Initiative for Vaccine Research (IVR)

Acute Respiratory Infections (Update September 2009)

Respiratory syncytial virus and parainfluenza viruses


Respiratory syncytial virus (RSV) and parainfluenza viruses (PIV) belong to the same family Paramyxoviridae. Their importance as respiratory pathogens in young children has been recognized for over 40 years, yet the development of vaccines against RSV or PIV has been hampered by several factors, including the risk of potentiation of naturally occurring disease, as was observed in the early 1960s (see below).

Respiratory syncytial virus (RSV) and parainfluenza viruses (PIV) belong to the same family Paramyxoviridae. Their importance as respiratory pathogens in young children has been recognized for over 40 years, yet the development of vaccines against RSV or PIV has been hampered by several factors, including the risk of potentiation of naturally occurring disease, as was observed in the early 1960s (see below). RSV is the most important cause of viral lower respiratory tract illness (LRI) in infants and children worldwide [16] , being responsible for 70 000 to 126 000 infant hospitalizations for pneumonia or bronchiolitis every year in the USA alone. The elderly also are at risk for severe RSV disease [19] [188] and 14 000 to 62 000 RSV-associated hospitalizations of the elderly occur annually in the USA [189].

Human parainfluenza viruses types 1, 2 and 3 (PIV1, PIV2 and PIV3, respectively) are second only to RSV as important causes of viral LRI in young children [10] , being recovered from 18% young children [190] , with upper respiratory illness (URI), 22% with LRI and 64% with croup [191] . PIV-1 and PIV-2 are the principal causes of croup, which occurs mostly in children 6 to 48 months of age, whereas PIV-3 causes bronchiolitis and pneumonia predominantly in children less than 12 months of age.

Disease Burden

RSV infection

Human RSV infection, the single most important cause of severe respiratory illness in infants and young children and the major cause of infantile bronchiolitis, is the most frequent cause of hospitalization of infants and young children in industrialized countries [192] . In the USA alone, from 85 000 to 144 000 infants with RSV infection are hospitalized annually [17] , resulting in 20%-25% of pneumonia cases and up to 70% of bronchiolitis cases in the hospital [193] [194] . Global RSV disease burden is estimated at 64 million cases and 160 000 deaths every year.

RSV disease spectrum actually includes a wide array of symptoms, from rhinitis and otitis media to pneumonia and bronchiolitis. Humans are the only known reservoir for RSV. Spread of the virus from contaminated nasal secretions occurs via large respiratory droplets, and close contact with an infected individual or contaminated surface is required for transmission. The virus can persist for several hours on toys or other objects, which explains the high rate of nosocomial RSV infections, particularly in paediatric wards. In the USA, nearly all children by 24 months of age have been infected at least once with RSV, and about half have experienced two infections [195].

Children who experience RSV infection early in life run a high risk of subsequent recurrent wheezing and asthma [196] [197] [198] , especially premature infants and infants with bronchopulmonary dysplasia, for whom preventive passive immunization with anti-RSV monoclonal antibodies such as Palivizumab is highly recommended [199] [200] [201] . RSV-infected infant sera and nasal secretions show a marked increase in levels of Th-2 cytokines and chemokines, including IL-4 and MIP-1?, as well as of non-neutralizing IgE antibodies. The RSV G protein is believed to induce the release of large amounts of Th-2 cytokines and MIP-1? from CD4+ T cells and mast cells, basophils and monocytes, that trigger increased pulmonary eosinophilia and asthma exacerbation [202] . Repeated RSV infection in a mouse model similarly induces persistent airway inflammation and hyperresponsiveness which are characteristics of asthma [203] [204] [205].

Infants who had been immunized with a formalin-inactivated RSV vaccine in the 1960s similarly experienced enhanced RSV disease and pulmonary eosinophilia upon subsequent RSV infection, leading to numerous hospitalizations and two deaths [206] [207] [208] , probably due to skewing of the immune response towards a Th-2 response and failure of the vaccine to induce a CD8+ T cell response [209] [210].

RSV also is a significant problem in the elderly [188] , in persons with cardiopulmonary diseases [211] and in immunocompromized individuals [212] . RSV attack rates in nursing homes in the USA are approximately 5%-10% per year with a 2%-8% case fatality rate, amounting to approximately 10 000 deaths per year among persons >64 years of age [213] ]. Among elderly persons followed for 3 consecutive winters, RSV infection accounted for 10.6% of hospitalizations for pneumonia, 11.4% of hospitalizations for obstructive pulmonary disease, 5.4% for congestive heart failure and 7.2% for asthma [214].

Few population-based estimates of the incidence of RSV disease in developing countries are available, although existing data clearly indicate that the virus accounts for a high proportion of ARIs in children. Studies in Brazil, Colombia and Thailand suggest that RSV causes 20-30% of ARI cases in children from 1-4 years of age, a proportion similar to that in industrialized countries. Another confusing aspect of the epidemiology of RSV infection is the seasonality of the disease. In Europe and North America, RSV disease occurs as well-defined seasonal outbreaks during the winter and spring months. Studies in developing countries with temperate climates, such as Argentina, have shown a similar seasonal pattern. On the other hand, studies in tropical countries often have reported an increase in RSV in the rainy season but this has not been a constant finding. Cultural and behavioral patterns in the community might also affect the acquisition and spread of RSV infection.

Parainfluenza virus infection

Parainfluenza viruses also cause a spectrum of respiratory illnesses, from upper respiratory infections, 30-50% of which are complicated by otitis media, to lower respiratory infections, about 0.3% of which require hospitalization. Most children are infected by human parainfluenza virus type 3 (PIV-3) by the age of two years and by parainfluenza virus types 1 and 2 (PIV-1 and PIV-2) by the age of five years [191] . PIV-3 infections are second only to RSV infections as a viral cause of serious ARI in young children. Pneumonia and bronchiolitis from PIV-3 infection occur primarily in the first 6-12 months of life, as is the case for RSV infection [190] . Croup is the signature clinical manifestation of infection with other parainfluenza viruses, especially PIV-1, and is the chief cause of hospitalization from parainfluenza infections in children two to six years of age [191] . The proportions of hospitalizations associated with PIV infection vary widely in hospital-based studies. Consequently, the annual estimated rates of hospitalization fall within a broad range: PIV-1 is estimated to account for 5,800 to 28,900 annual hospitalizations in the USA, PIV-2 for 1,800 to 15,600 hospitalizations, and PIV-3 for 8,700 to 52 000 hospitalizations. Along with RSV, parainfluenza viruses are also leading causes of hospitalization in eldrly with community-acquired respiratory disease.

PIV-1 causes large, well-defined outbreaks, marked by sharp biennial rises in cases of croup in the autumn of odd-numbered years. Outbreaks of infection with PIV-2, though more erratic, usually follow type 1 outbreaks. Outbreaks of PIV-3 infections occur on a yearly basis, mainly in spring and summer. Although PIV-1, -2 and -3 have been described as a cause of ARI in developing countries, the corresponding disease burden has not been accurately determined in these countries.

Reinfection with any of the parainfluenza viruses and/or with RSV can occur throughout life [215] , usually resulting in mild upper respiratory infections in young adults, but causing severe disease in immunocompromized patients [16] [216] [217].


RSV and parainfluenza viruses belong to the family Paramyxoviridae. These are enveloped viruses with a negative-sense single-stranded RNA genome.

Human RSV, together with its close relative bovine RSV, belongs to the subfamily Pneumovirinae, genus Pneumovirus. Its genome is a 15 222 nucleotide-long, negative-sense RNA molecule which encodes 11 viral proteins, among which the nucleoprotein (N), the fusion protein (F), the surface glycoprotein (G), the matrix protein (M) and several non-structural proteins including the L protein (replicase) and virulence factors NS1 and NS2 that mediate resistance to IFN-?/? [218] . The tight association of the RNA molecule with the viral N protein forms a nucleocapsid wrapped inside the viral envelope, from which protrude viral proteins F, G and SH. The RSV G protein was shown to be a structural and functional mimetic of fractalkine, a proinflammatory CX3C chemokine that mediates leucocyte migration and adhesion [219] , which explains its role in pathogenesis [220] [221] [222].

The fusion protein F and attachment glycoprotein G are the only two components that induce RSV neutralizing antibodies. The sequence of the F protein is highly conserved among RSV isolates. In contrast, that of the G protein is relatively variable [223] : two serogroups of RSV strains have been described, the A and B groups, based on differences in the antigenicity of the G glycoprotein. Current efforts are directed towards the development of a vaccine that will incorporate strains in both groups, or will be directed against the conserved F protein (for a review, see [224]).

Parainfluenza viruses belong to the subfamily Paramyxovirinae, itself subdivided into three genera: Respirovirus (PIV-1, PIV-3, and Sendai virus (SeV)), Rubulavirus (PIV-2, PIV-4 and mumps virus) and Morbillivirus (measles virus, rinderpest virus and canine distemper virus (CDV)). Their genome, a ~15 500 nucleotide-long negative-sense RNA molecule, encodes two envelope glycoproteins, the haemagglutinin-neuraminidase (HN), and the fusion protein (F), itself cleaved into F1 and F2 subunits, a matrix protein (M), a nucleocapsid protein (N) and several nonstructural proteins including the viral replicase (L) [225] . All parainfluenza viruses except PIV-1 express a non-structural V protein that blocks IFN signalling in the infected cell and acts therefore as a virulence factor [226].


General considerations

Development of vaccines to prevent RSV infection have been complicated by the fact that host immune responses appear to play a significant role in the pathogenesis of the disease. Early attempts at vaccinating children in the 1960s with a formalin-inactivated RSV vaccine showed that vaccinated children suffered from more severe disease on subsequent exposure to the virus as compared to unvaccinated controls (see above). These early trials resulted in the hospitalization of 80% of vaccinees and two deaths [206] [207] [208] . The enhanced severity of disease has been reproduced in animal models and is thought to result from inadequate levels of serum-neutralizing antibodies, lack of a cellular immune response, and excessive induction of a Th2 immune response with pulmonary eosinophilia and increased production of IL-4, IL-5 and MIP-1? [209] [210].

In addition, naturally acquired immunity to RSV is neither complete nor durable and recurrent infections occur frequently during the first three years of life. Older children and adults, however, usually are protected against severe RSV disease, suggesting that protection does develop after primary infection.

Passive immunization in the form of RSV-neutralizing immune globulin or humanized monoclonal antibodies given prophylactically has been shown to prevent RSV infection in newborns with underlying cardiopulmonary disease, and in small, premature infants [199] [200] [201] .This demonstrates that humoral antibody plays a major role in protection against infection. In general, secretory IgAs and serum antibodies appear to protect against infection of the upper and lower respiratory tracts, respectively, while T-cell immunity targeted to internal viral proteins appears to help terminate viral infections.

Although live attenuated RSV vaccines seem preferable for immunization of naive infants, nonreplicative vaccines may be useful for immunization of the elderly and older, high-risk children, as well as for maternal immunization. In addition, RSV vaccines should induce a balanced Th1-Th2 response and cover the two antigenically distinct serogroups RSVA and RSVB.

Subunit RSV vaccines

Three types of RSV subunit vaccines have been evaluated in clinical trials [227].

Candidate vaccines based on purified F protein (PFP-1, PFP-2 and PFP-3) prepared from RSV-infected cells were tested in a variety of rodent and nonhuman primate models and found to induce protection against RSV challenge [228] . These candidates were tested in human clinical trials involving elderly volunteers [229] [230] , pregnant women [231] and children with chronic lung disease [232] [233] . The vaccines were found to be safe and moderately immunogenic but the incidence of lower RSV ARI was not significantly diminished in the vaccinees. Vaccination of women in the 30th to 40th week of pregnancy induced RSV anti-F antibodies titres that were persistently fourfold higher in newborns to the vaccinated mothers than to those who had received a placebo and was not followed by increase in frequency or morbidity of respiratory illnesses in the seropositive infants.

Another subunit vaccine consisting of co-purified F, G, and M proteins from RSV A was tested in healthy adult volunteers in the presence of either alum or polyphosphazene (PCPP) as an adjuvant. Neutralizing antibody responses to RSV A and RSV B were detected in 76-93% of the vaccinees, but waned after one year, suggesting that annual immunization with this vaccine would be necessary [234].

Still another subunit approach was investigated using the central domain of the RSV G protein, whose sequence is relatively conserved among serogroup A and B viruses. A vaccine candidate, BBG2Na, was developed by fusing this domain (G2Na) to the albumin-binding region (BB) of streptococcal protein G and producing the fusion protein in a bacterial expression system. The candidate vaccine elicited a protective immune response in animals [235] , and was moderately immunogenic in adult human volunteers [236] . Its clinical development had to be interrupted due to the appearance of unexpected type 3 hypersensitivity side effects (purpura) in a couple of immunized volunteers.

Live attenuated RSV and PIV-3 vaccines

The development of reverse genetic systems for RSV and parainfluenza viruses has provided for the generation of a number of genetically designed vaccine candidates that harbor mutations or deletions in an effort to attenuate virus replication without compromising immunogenicity [237] . Achieving an appropriate balance between attenuation and immunogenicity has however been a major obstacle to the development of these vaccines.

A temperatrure-sensitive (ts) human PIV-3 (HPIV-3) strain, cp45, was selected after 45 passages of the virus in African green monkey cells at low temperature and evaluated as a intranasal vaccine in Phase I/II trials in RSV seropositive and seronegative children and in young infants. The HPIV-3 cp45 vaccine candidate was well tolerated and immunogenic in seronegative infants as young as 1 month of age [238] [239] [240] and showed little risk of transmission to unvaccinated children and toddlers [241] . Rcp45 is a promising HPIV-3 candidate vaccine that is likely to soon be evaluated in efficacy trials [16].

A cold-passaged (cp) derivative of RSV still caused mild respiratory illness in young children: the strain was further attenuated by chemical mutagenesis to produce the cpts 248/404 strain, which was, however, still reactogenic in 1-2 months-old infants [242] and had to be further mutagenized to produce suitably attenuated vaccine candidate strain rAcp248/404/1030-?SH [243] . Its immunogenicity remains to be tested. Another set of engineered candidate vaccines that have a deletion of the NS2 gene in common (rAcp248/404-?NS2) have been found to be overattenuated for children [244] . Meanwhile, a combination RSV and HPIV-3 intranasal vaccine was tested in 6-18 months-old children, using as a vaccine a mixture of the RSV cpts 248/404 and the HPIV-3 cp45 strains: both vaccines were found to be as immunogenic after simultaneous administration as after separate administration [245].

Another live attenuated candidate PIV-3 vaccine was developed using the Kansas strain of bovine PIV-3 (BPIV-3). BPIV-3 is closely related antigenically to HPIV-3, it can protect monkeys against challenge with HPIV-3, and it replicates poorly in humans, making a perfect Jennerian vaccine candidate. BPIV-3 was well tolerated and immunogenic in seronegative children and infants as young as 2 months old [246] [247] but the magnitude of the anti-HN response was lower in children who received the BPIV-3 vaccine than after immunization with the HPIV-3 cp45 strain.

Live chimeric and recombinant vaccines

A chimeric bovine/human PIV-3 (B/HPIV-3) strain was engineered by substituting in a BPIV-3 genome the HPIV-3 F and HN genes and the F/NH intergenic sequences to their bovine equivalent. The resulting B/H chimeric virus retained the attenuated phenotype of BPIV-3 and was highly immunogenic in rhesus monkeys [248].

The B/HPIV-3 chimeric strain was then used as a vector to express the F, or F and G open reading frames of RSV subgroup A or B [249] , thus providing a candidate intranasal vaccine against both RSV and PIV-3 infections [250] . African green monkeys immunized with the B/HPIV-3 chimera expressing either the native or soluble RSV F protein produced RSV-neutralizing antibodies and were fully protected against challenge with wild-type RSV [251] . The live attenuated nasal RSV/PIV-3 candidate vaccine (MEDI 534TM) was shown to be safe and well tolerated in Phase I clinical studies conducted by MedImmune in the USA in adults and seropositive children 1-9 years of age. The vaccine is presently entering Phase I/IIa clinical trials in 2 month-old infants and in 6-24 month-old children. The PIV-3-vectored RSV candidate vaccine could be a vaccine of choice to prevent RSV and PIV-3 infections in young infants.

Other virus vectors have been used to deliver RSV F and/or G proteins. Recombinant vaccinia virus and adenoviruses expressing RSV F, RSV G or RSV F and G were constructed and tested in animal models including chimpanzees but showed mediocre immunogenicity [252] [253] . Sendai virus (SeV), the murine PIV-1, which had been shown to be safe in human volunteers and to protect African green monkeys against human PIV-1 challenge, was also used as a vector to express RSV fusion protein F. The recombinant SeV-RSV F induced RSV-neutralizing antibodies and RSV-specific CTLs and protected cotton rats and mice against challenge with RSV of both A and B subgroups [254] [255] . Similarly, a SeV recombinant expressing the haemagglutinin-neuraminidase (HN) gene from HPIV-3 induced protection against both PIV-1 and PIV-3 challenge [256] . Sendai virus, however, does not seem to be sufficiently attenuated to be used as a Jennerian vaccine in human infants.

Venezuelan equine encephalitis virus (VEEV) replicon particles (VRPs) expressing RSV F or G similarly induced RSV-specific T cell and neutralizing antibody responses and protected mice and cotton rats against RSV challenge [257] [258].