Respiratory Syncityal Virus A and B: three bronchiolitis seasons in a third level hospital in Italy

Respiratory syncytial virus (RSV) is the main cause of respiratory infections and of hospitalization for bronchiolitis among infants < 12 months [1].

Globally, RSV causes nearly 34 million lower respiratory tract infections (LRTI) and 3.4 million hospitalizations per year in infants and children < 5 years of age, with an estimated annual increase of 10% [2, 3]. In many countries, including Italy, RSV currently represents a public health issue [4].

RSV was formerly considered a subfamily of Paramixoviridae but was recently reclassified as belonging to the Pneumoviridae family [5]; the virus is characterized by a large envelope and negative-sense RNA, coding for 11 glycoproteins. Two main proteins of the lipidic envelope, the fusion protein (F) and the G protein, are antigenically significant, thus inducing neutralizing antibody response. The F protein is highly conserved, while the carboxy-terminal portion of hypervariable region of G protein is subject to rapid mutations. Considering the different in vitro reaction to monoclonal antibodies due to the virion external G protein, RSV has been classified into two subtypes, A and B [6, 7]. These two major serotypes can simultaneously circulate during epidemic season, but, usually, one prevails over the other. Recently, RSV genome sequencing has enabled the identification of multiple genotypes and to date, the analysis of nucleotide sequences of G protein led to describe 11 RSV-A and 23 RSV-B genotypes [8].

Many studies have described different RSV genotypes and their clinical relevance, also in Italy [9, 10]. Monitoring genotypic characterization of the RSV virus could help in understanding the pathogenesis of the infection, in predicting the clinical severity of the disease, and in selecting candidate strains for development of therapeutic and vaccine strategies. It is recognized by many studies that the RSV infection has a more severe course in premature and immunocompromised children [11‐13]. It is clear that both the characteristics of the host (age, prematurity, comorbidity, immunodeficiency) and of the virus (including the subtype) influence the clinical course of infection, however the actual contribution of each factor is still unclear. To date, the clinical impact of viral factors related to RSV is still controversial. While many studies report no differences between the two serotypes [8, 9, 14‐16], many others relate the worst prognosis to RSV-A [10, 17‐20] and some others to RSV-B [21‐23]. In 2014 Cangiano et al. [24] highlighted the major role played by RSV in severe bronchiolitis and different rates of RSV related hospitalization in different epidemic season. A review conducted by Kuhdari et al. [4] confirmed the high impact of RSV infection on hospitalization in Italy, especially in children aged less than one year. Recently Midulla et al. [10] described the important influence of specific RSV genotypes on the clinical course of bronchiolitis in a previously healthy and term population. The authors reported deep differences in the demographic characteristics and in the clinical presentation, but not in the necessity and the duration of oxygen (O2) therapy.

It is well know that infants born at < 37 week gestational age (wGA) are at highest risk for severe RSV infection [25]. Prophylaxis with palivizumab (a monoclonal antibody anti-F protein) reduces RSV LRTI in late-moderate preterm population; however, infants are eligible for prophylaxis only according to specific national guidelines. According to the recommendation by the American Academy of Pediatrics, in 2016 the Italian Drug Agency (AIFA) limited the prescription of palivizumab to the group of < 29 wGA and age < 12 months at the beginning of the RSV season [26, 27]. Since then, several evidences in Italy reported an increase in vulnerability of all preterm infants [28‐31]. Therefore, in 2017 the AIFA authorized again the use of palivizumab for preterm infants of ≤35wGA aged less than six months at the start of epidemic season [32].

However, the impact of RSV A and B in terms of incidence and severity, both in full term and preterm population, has not yet been investigated.

The main aim of this study is to identify the trend of RSV-A and B in the last three epidemic bronchiolitis seasons in term born and preterm infants; secondly, investigate whether one of them was associated with a worse prognosis, in both populations. Finally we compared three bronchiolitis seasons in terms of length of stay, use of oxygen therapy and need for intensive care.

We performed a retrospective study involving patients admitted for RSV bronchiolitis to the Pediatric Academic Department of Bambino Gesù Children’s Hospital in Rome. We reviewed medical records of all patients hospitalized between November 2015 and April 2018 with a clinical diagnosis of bronchiolitis and nasopharyngeal aspirates positive for RSV. According to the Anglo-Saxon definition of bronchiolitis as the first episode of wheezing in the first twelve months of life, only children with less than one year of age were enrolled in the study. We excluded patients with incomplete medical record. We included in the study preterm infants as born < 37 wGA. Data processing was performed with the Microsoft Excel 2018 software. We obtained demographic characteristics of patients (including gestational age, sex and age), length of stay (LOS), medical history during recovery and necessity of O2 therapy administered in pediatric ward as low flow or high flow nasal cannulas. Finally, we evaluated admissions and LOS in pediatric intensive care unit (PICU) due to non-invasive Continuous Positive Airway Pressure (C-PAP) or invasive mechanic ventilation. In this study we selected the presence of hypoxemia as important clinical outcome, with the oxygen saturation detected at the pulso oximeter ≤92% in ambient air and need for O2 supplementation, the duration of O2 therapy and the PICU admission.

The identification of respiratory viruses on nasopharyngeal aspirates was accomplished by Multiplex RT-PCR. Samples were processed, immediately or after storage at − 80 °C, by AllplexTM Respiratory Panel Assays (Seegene, Korea). Nucleic acids were extracted using the STARMag Universal Cartridge kit (Seegene, Korea) on automated Nimbus IV platform. As recommended by the manufacturer, a total of 200 μl per sample were extracted and eluted in 100 μl of elution buffer. For each reaction, 7 μl of DNA/RNA in a final volume of 25 μl were used. Real time PCR was performed on CFX96 (Bio Rad Laboratories, Italy) with respiratory panel assay kit. The panel is made up of 3 mixes which allow the identification of 16 different viruses (Influenza A and B virus, Respiratory syncytial virus A and B, Adenovirus, Enterovirus, Parainfluenza virus 1, 2, 3 and 4, Metapneumovirus, Bocavirus, Rhinovirus, 3 Coronaviruses (NL63, 229E and OC43) and 3 Influenza A subtypes. An internal control was included in each sample to check both extraction efficiency and PCR inhibition. In every run, a negative control was used to monitor carry-over contamination. The results were analyzed automatically using Seegene software (Seegene Viewer V2.0).

Formal consent is not required for this kind of retrospective study: any personal data was protected accordingly to the Helsinki Declaration and to Italian law. (Legislative Decree of 30 June 2003, n. 196. Code on the protection of personal data).

In the present study we examined epidemiology and clinical characteristics of acute RSV bronchiolitis in term and preterm infants, hospitalized in the Pediatric Academic Department of Bambino Gesù Children’s Hospital in Rome, Italy. We studied 422 patients, of whom 50 were premature with wGA < 37.

The median age at recovery was 2 months with no significant difference between term born and preterm population, while LOS was longer in the group of premature infants.

During the whole reference period (3 years), 56,6% of all patients needed O2 therapy. Furthermore, 8,3% of children were transferred to intensive care unit necessitating C-PAP or ventilatory assistance with orotracheal intubation. We observed an increase in bronchiolitis severity over the three years: a significant increase in the percentage of patients requiring O2 supplementation (p < 0.0001) and intensive care cure (p: < 0.0001) was reported.

The description of a raising trend in RSV bronchiolitis severity is supported by other recent Italian studies [4, 28, 29], and could be related to the lack of immunity due to the selection of new RSV strains.

Furthermore, in our work, we described a conspicuous number of preterm children admitted to the pediatric department, and in particular of infants included in 29–36 wGA.

Comparing clinical course of bronchiolitis in full term population and in preterm, we found similar percentages of cases requiring O2 therapy. However, duration of oxygen therapy, need for PICU cure and medium of LOS in PICU were more relevant in preterm infants. We found a strong correlation between wGA and duration of oxygen therapy (Pearson r: -0,62). This result confirms the literature data that highlights aggressiveness of RSV in premature infants. [2, 28, 30, 33]. In 2017, a large retrospective work on prematures [13] identified an oxygen supplementation rate of around 80% and a PICU admission rate of 22%, while in our work emerged a lower rate both for O2 supplementation (62%) and PICU admission (16,1%), but a raising trend for both parameters.

During the years 2015–2018 RSV-A was prevalent among all infants enrolled in this study. Although this result is in line with the previous literature [8‐10] we studied three consecutive seasons, of which the intermediate was characterized by RSV-B prevalence, and this may be a confounding factor.

Considering the epidemiological trend of the virus, we observed interesting differences between the two age groups analyzed. A seasonal trend of RSV subtypes, which alternate from year to year, occurred in the term born population. In fact, we documented a clear prevalence of RSV-A in the first winter, of RSV-B in the second and again of RSV-A in the third. Conversely, in preterm infants, RSV-A was predominant in all three epidemic seasons. Comparing the severity of RSV bronchiolitis, our data show even more relevant differences between the two groups.

In full term infants, the prevalent serotype among the season was also responsible for most cases requiring oxygen supplementation and intensive care. In preterm infants instead, over the three investigated seasons, the rates of infection by RSV-A were higher compared with those caused by RSV-B, independently from the prevalent circulating serotype. Remarkably, the incidence of desaturation and the use of O2 therapy during RSV-A bronchiolitis was always prevalent compared to RSV-B bronchiolitis. According to this result, in premature infants RSV-A was related to an higher risk of O2 therapy compared to RSV-B (OR 2.39).

Our work, in line with current literature [17‐19], describes seasonal alternation of RSV serotypes A and B. This would explain RSV-A prevalence in our sample, since we have analyzed three sequential seasons. In our case series, there is no evidence of a serotype being more aggressive than the other, but we observed a relative increase of severe bronchiolitis for RSV-A, especially for premature children, as already considered by Esposito et al. [9]. Recently, an Italian study [10] described a worse clinical outcome in RSV-A related bronchiolitis, arguing that low weight and younger age at admission were more often associated with novel RSV-A genotype infection. In comparison with their results, we did not find differences in percentages of O2 therapy between RSV A and B, as the prevalent serotype showed a major virulence in general population. However, we found an increased relative risk of O2 therapy in RSV-A infected premature patients. Many studies have investigated the clinical relevance of environmental factors in severe clinical course of RSV infection [12, 34, 35]. In our study, however, the higher incidence of serotype A in preterm population, together with the resultant severe clinical course, suggests a higher virulence of RSV-A per se. This could suggest an important involvement of the immune response to the virus and a role for different immune mechanisms in response to the virus in preterm infants. Indeed, there are many evidences about lack of immunity in preterm children, regarding count and activity of all subsets of T cells [36] and deficient antibody production in younger infants infected with RSV [37]. Since various components of the innate and adaptive immune response are crucial for controlling RSV infection and disease [38], more studies are needed to focus preterm specific immune response. Moreover, some evidences in vitro reported different pathways of inflammatory activation by RSV-A and B [39].

The development of safe and effective vaccines might be greatly facilitated by the knowledge of natural immune response of infants to RSV infection [40]. Despite many RSV vaccine strategies and candidates [41], currently there is not any licensed for human use.

This study presents several limitations. First, it is a retrospective study, monocentric and mono departmental. In our hospital, indeed, many premature infants are hospitalized in the neonatal sub-intensive unit, and this could be an important bias regarding the effective numbers of RSV bronchiolitis in the preterm population. In addition, we were not able to perform viral load and viral genotype analysis. Since evidence about the role of co-infections in literature is not clear [42‐45], we decided to include 35% of patients co-infected by other respiratory virus. Moreover, the presence of a co-infection in our sample was not associated with an higher risk of O2 therapy or PICU admission.

Our analysis evidences a relevant increase in severity of bronchiolitis RSV-related in a third-level hospital and a significant higher clinical impact of bronchiolitis in premature infants affected by RSV-A. Considering the absence of appropriate therapeutic strategies and the evidence of increased vulnerability in premature population, our work emphasizes the need of urgent implementation in prophylaxis measures against RSV infection and, as recently reported by an Italian group [46], the importance of maintaining immunoprophylaxis also in infants ≤35 wGA.

We describe a seasonal trend of two different viral serotypes that alternate from year to year. Although there are no significant differences in severity between RSV-A and B in full term population, RSV-A is prevalent and causes most of the severe cases in premature infants.

Furthermore, an increase in numbers of severe bronchiolitis RSV-related is evident not only in the premature population.

These results could be due to the lack of immunity against the viral populations that alternate seasonally, but also to the selection of more aggressive viral strains. The temporary restriction of prophylaxis with palivizumab could be one of the causes of the increase in severity in premature infants.

In this scenario it is fundamental to continue active surveillance against the RSV.

Further correlation studies between viral type, genotype and clinical severity of bronchiolitis are needed. In our opinion, a better understanding of perinatal modification of the preterm infant immune system could be a starting point to explore the topic. The increase of current knowledge about RSV specific immune response could define the basis for achieving a safe and effective vaccine in order to decrease the RSV disease burden.