on the other side of administering unnecessary therapies by making inappropriate use especially of antibiotic therapy.
Therefore, in addition to the now well-known advantages of LUS in the pediatric population (non-invasive, non-ionizing radiation tool characterized by a rapid, affordable, point-of-care imaging modality that allow both real-time diagnosis and follow-up of respiratory diseases31-34), integrating LUS in the earliest phase of the diagnostic process of acutely ill patients also in the pediatric population allows for a greater possibility of earlier diagnosis and appropriate diagnosis in light of the result of the specific lung ultrasound pattern such as our results prove.
Our study has some limitations. First of all, due to the COVID-19 pandemic we have been able to enroll and evaluate a small number of acutely ill patients.
Secondly, having a small patient population, despite being a heterogeneous patient population by age, we were unable to stratify patients according to age and evaluate differences with respect to LUS findings and their distribution.
Finally, due to the scarcity of the comparison population, we could not adequately analyze the presence or absence of correlations between the ultrasound parameters analyzed, their distribution and the type of pathology (cardiac disease, systemic inflammatory disease, cardiac diasease and systemic inflammatory disease) and the values of pro-BNP, thus obtaining preliminary results regarding this aspect.
In conclusions, the LUS pattern we described in our “acutely ill children” population is in line with recent advances in SIS analysis in preclinical models and adult patients with respiratory disease and/or with cardiopulmonary interactions and reinforces the need to focus more efforts on studies of LUS artifacts also on the pediatric population.
Dedicated software, algorithms and transducers, and artificial intelligence are all needed to explore the information contained in US signals.35 However, pending the development and testing of new software that can obtain objective information (quantitative and qualitative), our results confirm that the knowledge of the different origin and the histological characteristics of the B lines allows us to perform a qualitative and semi-quantitative analysis of the SIS. In fact now, point-of-care physicians and sonographers can also apply our results in the first-line work-up in the diagnostic approach to this category of patients.
To date, it remains fundamental to have the clinical and laboratory background of each patient as a guideline in the interpretation of SIS even in acutely ill children . Indeed, new studies, including preclinical ones along with the use of artificial intelligence, are needed to validate our results.