Influenza virus causes annual, worldwide epidemics of respiratory disease that affect all segments of the population. Children, in particular, are susceptible to influenza infection and its complications. During the influenza pandemics of the 20th century, attack rates in children were as high as 70%. 1, 2 The mortality rates during these pandemics were highest at the extremes of age (children younger than 4 years and adults older than 65 years). 1 Even during interpandemic periods influenza exerts a significant toll on pediatric morbidity and quality of life. Attack rates in children average 20%, and up to one-half of infected children will seek medical care for an influenza-related illness. 3, 4 Similar to the mortality rates seen during pandemic influenza, hospitalization rates during interpandemic influenza are highest at the extremes of age (MR Griffin, CS Coffey, KM Neuzil, EF Mitchel, PF Wright and KM Edwards. Winter viruses: influenza and respiratory syncytial virus-related morbidity. Submitted for publication). 5, 6 Children with underlying medical conditions are at even greater risk than healthy children for adverse outcomes related to influenza. 6, 7
In addition to the morbidity suffered by children as a consequence of influenza infection, children play an important role in the transmission of influenza virus. 8, 9 Children more frequently acquire and shed influenza than adults, and children who attend day care or school may transmit influenza to their household contacts. 10 The arrival of influenza in a community is signaled by both an increase in school absenteeism and an influx of children with febrile respiratory illness into health care facilities. 11
Despite these well-known consequences of influenza in children, inactivated influenza vaccine is rarely given to children either for their own protection, or to prevent the spread of influenza to high risk individuals with whom they may be in contact. Increased awareness of the importance of influenza infection in children has led to discussions about broadening the use of influenza vaccine in the pediatric population. However, some experts are reluctant to embrace the routine use of influenza vaccine in children until the safety and efficacy of the vaccine are further evaluated. 12
From 1985 through 1990 a randomized, controlled trial of cold-adapted and inactivated vaccines for the prevention of influenza A disease was conducted at Vanderbilt University. 13 The study included 5210 healthy subjects, of whom 809 were younger than 16 years. Because influenza A did not circulate in the first year of the study, the 18 children who participated only in Year 1 were excluded from the analysis. Although the cumulative results of the trial have been previously published, the pediatric experience has not been reported. In this paper we report the analysis of the data from this trial in those children younger than 16 years, with an emphasis on inactivated influenza vaccine safety, immunogenicity and efficacy.
Subjects and recruitment
Details of the study have been previously described. 13 Briefly to be eligible for enrollment subjects had to be between the ages of 1 and 65 years and have no underlying cardiovascular, pulmonary, renal or endocrine disease. Potential subjects were excluded if they were allergic to eggs, pregnant or receiving steroids or other immunosuppressive agents. Subjects were assigned to one of three treatment groups (inactivated influenza vaccine, cold-adapted influenza vaccine or control), and subjects and clinical staff were blinded to vaccine assignment.
Inactivated influenza vaccines containing contemporary strains were commercially prepared in embryonated chicken eggs, inactivated, disrupted and purified by standard methods. 14 They contained 15 μg of hemagglutinin from each strain included in the vaccine preparation. During Year 1 of the study, bivalent inactivated vaccine contained only H1N1 and H3N2 antigens; during Years 2 to 5, the trivalent vaccine contained H1N1, H3N2 and influenza B strains.
Attenuated cold-adapted influenza A viruses containing contemporary hemagglutinin and neuraminidase proteins were derived from the cold-adapted master strain (A/Ann Arbor 6/60[H2N2]) and grown in the allantoic cavity of specific pathogen-free eggs as previously described. The titers of the cold-adapted vaccines ranged from 107.0 to 107.6 plaque-forming units/ml with the exception of A/Kawasaki in year 3, which had a titer of 106.0 plaque-forming units/ml. Cold-adapted influenza B vaccines were not sufficiently characterized to include in the study.
Subjects were immunized each fall for up to 5 consecutive years, staying in the same assigned vaccine group. In contrast to the recommendation that previously unvaccinated children age <9 years receive two doses of vaccine for satisfactory antibody responses, 15 all children in this study received only one dose of vaccine each year. Subjects were assigned to one of three vaccine groups: inactivated, cold-adapted or control. The strains of inactivated and cold-adapted vaccine used in each study year are listed in Table 1.
Subjects in all three groups received both an intranasal preparation, delivered as nasal drops, and an intramuscular injection in the deltoid, except for the children <2 years who received the injection in the thigh. Subjects in the inactivated vaccine group received the inactivated vaccine intramuscularly and uninfected allantoic fluid as a placebo in the nose. Subjects in the cold-adapted vaccine group received saline intramuscularly in Year 1 of the study and inactivated monovalent influenza B vaccine in Years 2 to 5, as well as 0.5 ml of each of the two cold-adapted influenza A strains intranasally. Children younger than 3 years received 0.25 ml of the inactivated vaccine or 0.5 ml of a 1/10 dilution of each of the two cold-adapted vaccine preparations as a single dose. Subjects in the control group received an intramuscular injection of saline placebo (Year 1) or monovalent influenza B vaccine (Years 2 to 5). They also received uninfected allantoic fluid as a placebo intranasally in all years of the study.
All subjects were asked to complete a vaccine reaction form to indicate any systemic or local symptoms during the 4 days after immunization. Serum samples were drawn before the first immunization, ∼1 month after each vaccination and in the spring of each year after the end of influenza season. Pre- and postimmunization antibody titers were measured to assess the immune response to vaccine. Postimmunization and postseason antibody titers were measured to assess seroconversion to the circulating strains over the influenza season.
Viral cultures were obtained by vigorously rotating two cotton swabs in the posterior pharynx and then processed and cultured for influenza virus as previously described. 13 All hemadsorbing viruses were characterized by inhibition of hemagglutination with appropriate antisera supplied by the Centers for Disease Control and Prevention. Sera were analyzed for influenza strain-specific antibody by hemagglutination inhibition assay (HAI) assay according to a standard protocol in which formalin-inactivated egg-grown whole virus antigen was used. 16
The influenza season in any study year started on the day that we obtained the first influenza A virus isolate in Nashville and ended on the day that our last isolate was obtained. Patients were notified by mail at the beginning of each influenza season, encouraged to report influenza-like illness to study personnel and given a financial incentive for each illness-related visit. Patients had influenza-like illness if they had fever of abrupt onset with at least one of the following: chills, headache, malaise, myalgia, cough, pharyngitis, or other upper respiratory complaints. They had culture-positive illness if they had influenza-like illness during the influenza season, if they presented for throat culture and if this throat culture was positive for influenza A virus. Seroconversion was defined as a 4-fold rise in serum titer from the postvaccination to postinfluenza season blood draw for the influenza A virus that circulated that season. Vaccines were evaluated for their efficacy in preventing two distinct outcomes: culture-positive illness and seroconversion. In Years 2 through 5, when patients reported for their spring blood draw, they were asked if they had had influenza in the preceding season.
Data management and analysis
Only those participants who were younger than age 16 years at time of vaccination were included in this analysis. Data analysis was performed with the SAS statistical package. 17 Comparability of the vaccine and controls groups with respect to baseline variables and vaccine side effects was assessed with standard contingency table tests on one way analyses of variance. Confidence intervals for proportions were derived using Fleiss’s method. 18 Vaccine efficacy was estimated as the percentage reduction in the incidence of culture-positive influenza illness or HAI seroconversion in each vaccine group relative to controls. This equals 100 (1 − RR), in which RR is the risk of influenza illness relative to controls. Confidence intervals for vaccine efficacy were estimated by a Taylor series approximation. 19
During the 5 years of the study 791 children younger than age 16 years received 1809 doses of vaccine or placebo. The participants were equally distributed among the three groups with respect to age, sex, previous history of influenza illness, number of family members and percent seronegative (HAI titer of <1/8). In general the percent of subjects seronegative to influenza A viruses before receiving any immunization decreased with increasing age, although the percentages seronegative in the 6- to <11- and 11- to <16-year-old groups were similar (Fig. 1).
Febrile, local and systemic reactions to vaccine were assessed in all children. The first year of the study was a pilot year, and only 68 children were enrolled. This was also the only year in which the control group received a saline placebo; in Years 2 to 5 the control group received inactivated influenza B. Although no severe reactions occurred, this study did not have the power to assess differences between local and systemic reactions during Year 1 when the control group was truly placebo.
Local and systemic reactions to vaccine in Years 2 to 5 are shown in Table 2. There were no severe reactions. Reaction rates after initial and repeat vaccination did not vary appreciably. Reaction rates were similar in all age groups and did not vary from year to year. There was a trend toward a higher frequency of arm induration among children who received inactivated trivalent vaccine than among children who received cold-adapted and inactivated influenza B vaccine or influenza B vaccine alone (control); this difference was statistically significant only for children age 11 years and older. Fever was more common in younger children than in older children but did not differ between vaccine groups. Children in the cold-adapted vaccine group had mild increases in the frequency of coryza as compared with children in the other 2 groups; this difference was statistically significant only for children younger than age 6 years. Children ages 11 to <16 years who received cold-adapted vaccine had a higher frequency of sore throat than children receiving other vaccines.
Although children younger than 9 years of age are routinely recommended to receive two doses of vaccine the first year they are immunized, only one dose was administered in this study. This afforded the opportunity to assess seroconversion after a single dose. Based on comparison of baseline and postimmunization serum samples, the immunogenicity of the vaccines in first-time recipients by age group, serostatus at baseline and vaccine serotype is shown in Figure 2. Among children who received inactivated vaccine and were seronegative at baseline, there was a trend toward an improved response to the vaccine with age. Responses to H1N1 and H3N2 serotypes were comparable among seronegative children in the inactivated vaccine group. Among seropositive children who received inactivated vaccine, response rates were higher for children older than 6 years than for children younger than 6 years for the H1N1 serotype. For the H3N2 serotype response rates to inactivated vaccine were vigorous in all age groups. Among children who received cold-adapted vaccine, the H1N1 serotype was more immunogenic than the H3N2 serotype, regardless of age or serostatus at baseline. Among the seronegative children who received cold-adapted vaccine, response rates were higher in children 6 years and older than in children younger than 6 years. Only a minority of seropositive children who received cold-adapted vaccines had a 4-fold rise in HAI titer after vaccination.
The onset of the influenza season each year was defined as the first isolate detected in Nashville through surveillance efforts in our vaccine evaluation clinic, in the children’s hospital and in Vanderbilt student health services. No influenza A illness was detected in the first year of the study (1985 to 1986), only influenza B. During 1986 through 1987 and 1988 through 1989, H1N1 subtypes circulated; during 1987 through 1988 and 1989 through 1990, H3N2 subtypes circulated. The lengths of the epidemic periods varied among the seasons. For the second year of the study (1986 through 1987) the epidemic period was 56 days, for the third year (1987 through 1988) 101 days, for the fourth year (1988 through 1989) 78 days and for the last year (1989 through 1990) 80 days. Based on positive influenza cultures in unvaccinated control subjects, attack rates for symptomatic influenza illness were 7.1 and 4.3% in H1N1 and H3N2 years, respectively. (Table 3) There were too few culture-positive illnesses to assess rates of disease or vaccine efficacy by age group. In all age groups combined both vaccines were highly efficacious at preventing culture-positive influenza illness in children during H1N1 years, with efficacy rates of 95% [95% confidence interval (95% CI), 67 to 99] for cold-adapted vaccine and 91% (95% CI 64 to 98) for inactivated vaccine. Although overall efficacy rates for H3N2 years were also similar, at 68% (95% CI 1 to 90) and 77% (95% CI 20 to 93) for cold-adapted and inactivated vaccine, respectively, confidence intervals were wide for both estimates.
Based on postvaccination to springtime conversion in unvaccinated control subjects, serologic attack rates of influenza infection for all age groups in the H1N1 and H3N2 years were 33 and 16%, respectively (Table 4). There was no association between age group and serologic attack rate. The cold-adapted vaccines were 78 (95% CI 64 to 86) percent and 26 (95% CI –14 to 52) percent efficacious at preventing seroconversion to H1N1 and H3N2 serotypes, respectively. The inactivated vaccines were 67 (95% CI 51 to 78) percent and 65 (95% CI 39 to 84) percent efficacious at preventing seroconversion to H1N1 and H3N2 serotypes, respectively.
Inactivated influenza virus vaccines have been in use for >50 years and remain the cornerstone of influenza prevention efforts in the United States. This trial provides additional evidence that such vaccines are safe and efficacious in healthy children. In this study 277 children younger than age 16 years received 635 doses of inactivated trivalent influenza vaccine. None had a serious adverse event. The frequency of side effects varied by age group. Coryza, fever and cough were the most common reported side effects in children younger than age 6 years, occurring in 16, 11.5 and 8.5%, respectively, of children immunized with inactivated trivalent vaccine. In children older than 6 years, coryza and arm induration were the most common reported side effects. The proportion of these reported side effects attributable to the trivalent vaccine is not possible to determine from this study, because children in the control group received an intramuscular injection of influenza B vaccine. However, the frequency of side effects compares favorably with those of other licensed pediatric vaccines. 20, 21 In previous trials involving >3000 children, split-virus influenza vaccines caused more mild local erythema and tenderness than placebo, but the frequency of systemic reactions was similar. 22, 23
In this study inactivated trivalent influenza vaccines were 91 and 77% efficacious in preventing symptomatic, culture-positive influenza A H1N1 and H3N2 illness, respectively, in children younger than age 16 years. The inactivated vaccine was highly efficacious in preventing culture-proved influenza despite the circulation of antigenically drifted strains during 2 of the 4 years. That inactivated influenza vaccine can protect against antigenically drifted influenza A virus was previously demonstrated in a prospective, nonrandomized controlled trial among children with asthma in Japan. 24 The efficacy of the inactivated vaccine based on HAI seroconversion, a more sensitive but less clinically relevant measure of influenza infection, was somewhat lower at 67 and 66%, respectively, for H1N1 and H3N2 serotypes. These results are similar to other large, controlled trials, in which trivalent inactivated vaccine was 50 to 95% effective at preventing laboratory-confirmed influenza A disease. 24–27 None of these prior studies reported efficacy rates based solely on culture-positive illness.
Although two doses of influenza vaccine are recommended for children receiving vaccine for the first time, 15 only one dose was given in this study. Because many of the subjects participated in the study for multiple years, this difference is relevant only for children in the first year of their enrollment. On the basis of prevaccine serology in this study, >50% of children younger than age 3 years and up to 30% of children younger than 6 years were seronegative to influenza before receipt of their first immunization and may have benefited from 2 doses of inactivated vaccine (Fig. 1). Two clinical trials have demonstrated the clinical efficacy of two doses of inactivated vaccine among children age younger than 3 years. In Finland two doses of inactivated influenza vaccine reduced influenza-associated otitis media by 83% and all otitis media by 36%. 28 Similarly among 6- to 30-month-old children attending day-care centers in North Carolina, inactivated influenza vaccine reduced acute otitis media rates by 41%. 29
The bivalent cold-adapted vaccine was 95.5 and 77.9% efficacious against culture-positive disease and seroconversion, respectively, in children during the 2 years in which influenza A/H1N1 strains circulated. When influenza A/H3N2 circulated, however, the cold-adapted vaccines were considerably less immunogenic and less efficacious (Fig. 2;Tables 3 and 4). Trivalent, cold-adapted vaccines, not available at the time of this study, have since been tested in large clinical trials in young children. Belshe et al. 30 demonstrated in an efficacy field trial that live-attenuated, cold-adapted trivalent intranasal influenza vaccines were 95 and 91% efficacious, based on culture-positive illness , against an antigenically similar influenza A H3N2 and influenza B, respectively, among children 15 to 71 months old. Likewise the vaccine was 86% efficacious against a significantly drifted influenza A/H3N2 virus the following year. 31 The efficacy against H1N1 strains could not be determined, because no H1N1 strains circulated during the time period of that study. 30
Little change has occurred in the manufacturing or delivery of inactivated influenza vaccines in the decade since the completion of this study, and our results should therefore be applicable to current inactivated vaccine use in the pediatric population. The live, attenuated vaccines have evolved, however, and the results from the cold-adapted arm of this study must be interpreted in the context of these changes and may account for the differing results between this study and the more recent study by Belshe et al. 30 The current live-attenuated investigational vaccines are administered as a nasal spray, rather than as nasal drops. The nasal spray may promote better delivery to mucosal surfaces and improve immunogenicity. Furthermore whereas children younger than 3 years in this study received a 1/10 dilution of cold-adapted vaccine, all children older than 15 months in the Belshe study received the full-strength vaccine. 30
The major strengths of this study include the large number and wide age range of the children enrolled, the inclusion of multiple influenza seasons and the virologically confirmed outcome measures. Several limitations exist as well. Only 50% of participants in this trial who reported illness over the influenza season presented for culture during their illness. This may explain the relatively low rates (5.7%) of culture-positive illness among our control group. Using weekly telephone surveillance, albeit in a younger population, Belshe et al. 30 reported culture-positive influenza rates of 18% among control children. A second limitation of this study was the high rate of postimmunization seroconversions among children who did not receive influenza vaccine. These false positive serologies may affect our estimates of vaccine immunogenicity and vaccine efficacy based on seroconversion. However, these false positive serologies would not affect our standard outcome measure, which was efficacy based on culture-positive illness.
This study substantiates the currently licensed, inactivated influenza vaccine as a safe and effective prophylactic agent against influenza A in healthy children. These results should aid parents, health care providers and policy makers in their decisions regarding influenza immunization of healthy children. These findings should also encourage parents and providers to immunize the nearly 70% of high risk children who still do not receive the recommended inactivated vaccine.
We thank Mr. Dale Plummer for statistical assistance and Dr. Marie Griffin for critical review of the manuscript.
Grant support was provided by the National Institutes of Health (AI-52594).
1. Dauer CC, Serfling RE. Mortality from influenza. Am Rev Respir Dis 1961; 83: 15–28.
2. Dunn FL, Carey DE, Cohen A, Martin JD. Epidemiologic studies of Asian influenza in a Louisiana parish. Am J Hyg 1959; 70: 351–71.
3. Glezen WP, Keitel WA, Taber LH, Piedra PA, Clover RD, Couch RB. Age distribution of patients with medically-attended illnesses caused by sequential variants of influenzaA/H1N1: comparison to age-specific infection rates, 1978–1979. Am J Epidemiol 1991; 133: 296–304.
4. Wright PF, Thompson J, Karzon DT. Differing virulence of H1N1 and H3N2 influenza strains. Am J Epidemiol 1980; 112: 814–9.
5. Neuzil KM, Mellen BG, Wright PF, Mitchel EF, Griffin MR. The impact of influenza on hospitalizations, outpatient visits, and antibiotic prescriptions in children. N Engl J Med 2000; 342: 225–31.
6. Izurieta HS, Thoompson WW, Piotr K, et al. Influenza and the rates of hospitalization for respiratory disease among infants and young children. N Engl J Med 2000; 342: 232–9.
7. Neuzil KM, Wright PF, Mitchel EF, Griffin MR. The burden of influenza illness in children with asthma and other chronic medical conditions. J Pediatr 2000; 137: 856–64.
8. Fox JP, Hall CE, Cooney MK, Foy HM. Influenzavirus infections in Seattle families, 1975–1979. Am J Epidemiol 1982; 116: 212–27.
9. Longini IM, Koopman JS, Monto AS, Fox JP. Estimating household and community transmission parameters for influenza. Am J Epidemiol 1982; 115: 736–51.
10. Hurwitz ES, Haber M, Chang A, et al. Effectiveness of influenza vaccination of day care children in reducing influenza-related morbidity among household contacts. JAMA 2000; 284: 1677–82.
11. Glezen WP, Cough RB. Interpandemic influenza in the Houston area, 1974–76. N Engl J Med 1978; 298: 587–92.
12. McIntosh K, Lieu T. Is it time to give influenza vaccine to healthy infants? N Engl J Med 2000; 342: 275–6.
13. Edwards KM, Dupont WD, Westrich MK, Plummer WD, Palmer PS, Wright PF. A randomized controlled trial of cold-adapted and inactivated vaccines for the prevention of influenza A disease. J Infect Dis 1994; 169: 68–76.
14. Murphy BR, Webster RG. Orthomyxoviruses. In: Field BN, ed. Field’s virology. New York: Raven Press, 1990: 1091–152.
15. Centers for Disease Control and Prevention. Prevention and control of influenza. MMWR 2000; 49: 1–38.
16. Subbarao EK, Kawaoka Y, Ryan-Poirier K, Clements ML, Murphy BR. Comparison of different approaches to measuring influenza A virus-specific hemagglutination inhibition antibodies in the presence of serum inhibitors. J Clin Microbiol 1992; 30: 996–9.
17. SAS language: reference, version 6. 6th ed. Cary, NC: SAS Institute, 1990.
18. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: Wiley, 1981: 13.
19. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research: principles and quantitative methods. New York: Van Nostrand Reinhold, 1982:297–9.
20. Decker MD, Edwards KM, Steinhoff MC, et al. Comparison of 13 acellular pertussis vaccines: adverse reactions. Pediatrics
1995; 96: 557–66.
21. Rennels MB, Edwards KM, Keyserling HL, et al. Immunogenicity and safety of 7-valent pneumococcal-CRM197 conjugate vaccine. Pediatrics
1998; 101: 606–11.
22. Wright PF, Cherry JD, Foy HM, et al. Antigenicity and reactogenicity of influenza A/USSR/77 virus vaccine in children: a multicentered evaluation of dosage and safety. Rev Infect Dis 1983; 5: 758–64.
23. Wright PF, Thompson J, Vaughn WK, Folland DS, Sell SHW, Karzon DT. Trials of influenza A/New Jersey/76 virus vaccine in normal children: an overview of age-related antigenicity and reactogenicity. J Infect Dis 1977; 136: S731–41.
24. Sugaya N, Nerome K, Ishida M, Matsumoto M, Mitamura K, Nirasawa M. Efficacy of inactivated vaccine in preventing antigenically drifted influenza type A and well-matched type B. JAMA 1994; 272: 1122–6.
25. Clover RD, Crawford S, Glezen WP, Taber LH, Matson CC, Couch RB. Comparison of heterotypic protection against influenza A/Taiwan/86 (H1N1) by attenuated and inactivated vaccines to A/Chile/83-like viruses. J Infect Dis 1991; 163: 300–4.
26. Hoskins TW, Davies JR, Allchin A, Miller CL, Pollock TM. Controlled trial of inactivated influenza vaccine containing the A/Hong Kong strain during an outbreak of influenza due to the A/England/42/72 strain. Lancet 1973: 116–20.
27. Hoskins TW, Davies JR, Smith AJ, Miller CL, Allchin A. Assessment of inactivated influenza-A vaccine after three outbreaks of influenza A at Christ’s Hospital. Lancet 1979: 33–5.
28. Heikkinen T, Ruuskanen O, Waris M, Ziegler T, Arola M, Halonen P. Influenza vaccination in the prevention of acute otitis media in children. Am J Dis Child 1991; 145: 445–8.
29. Clements DA, Langdon L, Bland C, Walter E. Influenza A vaccine decreases the incidence of otitis media in 6- to 30-month-old children in day care. Arch Pediatr Adolesc Med 1995; 149: 1113–7.
30. Belshe RB, Mendelman PM, Treanor J, et al. The efficacy of live attenuated, cold-adapted, trivalent, intranasal influenzavirus vaccine in children. N Engl J Med 1998; 338: 1405–12.
31. Belshe RB, Gruber WC, Mendelman PM, et al. Efficacy of vaccination with live-attenuated, cold-adapted, trivalent, intranasal influenza virus vaccine against a variant (A/Sydney) not contained in the vaccine. J Pediatr 2000; 136: 168–72.