Respiratory syncytial virus (RSV) is the most important cause of lower respiratory tract infection in infants and children. Fifty to 90 percent of hospitalizations for bronchiolitis, 5 to 40 percent of those for pneumonia, and 10 to 30 percent of those for tracheobronchitis are caused by RSV [1]. Substantial increases in the number of admissions for RSV bronchiolitis (up to 126.300 hospitalizations and 500 deaths annually) have been documented in North America [2]. In Canada, inpatient care of RSV illness costs $ 18 million (US dollars) yearly, accounting for 62 percent of the total cost of this disease [1]. The magnitude of the costs is understandable, since virtually all children become infected with RSV within two years after birth, and one percent requires hospitalization [3].

In developed countries, there are well-defined high-risk groups, generally with chronic underlying disorders, in whom infection with RSV is more likely to progress into severe lower-respiratory-tract infections. Healthy infants younger than 3 months old are also susceptible to such infections. High risk groups for severe RSV disease include infants below six months of age, premature infants with or without chronic lung disease (bronchopulmonary dysplasia – BPD), infants with hemodynamically significant congenital heart disease (CHD), immunosuppressed patients (including those undergoing chemotherapy, bone-marrow and solid-organ transplantation, and those with underlying disorders of cellular immunity) or cystic fibrosis, and infants with neuromuscular diseases [4-6]. There is no particular age-group that is not at risk for RSV infection, but certain risk factors have been implicated in more severe disease:low socioeconomic status, crowded living conditions, exposure to indoor smoke pollution, a family history of asthma or atopy, and perhaps infection with the A subgroup of RSV [4]. Preterm infants with or without BPD and infants with CHD are also known to be at increased risk for hospitalization including admission to an intensive care unit (ICU), ventilatory support, or prolonged supplemental oxygen [7-9].

This review focuses on the burden of RSV disease in preterm infants with and without BPD.

The year-to-year and seasonal variations in RSV activity are key aspects of RSV epidemiology. RSV is a seasonal virus, with peak rates of infection occurring annually in the cold season in temperate climates, and in the rainy season, as temperatures fall, in tropical climates [4]. Among Europe RSV related re-hospitalizations of preterm infants show a seasonal distribution mainly during the winter and spring months ranging from October to May, and peaking between December/January and March [9-14]. The distribution of hospitalizations and ICU admissions related to confirmed RSV infection throughout the study period (IRIS study) peaked in December and January [14] and the distribution of RSV related re-hospitalizations in the Munich RSV Study extended from October to May, with 29 cases (75.7%) occurring between January and March [10]. The Austrian RSV 29-32 Study Group including 863 infants aged 29 to 32 weeks of gestational age revealed 18% of RSV related re-hospitalizations occurring outside the typical RSV season [11]. An even higher rate has been reported by Duppenthaler et al., [15] during the 1999/2000 RSV season in Switzerland with 36% of hospitalisations occurring after April 1, 2000 that would not have been preventable by palivizumab prophylaxis initiated in November 1. Overall there are major differences between regions throughout the northern hemisphere [16-22]. In the southern hemisphere, for example in Gambia, RSV activity peaked during the summer months between August and September over a 4-year (1993 – 1996) period [23].

Thus, knowledge about local RSV epidemics is mandatory for targeted prophylaxis with palivizumab in high-risk infants. In Austria an epidemiological monitoring system has been established in 2002 called RSV-hotline incorporating data of infants hospitalized due to RSV disease that are entered into the system voluntarily by Austrian pediatricians (https://hc4you.hcsolutions.at/rsv). The seasonal distributions of RSV attributed hospitalizations of preterm infants in Austria over the seasons 1998 to 2001 and 2001 to 2003 are shown in Fig. (1a, b), respectively [11, 24].

Fig. (1). (a). Seasonal distribution of rehospitalisations due to respiratory illness (proven RSV and non-RSV infections) in premature infants of 29–36 weeks gestational age [24]. (b). Seasonal distribution of RSV hospitalizations (n=38) in premature infants of 29–32 weeks’ gestational age between 1 June 2001 and 1 June 2003 [11].

The most common infection caused by RSV is of the upper respiratory tract; such infections are characterised by rhinitis, cough, and sometimes fever. Acute otitis media occurs in up to a third of children with RSV illness; both RSV and bacterial pathogens have been isolated from the middle ears of children with RSV. Croup also occurs with RSV infection, but bronchiolitis and pneumonia are the most common manifestations in children. Signs of upper-respiratory-tract involvement commonly precede those of the lower respiratory tract by a few days, and fever, when present, is usually low grade. Dyspnoea, lower chest-wall indrawing, and difficulty in feeding characterise lower-respiratory-tract infection. In bronchiolitis, wheeze may be audible with or without a stethoscope, and a prolonged expiratory phase and crackles are characteristic. Air trapping results in very fast breathing and a palpable liver and spleen. The typical radiographic pattern includes hyperinflation with diffuse interstitial markings and peribronchial thickening. Segmental atelectasis, which usually clears spontaneously, is often seen. Children with pneumonia, on the other hand, have fine crackles and a radiographic pattern of alveolar, segmental, or lobar consolidation.

Severe bronchiolitis may lead to acute respiratory failure associated with severe bronchospasm, moderate to severe hypoxia, and carbon dioxide retention. With lung function tests [25] two patterns of severe disease are seen: in about two-thirds of cases there is obstructive small airways disease (bronchiolitis), and in the remainder there is a restrictive pattern (pneumonia). Most of the latter cases meet the criteria for acute respiratory distress syndrome. They tend to be younger, have more predisposing underlying disease, and are ventilated for longer. In severely ill children, complications include pulmonary hypertension and cardiovascular compromise requiring inotropic support [26].

Although bacterial superinfection is rare in developed countries, it is more common in developing countries. This may partly explain the higher fatality rates seen in developing nations [4]. Over a five years period we observed a general low total rate of bacterial co-infection of 1.9% excluding ICU patients [27]. The risk of concurrent bacterial infection in preterm infants hospitalized due to respiratory syncytial virus infection was three times higher compared to term infants (9.5 vs. 3.1%, p=0.017) and associated with prolonged hospitalisation and ICU admission). The predominant pathogens were Streptococcus pneumoniae and Haemophilus influenzae. Mean length of stay in preterm infants with bacterial co-infection was 22.3 days compared to 10.3 days without bacterial co-infection (p < 0.006).

Apnoea tends to occur in infants under 2 months of age with atelectasia on chest radiography [28] and is common in those born prematurely. In cases of severe apnoea, mechanical ventilation may be necessary, despite the absence of respiratory failure. The pathophysiology of this manifestation is unknown, but postulated mechanisms include the immaturity of the respiratory centre in the brain-stems of premature infants, and RSV associated hypersensitivity of the laryngeal chemoreceptors [4]. In a recent systematic review the incidence of apnoea including a study population of 5575 hospitalized patients with RSV ranged from a high of 23.8% to a low of 1.2% [29]. Comparison of the cohort by term and preterm birth revealed an expected excess in apnoea incidence in preterms, even in the absence of information about chronologic age In detail, the authors cite five of seven studies reporting on the incidence of apnoea with significant rates comparing term to preterm infants (range 0.5 – 12.4% compared to 4.9 – 37.5%, respectively, overall p<0.001), see Table 1 [29]. A study by Gleeson et al., [30] examined the hypothesis that dysregulation of mucosal immune responses to respiratory infections is a critical event, which could be causal in respiratory arrest of some previously healthy infants. The salivary IgA and IgM concentrations in the apparent life threatening event (ALTE) infants at presentation to hospital indicated that a significant mucosal immune response had already occurred with nearly 60% of the IgA concentrations significantly above the population-based reference ranges. This hyper-immune response was most evident in the ALTE infants with pathology evidence of an infection; and the most prevalent pathogen identified was RSV (64%). The group compromised 26 infants with a gestational age ranging from 26 to 41 weeks at a median age of 60 days.

Encephalopathy associated with respiratory syncytial virus infection is not well recognized. Ng et al., [31] reported on an incidence of 1.8% in a total of 487 patients with respiratory syncytial virus bronchiolitis studied over a period of four years with seizures being the presenting complication. Interestingly, three of the nine infants showing neurologic complications had a history of prematurity (27, 28, and 31 weeks of gestational age, respectively).

During a five-year period (1984 to 1989) Meert et al., [7] reported on 484 previously healthy infants (including 27% preterm born) admitted to the hospital with RSV infection. No differences were found in the presenting symptoms of respiratory distress, cough, fever or shock, although the preterm group was more likely to present with apnea (13 vs. 4%, p<0.001). Chest roentgenograms revealed that preterm infants had a higher incidence of atelectasis/infiltrate and hyperinflation (47 vs. 36%, p<0.05, and 28 vs. 19%, p<0.05, respectively). Preterm infants had longer hospital stays (mean 5.4 vs. 3.6 days, p<0.001) as well as a higher Physiologic Stability Index and Therapeutic Intervention Score (6.4 vs. 4.3, p<0.001, and 7.4 vs. 5.4, p<0.001, respectively). They were also more likely to receive supplemental oxygen (39 vs. 19%), ICU admission (13 vs. 4%), mechanical ventilation (12 vs. 2%), and nothing by mouth status (17 vs. 7 %, for all p<0.001, respectively). There was no difference in postnatal age (mean 5.5 vs. 5.1 months), but preterm infants had a younger postconceptual age (mean 3.9 vs. 5.0 months, p<0.05).

The 37 preterm infants with RSV related rehospitalisation of the Munich RSV study [10] presented with pneumonia with and without clinical signs of obstructive bronchitis (27%); obstructive bronchitis, bronchiolitis, or acute bronchitis (67.6%) or acute upper airway infection (5.4%).

In the early 1960s, vaccination of infants with a formalin-inactivated RSV vaccine failed to be successful despite significant increases of compliment fixatory antibodies in the vaccinated infants [58,59]. The vaccine augmented the severity of wild virus infection and led to the death of two infants in the following season. Standard polyclonal immunoglobulin preparations containing substantial levels of neutralizing antibodies to RSV were not successful in the prevention of RSV hospitalisation. In contrast, the hyperimmmune RSV globulin (RespiGam^®) given intravenously at dosages of 750 mg/kg monthly over the RSV season proved to be save and effective in prevention of lower respiratory tract infections in high-risk preterm infants and young children [60]. Children with cyanotic congenital heart disease however were more likely to experience adverse events, such as sudden unexplained death associated with cardiac surgery, if they received RSV-IGIV. Thus, recommendations for the use of RSV-IGIV published by the American Academy of Pediatrics excluded those infants with CHD [61]. Despite its efficacy, RSV-IGIV had the disadvantage of monthly infusions that were cumbersome, expensive, time-consuming, and stressful for both the patient and the parents, and there was always the remote possibility of blood-borne pathogen transmission. The development of a highly potent RSV-neutralizing monoclonal antibody (Mab) was the next approach to improve the specific activity of anti-RSV immunoglobulin. In 1997, researchers at MedImmune, Inc. developed a humanized murine monoclonal antibody called palivizumab (Synagis^®) that recognizes a conserved neutralizing epitope at the “A” region on the F glycoprotein of RSV [62].

In a randomized, placebo controlled (palivizumab:placebo 2:1), double-blind, multicenter phase III trial including 139 centers in the United States, Canada, and the United Kingdom 1502 young children with prematurity (≤ 35 weeks of gestation, ≤ 6 months of age) or bronchopulmonary dysplasia (≤ 24 months of age, requiring medical treatment within the past 6 months for their chronic lung disease) were included during the 1996 to 1997 RSV season to receive 5 monthly (every 30 days) injections of either 15 mg/kg palivizumab or an equivalent volume of placebo intramuscularly [45]. Children were followed 150 days for the primary endpoint hospitalization due to RSV infection. Ninety-nine percent of the children in both groups completed the study protocol and 93% received all five scheduled injections. Palivizumab prophylaxis resulted in a 55% reduction in hospitalization as a result of RSV infection. Details of study results including secondary outcome measures are presented in Table 4 [45].

The incidence of RSV related hospitalization following palivizumab prophylaxis was assessed in several retrospective and prospective studies conducted in Europe and North America [63]. In general, the incidence of RSV related hospital admissions after palivizumab prophylaxis in post-marketing surveillance studies was lower than the reported rates of the IMpact study [46]. Data from the American Palivizumab Outcomes Registry provide the largest data set available on infants born prematurely with or without BPD and children with CHD (55-58) and are summarized in Table 5 [64].

Data from the French Pediatricians' Group of Synagis Patients' Name-Based Programs reported on 7.6% re-hospitalisation rate in a population of 515 infants with a high rate of 81% of the infants included having BPD [65]. In the large Spanish trial of the IRIS Study Group [48] including two study cohorts of a total of 3502 infants over four RSV seasons (1998 to 2002) an overall reduction of 70% of RSV related hospital admissions in premature infants with and without BPD was demonstrated. Infants without prophylaxis had a 4-fold increased risk of RSV hospitalization. No deaths were reported in palivizumab recipients.

Adherence to the monthly injections scheme of palivizumab prophylaxis remains a major problem. Parnes et al., [66] noted lower re-hospitalisation rates in infants with higher adherence to this scheme, and they observed that approximately half of RSV related hospitalizations occurred between the first and second injection. Similar findings were also reported by Manzoni et al., [67], consistent with the trends of serum palivizumab levels that may still be inadequate and not fully prototective after the first monthly dose [68]. Missed or delayed palivizumab injections resulted in an increase of RSV related hospitalizations from 2.4 % to 4.4 % (p=0.02). In comparison of a home versus office setting during the 2000/2001 season 969 infants received palivizumab injections at home by a nurse and 477 in the office of a pediatrician [69]. Comparing home versus office, compliance was 98% compared with 89% (p<0.001), and RSV related hospitalization rate was 0.93% compared with 3.57% (p<0.001). During a multicenter observational study of respiratory syncytial virus-associated hospitalizations and use of palivizumab in premature infants aged 29-32 weeks over two RSV seasons (2001 to 2003) in Austria 238 infants received palivizumab prophylaxis [11]. The rate of inadequate or incomplete injections was very high (62%) resulting in an increased rehospitalisation rate of 8.1% compared with 3.3% of adequate prophylaxis (12/148 vs. 3/90, p=0.07).

A vaccine for RSV is needed, and a protective live attenuated vaccine administered at or around birth would be ideal. However, problems like insufficient attenuation of the vaccine strain of RSV, its thermolability, and the need to include A and B strains in each vaccine have hindered trials in infants under 3 months of age [4]. Prevention of RSV related hospitalizations in preterm infants by monthly injections of monoclonal antibodies like palivizumab or in the near future motavizumab is recommended in selected high-risk groups [70] but still discussed due to the high costs of the product. Recent studies focused on RSV risk factor models aiming to tailor prophylactic treatment with monoclonal antibodies to those preterm infants with the highest risk for severe RSV related disease [71-74]. Until a vaccine is developed efforts to offer prophylaxis to high-risk infants is mandatory.