Respiratory syncytial virus (RSV) is the most important viral respiratory pathogen in infancy and early life. Respiratory syncytial virus infection is a substantial cause of infant morbidity and occasional mortality in all geographic and climatic regions of the world. In developed countries, approximately one half of all infants under the age of 1 year will acquire RSV infection during their first winter season, and nearly all infants will be infected by 2 years of age. Approximately 1% of all infants will be hospitalized for RSV disease in the first year of life. In addition, while mortality is much less than 1% in otherwise healthy infants with RSV infection in the presence of good supportive medical care, mortality rates exceed 3% in infants with underlying heart or lung disease even with the best supportive care. Both morbidity and mortality are increased for RSV infections occurring in developing countries.
Once RSV infection has occurred, there is fairly little that practitioners can offer in terms of treatment. Several recent studies1 of oral bronchodilator therapy in infants with bronchiolitis suggest that use of these agents can lessen the work of breathing, reduce respiratory rates, and result in some improvement in oxygenation. However, these changes are comparatively minor and do not result in reduced need for hospitalization of outpatients or in shorter hospital stays for inpatients. Measurements of pulmonary function improve in patients hospitalized with bronchiolitis following institution of corticosteroid therapy, but not at a faster rate than in individuals who receive only other forms of therapy.2 Finally, most studies of the use of ribavirin in bronchiolitis have demonstrated only modest improvements in clinical measures or in oxygénation in ribavirin recipients compared with controls. Only one study on the use of ribavirin in patients on mechanically assisted ventilation has demonstrated meaningful benefits.3 In these patients, the time on ventilators, the time on oxygen, and the duration of hospital stay were reduced. In essence, the vast majority of patients with RSV infection, including most of those requiring hospitalization, cannot expect significant benefit from the use of any of these approaches to therapy.
Because of the appreciable morbidity resulting from RSV infection and because of the lack of agents that are effective in management, the development of an effective vaccine merits high priority.
WHAT PROBLEMS CAN BE EXPECTED IN DEVELOPING AN RSV VACCINE?
Perhaps the major difficulty in developing an RSV vaccine is the feet that natural infection confers only temporary, if any, protection against reinfection. Studies of natural and experimental RSV infection indicate that even high titers of antibody in serum and in respiratory secretions provide only partial protection against infection. These findings raise the question of whether a vaccine actually could provide protection against infection. In response to this, it should be noted that titers of RSV neutralizing antibody in cord blood are directly related to age at time of infection in infancy. That is, individuals born with higher transplacentally acquired serumneutralizing antibody titers have some degree of protection against RSV infection until the natural decline in such antibody titers occurs. More important, recent studies of administering an RSV hyperimmune globulin preparation to children with underlying heart or lung disease have demonstrated that passive immunization with this product reduces the incidence of lower respiratory tract disease, frequency of hospitalization, and need for intensive care unit admissions for RSV infection.4
Finally, following experimental challenge with RSV, infection occurs with equal frequency in individuals with and without high titers of antibody before challenge, but the quantity of virus shed is reduced in those with higher titers.5 Therefore, it seems feasible that the use of a vaccine that produces very high neutralizing antibody responses in recipients might protect them from lower respiratory tract involvement following RSV infection, if not against RSV infection itself.
Another important aspect of developing a vaccine is that it must not increase the severity of illness. There are some suggestions that immunologie mechanisms are responsible for increasing the severity of RSV infection.6'7 An ideal vaccine would stimulate those components of the immune response that provide protection, but not those that increase the severity of disease.
These are more than theoretical concerns. The first RSV vaccine used was prepared by exposing live virus to formalin to inactivate it. Recipients of this vaccine developed very high serum antibody titers against the fusion protein of RSV, but had poor neutralizing responses. Cell-mediated immunity to RSV was markedly enhanced, beyond a level that would occur with natural infection. Vaccinées who then developed natural RSV infection experienced markedly more severe bronchiolitis compared with recipients of a control preparation. In one study, as many as 40% of vaccine recipients required hospitalization for RSV infection, and two deaths occurred.
In summary, it appears that serum antibody is perhaps the major mediator of immunity to natural infection with RSV. Any candidate vaccine therefore should induce high serum-neutralizing antibody titers. Induction of local antibody in the respiratory tract appears, by simple logic, to be a desirable effect of vaccination also, although available data suggest that induction of local antibody is not essential to developing protection. It is not known whether a combination of local and serum antibody would provide optimal protection. Cell-mediated immunity appears to be important in recovery from RSV infection because viral shedding persists for markedly prolonged periods in individuals with defects in cell-mediated immunity. However, these individuals do not appear to have especially severe forms of RSV illness, and there is no evidence to date that cell-mediated immune responses play any role in prevention of infection. Past studies suggest that induction of exaggerated cell-mediated immune responses, and perhaps induction of RSV-specific IgE responses, might be undesirable effects of a vaccine.
WHAT CLUES ARE AVAILABLE FROM THE BIOLOGY OF RSV TO SUGGEST THE IDEAL COMPONENTS OF AN RSV VACCINE?
Respiratory syncytial virus is an enveloped virus with two major types of glycoproteins extending from the surface of the virus. One of these proteins (G protein) is apparently responsible for attachment of RSV to target cells. A second protein (F protein) enables the lipid membrane of the virus to fuse with the lipid membrane of the target cells, allowing the RNA of the virus to be inserted into the target cells. These two are the only RSV proteins that are effective in stimulating neutralizing antibody responses in humans or in various animal models. Therefore, they appear to be the best candidates for RSV subunit vaccine preparations.
Candidate Respiratory Syncytial Virus Vaccines
Respiratory syncytial virus grows well in the range of temperatures maintained in the human respiratory tract. Lung temperatures are usually maintained at 370C and rise with fever. Temperatures in the upper respiratory tract are generally several degrees lower. It is possible to induce mutant strains of RSV that grow only at the lower temperatures of the upper respiratory tract. These temperature-sensitive mutants are potentially good candidates in developing a vaccine because theoretically they would infect only the upper respiratory tract and would not be capable of replication in the lower respiratory tract. It is not known, however, whether infection of the upper respiratory tract alone might be sufficient to stimulate lower respiratory tract disease via reflex mechanisms.
Finally, there are strains of RSV that grow in nonhuman species, but are not entirely species specific. It is possible that these live viruses might replicate well enough to induce immune responses in humans, without having virulence for humans.
WHAT DEGREE OF SUCCESS HAS BEEN ACHIEVED IN RSV VACCINATION IN HUMANS?
The candidate vaccine that has undergone the most extensive testing recently in humans is a purified preparation of the fusion protein of RSV (Table). This vaccine (PFP) was first administered to adults and has been tested in progressively younger age groups for both safety and immunogenicity. In all these age groups, local and systemic reactions immediately following vaccination were minimal, and the vaccine induced modest increases in antibody titer. It should be noted that many of these individuals had high pre-existing titers because of the frequency of RSV infection in childhood and early adult life. More recently, limited studies of safety, immunogenictty, and efficacy of PFP have been carried out in 12- to 36-month-old children.
Study subjects recruited for these trials had been hospitalized previously for RSV infection. Because of this, they were already partially immunized to RSV, but also were somewhat more likely to have lower respiratory illness with subsequent RSV infection. Therefore, it could be determined whether the vaccine could prevent lower respiratory disease as well as RSV infection in general. In children 18 to 36 months of age, PFP vaccination resulted in an average eightfold increase in RSV serum-neutralizing antibody titers. These titers were sustained through a 6-month follow-up period. None of the 13 vaccine recipients became infected, while 7 of 13 saline placebo recipients developed RSV infection.8 Again, there were no significant immediate side effects of vaccination, and no unusually severe RSV-related illnesses occurred in a 2-year followup period. Infection rates in vaccinées were still slightly lower during the second year of follow-up than in placebo recipients.
Intramuscular administration of PFP did not regularly induce antibody responses in secretions, either in IgA or IgE isotypes, yet still appeared to be protective. This further suggests that serum antibody concentrations are related more closely to protection than to secretory antibody titers.
Many individuals with bronchopulmonary dysplasia or cyanotic congenital heart disease are in the 18- to 36-month-old age group. These individuals are at risk for serious illness at the time of RSV infection and would be good candidates for a vaccine such as PFP. However, the group at greatest risk includes younger infants with and without underlying illness. Unfortunately, infants under the age of 18 months have not responded well to PFP vaccination, although infants undergoing natural infection at a similar age develop excellent antibody responses. An explanation for the poor antibody responses to PFP in this age group is not immediately obvious, especially since other viral pro' tein vaccines (influenza, polio, and hepatitis B) are immunogenic in this population. TTie success of PFP in this age group and in younger infants probably will depend on the use of different adjuvants or perhaps conjugation of PFP to immunologically recognized carriers.
Attempts also have been made to develop live, attenuated RSV strains for use in direct nasopharyngeal immunization.9'10 A cold-adapted candidate strain of RSV was grown in cell culture at 260C before administration to human adults and children. In seropositive infants, this mutant induced immune responses without causing symptoms. However, when seronegative infants were challenged, fever and otitis media occurred.
Further testing with temperature-sensitive mutants of RSV has been carried out in infants and young children.9'10 In one study, a temperature-sensitive mutant of RSV was administered to 32 children. Each of these showed evidence of infection. However, increases in serum-neutralizing antibody titer occurred in only 7 of 25 initially seropositive children. In contrast, all seven initially seronegative infants developed excellent serum -neutralizing antibody responses. Those children who were initially seropositive did not develop any clinical manifestations of illness. All seven initially seronegative infants developed rhinitis, and one developed otitis media.
A limitation of these live mutant vaccines is that some of the vaccine strains recovered following challenge had lost their temperature-sensitivity during replication in the human host. These temperatureinsensitive strains constituted only a small fraction of the virus recovered, but it is not known if they would be capable of spreading to others or if temperaturesensitive strains would retain their sensitivity if they spread to other humans. No infant challenged with live virus has developed more severe disease when undergoing subsequent natural infection with RSV.
Temperature-sensitive mutants that are more stable have been developed by inducing further mutations during growth in cell culture. These strains have been administered to adults, and plans are proceeding to immunize infants and young children with similar strains. A true challenge is to develop live vaccines that do not cause the symptoms observed in past studies of seronegative infants, yet stimulate immune responses in infants who are seropositive because of the presence of transplacentally acquired antibody.
Both inactivated and live RSV candidate vaccines will continue to be tested in infants and young children. Sequential vaccination, with a first dose of live attenuated vaccine followed by boosting with intramuscular subunit vaccines, also is an option. We are encouraged by the fact that influenza subunit and cold-adapted live vaccines are both safe and immunogenic in infants and children of the same age group. Testing of RSV vaccines must proceed at a slower pace because of the phenomenon of vaccine-induced enhanced disease. Curiously, this phenomenon of disease enhancement has not been demonstrated in the case of inactivated influenza or parainfluenza virus vaccines.
Another important step in the development of RSV vaccines is to determine a target population. Clearly, children with underlying cardiac or pulmonary disease would benefit from an RSV vaccine, it can be expected that 1% of all infants in the general population will be hospitalized for RSV infection during their first year of Ufe. These infants also would appear to be good candidates for an RSV vaccine, but it is unclear how they would be identified before infection occurs. Immunization of the entire population of infants to protect these 1% would be feasible only if the vaccine were inexpensive and easily administered.
1 . Klassen TP, Rowe PC, Sutcliffe T. Ropp LJ. McDowell IW, Li MM. Randomized trial of salbutamol in acute bronchiolitis. J Pediotr. 1991:118:807-811.
2. Springer C, Bar-Yishay E, Uwayyed K, Avital A, Vibrai O, Godfrey S. Corticosteroids do not affect the clinical or physiological status of infants with bronchiotitis, ftdietrPaJmonoJ. 1990;9:181-1B5.
3. Smith DW, Frankel LR, Mathers LH. Tang ATS, Ariagno RL, Prob« CG. A control!«! trial of aerosolized ribavirin in infants receiving mechanical ventilation for severe respiratory syncytial virus infection. N EngU Med. 1991:325:24-29.
4. Oroothuis JR, Simoes EAF, Levin MJ, et al. Prophylactic administration of respiratory syncytial vims immune globulin to high risk infants and young children, NEngUMíd. 1993;329:1524-1530.
5. Mills VJ, Van Kirk JE, Wright PF, Chanock RM. Experimental respiratory syncytial virus infection of adults: possible mechanisms of resisrance to infection and illness. J Immimol. 1971:107:123-130.
6. Welliver RC, Wong DT, Sun M, Middleron E Jr, Vaughan RS, Ogra PL. The (development of respiratory syncytial virus-specific IgE and the release of histamine in nasopharyngeal secretions after infecí ion. N EngiJ Med. 1981:305:841-840.
7- Welliver RC, Kau! A, Ogra PL. Cell-mediated immune response to respiratory syncytial virus infection: relationship to the development of reactive airway disease. ; Pafcotr. 1979;94;370-375.
8. Tristram DA, Welliver RC, Mohar CK. Hogerman DA, Hildreth SW, Paradiso P. lmmunogenicity and safety of respiratory syncytial virus subunir vaccine in séropositive children IS to 36 months old. ] Infect Dis. 1993;167: 19 1-195.
9. Kiro HW, Anobio JO, Brandt CD, et al. Safety and anligenicity of temperature sensitive (TS) mutant respiratory syncytial virus (RSV) In infants and children, ftdiarrics. 1973:52:56-63.
10. Wright PE= Shinoiaki T, Fleet W, Sell SH. Thompson J. Kanon DT. Evaluation of a live, attenuated respiratory syncytial virus vaccine in infants. / Pediarr. 1976:88:931936.
Candidate Respiratory Syncytial Virus Vaccines