Clinical experience with respiratory syncytial virus vaccines

Abstract

Respiratorysyncytial virus (RSV) infection is at times associated with life-threatening lower respiratory tract illness in infancy. Severe infection during the first year of life may be an important risk factor or indicator for the development of asthma in early childhood. Severe infections primarily occur in healthy infants, and young infants and children with specific risk factors. However, RSV causes respiratory infections in all age groups. Indeed it is now recognized that RSV disease is responsible for significant morbidity and mortality in the geriatric population. RSV infection remains difficult to treat, and prevention is a worldwide goal. For this reason there has been an intensive effort to develop an effective and safe RSV vaccine.

Initial infection with RSV affords limited protection to reinfection, yet repeated episodes decrease the risk for lower respiratory tract illness. In the 20 years from 1960 to 1980, trials of several candidate RSV vaccines failed to attain the desired safety and protection against natural infection. Some vaccine types either failed to elicit immunogenicity, as with the live subcutaneous vaccine, or resulted in exaggerated disease on natural exposure to the virus, as with the formalin-inactivated (FI) type.

Currently vaccine candidates are being developed based on the molecular virology of RSV. Recent formulations of candidate RSV vaccines have focused on subunit vaccines [such as purified fusion protein (PFP)], subunit vaccines combined with nonspecific immune activating adjuvants, live attenuated vaccines (including cold passaged, temperature-sensitive or cpts mutants), genetically engineered live attenuated vaccines and polypeptide vaccines.

INTRODUCTION

Respiratory syncytial virus (RSV) infection causes respiratory tract illness in individuals of all ages. ¹ Morbidity is greatest in young healthy infants (<6 months), children with underlying risk factors such as prematurity, ² bronchopulmonary dysplasia, ³ cystic fibrosis, ⁴ congenital heart disease ⁵ or immunodeficiency ⁶ and, data now suggest, the elderly. ^7–9

Antibodies are generated in response to most RSV proteins. ^10–12 Humoral and secretory antibodies targeted at the fusion (F) and attachment (G) viral surface proteins are probably most important; however, these antibodies offer only limited protection against infection. Reinfection is common throughout life and can occur even when high levels of virus-neutralizing antibodies are found. ¹³ High levels of serum neutralizing antibodies do reduce the risk of lower respiratory tract illness, with limited benefit against upper respiratory tract infection. Maternally derived antibodies are important in very young infants in protecting against RSV disease. ^{14, 15}

Strategies for developing vaccines include subunit vaccines, subunit vaccines combined with nonspecific immune activating adjuvants, live attenuated vaccines, live attenuated genetically engineered vaccines and polypeptide vaccines. Of these the subunit PFP, the polypeptide G (BBG2Na, amino acids 130 to 230 of G fused to the albumin-binding domain of streptococcal G protein) and live attenuated virus (cold passaged, temperature-sensitive mutants and genetically engineered strains) candidate vaccines are being evaluated clinically and may show promise with further development. These vaccines will be briefly reviewed, and results of relevant clinical trials in the public domain summarized.

RSV AND THE IMMUNE RESPONSE

RSV infection induces secretory antibodies, serum neutralizing antibodies and T cell-specific immunity. Maternally derived antibodies in the very young will help protect against RSV infection, and the development of secretory antibodies, serum antibodies and cytotoxic T cells will affect recovery from illness. ¹⁴

In the initial stages of infection, RSV attaches to and invades the respiratory epithelium lining the nasal cavity. ¹⁴ Once a host is infected or exposed to RSV, the innate mucosal immune response is triggered. The innate response is important in inflammation and in the development of the adaptive immune response. If innate immunity is not sufficient to prevent viral replication in the upper respiratory tract, an upper respiratory tract illness ensues. The mucosal immune response that includes secretory antibodies will help curtail the infection. If the virus spreads to lower respiratory tract via direct cell fusion and/or aspiration, high levels of serum neutralizing antibodies can prevent lower respiratory tract illness.

Once infection is established, the cellular immune response (including cytotoxic and helper T cells) promotes viral clearance. ¹⁴ Evidence for this appears in studies that compare viral shedding in immunocompetent infants vs. children with deficient cellular immunity. RSV shedding generally stops within 21 days after infection in healthy infants and young children; in contrast children with cellular immunodeficiency can shed virus for several months. ^{6, 16} Research is currently being conducted to better understand the contribution of cellular immunity and the different T cell subsets in recovery from RSV infection. ¹⁷ In the mouse model a T helper cell (Th) 1 response to RSV infection is demonstrated, with subsequent production of interferon-gamma, interleukin-2 and IgG2a. ^18–22 However, if RSV antigen (as opposed to natural wild-type virus infection) is used in eliciting the primary immune response, a Th2 cytokine expression pattern is evoked. ²³ Further, use of a live virus stimulates a Th1-type response, whereas use of an inactivated virus results in a Th2 cytokine expression pattern. ²³

In infants a primary RSV infection elicits a poor immune response, particularly among infants <6 months of age, and has limited effects on subsequent reinfection. ²⁴ In fact reinfection and the incidence of lower respiratory tract illness are not significantly decreased until the third infection with RSV. ^{15, 25} After primary infection in young infants, virus-specific neutralizing antibody and antibodies directed to the two main viral surface proteins (F or G) are often reduced or minimal. ²⁶ More appropriate immune responses occur in older infants (>9 months) and young children after primary infection and reinfection, although the response is still less than that of an adult. Decreased rates of lower respiratory tract illness observed in children after multiple RSV infections suggest that vaccines may be successful in preventing lower respiratory tract illness.

GOALS AND COMPLEMENTING VACCINE STRATEGIES

RSV is a prime candidate for vaccine development, and because the greatest morbidity and mortality from RSV occurs in infants, there is need for a vaccine that can be administered to very young individuals. However, attempts to create effective and safe RSV vaccines have been hindered by several factors, including: weak immune responses in newborns and young infants; suppression of immune response by circulating maternally derived antibody ^{8, 26, 27}; and early experience with a vaccine that potentiated RSV disease. ^28–30

The first RSV vaccine tested in the mid-1960s was a formalin-inactivated (FI) and aluminum-precipitated vaccine administered via the intramuscular route. ^28–31 This vaccine was associated with an exaggerated respiratory disease on subsequent natural infection with RSV. ^28–31 This phenomenon was observed only in children <2 years of age. The most widely accepted explanation for the enhanced disease is that an imbalance occurred in the Th1 and Th2 lymphocyte responses to the vaccine. ¹⁴ Normally Th1-type responses with interferon-gamma and interleukin-2 production are expressed upon virus-induced infection. However, the pulmonary histologic findings in mice immunized with the FI-RSV vaccine were more consistent with a Th2-type response with an ineffective immune response to RSV. ¹⁴

Although a licensed vaccine for the prevention of RSV is not yet available, there has been significant progress in the development and evaluation of both subunit and live attenuated RSV vaccine candidates. ²⁴ For a vaccine to be effective, it must target the multiple strains of RSV that circulate during an outbreak. In addition, the immune response that is elicited should be directed against both of the main RSV antigenic groups: A and B. ^{24, 32} An effective vaccine for RSV would ideally elicit robust humoral, secretory and cellular immune responses of sufficient magnitude to protect the respiratory mucosa, but not even natural infection induces such a robust immune response in healthy adults. Thus a realistic goal of a vaccination strategy might not be to prevent infection, but rather to prevent serious disease. Recently the goals of a vaccination program and a number of complementing strategies were proposed that take into consideration the advantages and disadvantages of the current candidate subunit and live attenuated vaccines. ²⁴

Goals of vaccination

With our limited understanding of the immunologic requirements for protection against infection in humans, and with the establishment of serum neutralizing antibody as an immune correlate of protection against lower respiratory tract illness in the immunocompetent host, it would seem reasonable that an efficacious vaccine must prevent lower respiratory tract illness. Thus the potential goals of an RSV vaccination program are to prevent RSV lower respiratory tract illness and the hospitalizations and deaths associated with RSV infection. The indirect benefits of a successful vaccination program would be reflected in diminished secondary complications of otitis media, sinusitis and the infrequent bacterial superinfections of the lungs. It would also be anticipated that there might be a reduction in the inappropriate use of antibiotics for the treatment of common respiratory infections that are primarily viral in origin. A reduction in antibiotic usage contributes to the control of antimicrobial resistance in community-acquired bacterial pathogens.

Vaccine approaches

Strategies for the prevention of RSV disease include maternal immunization in the third trimester, a primary immunization schedule with the first dose administered at birth or shortly after birth, immunization for high risk groups utilizing CDC guidelines similar to those for the influenza vaccine and, perhaps ultimately, universal immunization of preschool and school age children. Universal immunization of preschool and school age children might be considered if they have high infection rates and are responsible for transmission of RSV to susceptible members in households and the community at large. Such a vaccination strategy in children is being evaluated to control influenza outbreaks in the community.

Table 1 presents some of the recent formulations that have been or are being investigated for RSV vaccines. The initial attenuated live virus mutants were temperature-sensitive or cold-passaged mutants of the RSV/A subgroup virus (not included in Table 1). Initial live virus vaccine candidates were either underattenuated or overattenuated, and research on these vaccine types was not pursued. ²⁴

TABLE 1: Current approaches in RSV vaccine formulation

Interest in RSV vaccines was revived with the purification and characterization of the fusion protein. ³⁴ The principal antigen in the PFP vaccines is the fusion (F) glycoprotein, which is highly conserved among the major RSV groups. The F protein also induces serum neutralizing antibodies. ³⁴ Currently there is great interest in administering subunit vaccines with adjuvants that are nonspecific immune activators to elicit a balanced humoral and cellular immune response.

Maternal immunization. The goal of maternal immunization is to boost serum neutralizing antibodies in the mother during the second or third trimester of pregnancy, which in turn would enhance the active placental transport of serum neutralizing antibodies from mother to infant. Infants born before or during the RSV season, with higher serum neutralizing antibodies against RSV, should be better protected against severe RSV illness. ³⁵ Because the majority of infants hospitalized for RSV are younger than 5 months of age, maternal immunization could potentially prevent a significant proportion of serious lower respiratory tract illness in early infancy. ³⁶ At birth RSV-neutralizing antibodies in full term infants approximate maternal levels. ³⁷

Munoz et al. ^{24, 38} evaluated maternal immunization with a second generation purified F protein (PFP-2) vaccine. In a Phase II randomized, double blind, placebo-controlled study to assess safety and immunogenicity, healthy pregnant women received either vaccine or saline in a 4:3 ratio. A total of 35 healthy women, ages 18 to 45 years in weeks 30 to 34 of gestation, participated in the study. The time between vaccination and delivery was 46.9 days; this is important because time is necessary for the mother to develop an appropriate immune response and to actively transfer RSV-specific antibodies to the neonate. The vaccine was well-tolerated by both mother and fetus. The study was not designed to evaluate efficacy, but it did demonstrate that the PFP-2 vaccine was safe and not associated with enhanced disease (as indicated by no increase in febrile disease, lower respiratory tract illness, otitis media, antibiotics use or need for medical attention). ^{24, 38}

The PFP-2 vaccine in the pregnant healthy women was only modestly immunogenic. However, there was virtually 100% transmission of neutralizing antibodies to the infant, and the antibody decay rate was ∼36 to 37 days. Thus with a vaccine that is more immunogenic, this strategy may be a potential avenue for protecting young infants during the greatest vulnerability for severe disease caused by primary RSV infection. ²⁴

Primary immunization with live attenuated vaccines. Because of historic adverse experiences with the FI-RSV vaccine, it is unlikely that a subunit vaccine will be evaluated in infants in the near future. Live attenuated RSV vaccines, however, were never associated with enhanced disease severity. Thus primary immunization is being actively investigated with attenuated and genetically engineered live RSV vaccines. This strategy probably will require two to three serial doses to protect young infants during the initial season. A series of live attenuated, cold-passaged, temperature-sensitive (cpts) RSV mutants have been evaluated in healthy adults, seropositive children and RSV-naive infants. The cpts 530/1009, cpts 248/955, and cpts 248/404 mutants have been shown to be safe in adults and RSV-seropositive children. These cpts mutants, however, have been underattenuated in young infants and thus are not sufficiently safe vaccine candidates to proceed further in vaccine trials in infants. ^{39, 40}cpts 530/1009 and cpts 248/955 produced prolonged viral replication in infants and were insufficiently attenuated; the virus was transmitted to some of the placebo recipients. Wright et al. ⁴⁰ evaluated cpts248/404 in seronegative infants. The vaccine was given in a two dose regimen to infants as young as 1 to 2 months. The first dose was associated with significant viral shedding, with titers in the range of 10³ and 10⁴ plaque-forming units per ml in secretions. Infants shed virus for up to 3 weeks. cpts 248/404 was associated with an increased frequency of upper respiratory symptoms and little evidence of lower respiratory illness. ⁴⁰ Although the serum neutralizing antibody induced by this vaccine was poor, there was a reduction in reinfectivity with the second dose, suggesting that the first dose was at least somewhat protective.

It is important for a live RSV candidate vaccine to retain genetic stability. Attenuation of the vaccine may be the result of only a relatively small number of nucleotide and amino acid changes. ^{41, 42} Factors that are critical in development of an effective live vaccine are the balance between a level of attenuation that is safe for infants and one with immunogenicity that is sufficient to protect against natural infection. Recent developments in reverse genetic technology have allowed for the construction of an infectious virus with multiple specific mutations (rABcp 248/404/1030, rA2cp 248/404/1030εSH). These genetically engineered mutants have recently been evaluated in infants and appear to be promising as vaccine candidates.

The potential advantages of live attenuated vaccines include the generation of both local and systemic immunity, the ability to administer multiple times and the ability to infect in the presence of maternal antibody (Table 2). However, a major issue is that live vaccines based on mutant strains may not be sufficiently attenuated and genetically stable for use in very young RSV seronegative infants. Genetically engineered mutant strains may offer a successful approach for creating a vaccine for use in infants that is attenuated, immunogenic, genetically stable and nontransmissible. ²⁷ Vaccination of infants could be performed in conjunction with maternal immunization. Maternal immunization might provide the infant additional protective time to delay the primary vaccination series with a live attenuated vaccine. Thus instead of starting at birth, the initial dose may be given at 4 to 6 months of age. This might allow the infant to better tolerate the vaccine, elicit a more robust immune response and still remain protected during its most vulnerable period in life.

TABLE 2: Live attenuated RSV vaccine in infants. Immunization of neonates with live attenuated RSV vaccines is under evaluation, and limited studies have been conducted in recent years. Protection against natural RSV infection has not yet been demonstrated. A live attenuated vaccine virus that is sufficiently attenuated and immunogenic in seronegative infants and not transmissible must first be developed before this approach can be effectively tested. ²⁴

High risk groups using subunit vaccines. The RSV F and G proteins are the main targets for induction of neutralizing antibodies. Therefore they have been the focus of subunit vaccine development. ^{24, 43} Subunit vaccines are designed to induce a neutralizing antibody response capable of protecting the lower respiratory tract. Several vaccine candidates have been evaluated in RSV-seropositive children and adults. The subunit vaccines PFP-1 and PFP-2 have been tested in healthy seropositive children, ^{44, 45} children with cystic fibrosis ⁴⁶ and children with bronchopulmonary dysplasia. ⁴⁷ These vaccines are safe and at times reduce respiratory problems during RSV infection.

The current subunit vaccines [PFP, BBG2Na (a polypeptide containing a conserved region of the G protein fused to the albumin-binding domain of streptococcal G protein) and F and G chimeric] are not expected to induce a strong cytotoxic T lymphocyte response. ^{44, 48, 49} These types of vaccines elicit neutralizing antibodies and are generally safe in primed individuals (previously infected). They are thus suitable for vaccinating persons with underlying chronic conditions that increase their risk for more severe disease. ¹ These high risk groups include individuals with asthma, children with cystic fibrosis (CF), and the elderly. Subunit vaccines and BBG2Na are being analyzed in these groups because they appear safe and immunogenic and may provide protection against RSV lower respiratory tract infection. ²⁴

A small study of children with cystic fibrosis who were given the PFP-2 vaccine had fewer lower respiratory illnesses during the RSV season than did unvaccinated CF children. ⁴⁶ On the basis of these promising early results, Hiatt et al. ⁵⁰ tested a third generation (PFP-3) vaccine in RSV-seropositive CF children. The study was multicenter, placebo-controlled and designed to determine the safety, immunogenicity and effectiveness of the PFP-3 vaccine against lower respiratory tract illness during the RSV season. A total of 297 children ages 1 through 12 years with cystic fibrosis, who were serum-positive to RSV, participated. Demographics were comparable between the placebo and vaccination groups.

The primary outcome measure was reduction of lower respiratory tract illness events (regardless of causality) throughout the trial. Overall although it was not statistically significant, there was an ∼10% reduction in lower respiratory tract illness events in the vaccine recipients compared with controls (38% in the vaccine group vs. 48% of the control subjects). Similarly the cumulative mean number of lower respiratory tract illness episodes was lower in the vaccination group vs. control (9% reduction, P not significant). During the trial planning it was estimated that there would be a mean of two lower respiratory tract illness events per patient, but in actuality slightly less than one event occurred per patient. Therefore the study was underpowered to detect significant differences. This trial also did not evaluate culture-positive illness (or efficacy), but rather the overall impact of the vaccine on lower respiratory tract morbidity (effectiveness).

SUMMARY

The incubation period for RSV is ∼4 days. Mounting an effective immune response takes at least this long. Consequently by the time immune mechanisms are fully recruited, the virus has been actively replicating for a number of days.

To be effective an RSV vaccine must be administered prophylactically. Further the vaccine must be used in time to prevent primary RSV infection, because it is the initial episode that is associated with the highest rates of morbidity and mortality.

Given that naturally occurring RSV infection fails to afford complete protection against reinfection, an RSV vaccine should most likely target prevention of lower respiratory disease. In that regard subunit and cpts candidate vaccines appear promising, although they will not be commercially available for a number of years.