Influenza A and B viruses are a major cause of respiratory tract illnesses in children. 1 Recent studies have demonstrated increased risk of hospitalizations because of influenza in young children. 2, 3 Children can also act as a reservoir for transmission of virus to other household members and to the general population. 4–6 For 2002 to 2003 the ACIP encouraged the use of the influenza vaccine in children 6 to 23 months of age. 7 The lack of a rapid and easy means of administering influenza vaccine on an annual basis to children may slow down the implementation of this more widespread vaccination program. Live attenuated, cold-adapted, trivalent intranasal influenza virus vaccine (CAIV) could represent an alternative approach to influenza vaccination in children and adults.
CAIV is produced by genetic reassortment between a wild-type influenza virus and an attenuated master donor virus for each 6:2 reassortant strain to be included in the vaccine. Each vaccine virus derives two gene segments encoding the hemagglutinin and neuraminidase from a wild-type influenza virus, and the remaining six gene segments encoding proteins are responsible for cold adaptation, temperature sensitivity and attenuation from the master donor virus. The vaccine is grown in the allantoic cavity of pathogen-free eggs and is a combination of two influenza A strains and one influenza B strain. 8–11 Prior studies have shown that CAIV was effective in preventing influenza and influenza-related otitis media in children. 12, 13 It was the purpose of this study to evaluate the safety of CAIV containing influenza A and influenza B strains in a study population of ∼10 000 children with medical utilization in the hospital, clinic and emergency department as outcomes.
Vaccine and placebo.
MedImmune Vaccines, Inc., supplied CAIV in frozen single dose intranasal applicators. Each 0.5-ml dose was delivered as an intranasal spray. Each applicator had a dose divider that was to allow for the delivery of approximately one-half the contents to each nostril. Each dose of CAIV contained 107 median tissue culture infectious dose of each of the three 6:2 reassortant strains of influenza recommended by the Food and Drug Administration for the 1999 to 2000 influenza season: A/Beijing/262/95 (H1N1); A/Sydney/05/97 (H3N2); and B/Yamanashi/166/98 influenza strains. Vaccine and placebo contained allantoic fluid containing sucrose-phosphate-glutamate. 14–16
The protocol was reviewed and approved by the Kaiser Permanente Institutional Review Board. Healthy children and adolescents 1 through 17 years of age were eligible to enroll in the study. To be eligible children must not have received the 2000 to 2001 formulation of trivalent influenza vaccine. Children also could not have received any live virus vaccine within 1 month of enrollment or any inactivated vaccine within 2 weeks of enrollment in this study. Children were excluded from the study if they had a history of egg allergy, a history of asthma (by parent report), a fever (>100.0°F oral) or respiratory illness within 72 h of enrollment. In addition children with immunodeficiency or children taking immunosuppressive agents were excluded.
This was a randomized, double blind, placebo-controlled study. The randomization ratio was 2:1 (CAIV to placebo). Children <9 years of age received a second dose of the same agent 28 to 42 days after the first dose. Enrollment began in October 2000 and was completed at the end of December 2000. Enrolled children were then followed for 42 days after each vaccination for any medically attended event or serious adverse event (SAE).
Data collection and statistical analysis.
Within Kaiser Permanente (KP) children are assigned a unique medical record number that identifies them for life. Kaiser Permanente maintains automated clinical information system databases that contain diagnoses for all clinic, emergency and hospital visits. In addition, because KP is self-insured, claims for utilization outside of KP are also available from computerized data. For the purposes of this study, a study population database was created from the vaccination logs submitted by study nurses at the 31 participating sites. On a weekly basis this database was linked with utilization databases to identify any hospitalizations, emergency visits or clinic visits within 42 days of receipt of a dose of vaccine. The investigators reviewed each utilization to confirm the diagnosis and assess severity and possible relationship to vaccine. For serious adverse events (SAEs), additional follow-up was obtained through contact with the patient’s personal physician, with their parents and through medical record review.
Outcomes in this study included four prespecified diagnostic groupings and an analysis that included all observed diagnostic categories. Rates of events were compared for each diagnostic group and individual diagnostic category, and a rate ratio was calculated. The first event for each diagnosis for each child in each observation window was used. The relative risk (RR) and its two-sided 90% confidence interval (CI) were calculated according to the midprobability exact binomial method adjusted for follow-up time. 17 The four prespecified diagnostic groups included acute respiratory tract events (mastoiditis, sinusitis, laryngitis, tracheitis, laryngotracheitis, epiglottitis, croup, bronchitis, bronchiolitis, viral pneumonia, bronchopneumonia with unspecified organism, pneumonia with unspecified organism, influenza with pneumonia, influenza with other respiratory manifestations, extrinsic asthma, intrinsic asthma, unspecified asthma, wheezing, pulmonary congestion, shortness of breath), systemic bacterial infection, acute gastrointestinal tract events and rare events potentially associated with wild-type influenza (encephalitis, Reye syndrome, myocarditis and pericarditis, Guillain-Barré syndrome, polymyositis and seizures).
Comparisons were made for each setting separately and for all three settings combined, for each dose separately and for both doses combined, as well as for each of four age groups: 1 to 8 years of age, 9 to 17 years of age, 12 to 17 months of age and 18 to 35 months of age and all ages combined. No adjustment was made for these multiple comparisons in the statistical analysis. A statistically significant increase in the CAIV group was declared if the lower bound of the 90% CI for the rate ratio exceeded 1. Detecting significant decreases in the CAIV group was not a goal of the trial. Such results are presented for contextual purposes.
Between October 2, 2000 and December 22, 2000, 9689 evaluable children were enrolled. Of these, 3769 children 1 to 8 years of age received CAIV and 1868 received placebo, whereas in children 9 to 18 years of age, 2704 children received CAIV and 1348 received placebo. Overall 86.6% of the younger children received both a first and second dose. CAIV participants were 51% female compared with 50% in controls. The population was 6% African-American, 55% White, 20% Hispanic, 10% Asian and 9% other.
Overall utilization rates by setting (emergency room, clinic, hospital) are summarized in Table 1. In comparing overall numbers of events, it is important to note that because of the 2:1 randomization, twice the number of events is expected in the CAIV group than in the placebo group. The percentages of children with utilization in the hospital, emergency room and clinics are almost identical in the two groups.
The rates of utilization for the prespecified diagnostic groups are shown in Table 2 for all utilization settings combined. There were no significant increases for these outcomes in vaccinees when utilization at all settings was combined; however, significantly elevated rates were observed in some analyses in the CAIV group for acute respiratory tract events and for acute gastrointestinal events when analyzed for individual utilization settings as shown in Table 3. Of note, but not shown, is that significantly decreased rates were observed in 16 comparisons for these prespecified outcomes by individual utilization setting, including 8 analyses for acute respiratory events and 8 analyses for acute gastrointestinal events.
Medical utilization occurred for 170 unique individual diagnostic categories. When accounting for the comparisons by dose, age and setting, >1500 statistical comparisons were made in this study. The comparisons for which we observed a statistically elevated risk ratio are shown in Table 4. During the 42-day observation period, significantly elevated risks were observed in one or more comparisons for asthma, abdominal pain, upper respiratory infection (URI), musculoskeletal pain, otitis media with effusion and adenitis/adenopathy as well as for several diagnostic categories not thought to have biologic likelihood. Further evaluation was conducted for events thought to have a biologically plausible association.
Overall asthma diagnoses were observed in 0.9% of CAIV recipients and 0.9% of controls. Elevated risk ratios were observed in 4 of the 31 separate comparisons, all of these in children 18 to 35 months of age. In this age group there were 16 asthma events (all settings and doses combined) in CAIV recipients and 2 in placebo recipients (RR 4.06; 90% CI 1.29 to 17.86). When the time association for asthma was evaluated (Fig. 1), no consistent time association of these events with receipt of vaccine was noted. All of these events occurred in the clinic except for one in the emergency department. No child required hospitalization. Treatment consisted of beta-2 agonist for 94% (17 of 18), antibiotics for 56% (10 of 18), systemic corticosteroids for 33% (6 of 18) and inhaled corticosteroids for 17% (3 of 18). The data were subsequently also evaluated by combining the diagnoses of asthma, shortness of breath and wheezing. The relative risks were no longer significant using this combined diagnostic category, peaking at RR = 2.19; 90% CI 0.90 to 6.00 after the first dose of vaccine in children 12 to 35 months (17 of 904 for the vaccine group and 4 of 465 for the placebo group).
Although history of asthma or possible asthma according to the parent was an exclusion criteria for participation in the trial, 7 of the 16 (44%) CAIV participants 18 to 35 months of age with the diagnosis of asthma had a prior visit for asthma in their medical record. To further evaluate a possible increased risk of asthma in CAIV recipients with a prior history of asthma, the electronic medical record for all participants was reviewed to identify children with a visit before the study for asthma/reactive airway disease. This review revealed 8.8% of study participants with a history of asthma/reactive airway disease. In these children receipt of CAIV was not associated with an increased risk of an asthma diagnosis in the 42 days after vaccine with a RR of 1.11; 90% CI 0.59 to 2.14.
In all settings combined, abdominal pain occurred in 0.7% (47 of 6473) of the vaccine group and 0.8% (26 of 3216) of placebo recipients. However, in 2 of 26 analyses, there was an elevated risk ratio for abdominal pain, both events in the emergency department (Table 4). There were 14 children with abdominal pain in the emergency department in the CAIV group compared with one in the placebo group. No consistent time association was observed for abdominal pain in the emergency department setting. To evaluate a possible increased risk of potentially severe medical events that can present with abdominal pain, medical utilization for the following diagnostic categories was identified: appendicitis; gastroenteritis; intestinal obstruction; mesenteric adenitis; pancreatitis; intestinal perforation; ulcer; volvulus; and intussusception. Apart from gastroenteritis, where a relative risk of 0.87 was observed (90% CI 0.60, 1.11), 1 event each was observed for appendicitis and “rule-out” appendicitis. Review of the medical record for the child with appendicitis revealed onset of abdominal pain before receipt of CAIV.
The diagnostic category of URI was significant for increased risk in 3 of 41 comparisons. The comparison containing the greatest number of events was 18 to35 month olds in the combined setting. The RR was 1.30, and the 90% CI was 1.01 to 1.67. Over the 42-day observation period, there was no temporal relationship with vaccination.
There was at least one analysis demonstrating increased risk for the diagnostic categories of musculoskeletal pain, otitis media with effusion and adenitis/adenopathy. In each of these categories, the number of events was small. When these diagnostic categories were evaluated for a temporal relationship with vaccination, no consistent time association was observed. However, because otitis media with effusion is a chronic disorder without an acute onset, evaluating time association therefore is difficult.
When the children with enuresis, speech delay and seborrhea were reviewed, each of these children had a history of these conditions before receipt of vaccine. The majority of children with otitis externa and urinary tract infection had an etiologic agent identified. The children hospitalized for elective procedures were each receiving treatment for conditions identified before trial participation.
The diagnostic categories with observed decreased risk ratios are shown in Table 5. As can be seen, decreased relative risks were observed for 51 comparisons compared with 24 with elevated risk ratios. Some of the comparisons with significantly decreased relative risks included the same diagnostic categories for which elevated risks were observed in other comparisons tending to decrease the likelihood of a true causal relationship with vaccination.
The rate of SAE was 0.2% and equally distributed between vaccine (13 of 6473) and placebo (7 of 3216). No SAE in vaccine recipients was judged related to CAIV.
Several studies have evaluated the efficacy of cold-adapted intranasal influenza vaccines in children. 12, 13, 18–22 A recent study also reported on the safety of the CAIV in children. 22 Our study was designed exclusively to evaluate the safety of this vaccine in children and adolescents. In this study there was utilization for 170 unique diagnoses resulting in >1500 statistical analyses (Tables 4 and 5). Therefore one would expect some observations to achieve statistical significance by chance alone. Many of the diagnoses that achieved significance were not thought to be biologically plausible by the investigators.
There were six individual diagnoses that were associated with increased risk and were believed to be biologically plausible: asthma; URI; abdominal pain; musculoskeletal pain; adenitis/adenopathy; and otitis media with effusion.
We observed an increased risk of asthma in young children 18 to 35 months of age after receipt of CAIV. It is possible that the long period of replication, up to 21 days, observed by Vesikari et al. 23 could also be associated with an increased risk of reactive airway disease in this age group. This increased risk of asthma with CAIV has potential significance for the use of vaccine in young children with a history of or who are predisposed to asthma. It is possible that use of vaccine in children <3 years of age, who are more likely to be influenza-naive, could be associated with induction of wheezing. Additional studies in asthmatics and in young children to evaluate a possible risk of reactive airway disease are warranted. The CAIV vaccine has been licensed and is now available for healthy individuals 5 to 49 years of age for the 2003 to 2004 influenza virus season. Because of the increased risk for medically attended asthma visits observed in this study in young children, the manufacturer did not seek an initial indication in children <5 years of age.
There was an increased risk of abdominal pain in the vaccine group in the emergency department setting but a decreased risk in the clinic and in the combined settings. The lack of close temporal relationship between abdominal pain and vaccination also makes it unlikely that the vaccine is a cause of abdominal pain.
The increased risk observed in the vaccine group for musculoskeletal pain, adenitis/adenopathy and otitis media with effusion were thought to be biologically plausible. However, in each case it was reported in only a small number of the total analyses for each diagnosis, and the incidence of each diagnosis was low (2.13 visits/1000 person-months for adenitis/adenopathy, 8.20 visits/1000 person-months for musculoskeletal pain and 10.86 visits/1000 person-months for otitis media with effusion in the vaccine group for the analysis identified as at increased risk). The CAIV may be associated with these diagnoses, but its impact appears small.
Conjunctivitis did not appear to be at increased risk during the 42-day observation period (RR 0.65; 90% CI 0.44 to 0.96), but on evaluation of the first 14 days postvaccination it did appear to have an increased risk. This evaluation was conducted because of the close proximity of the eye to the administration site. This risk occurred for both vaccine doses and was most significant for younger children. The overall number of conjunctivitis events was small: 15 of 2716 vaccine recipients (0.55%) after the first dose in children 12 to 77 months of age; and 13 of 1244 vaccine recipients (1.0%) after the second dose in children 12 to 47 months. All cases of conjunctivitis were mild. Further virologic studies of this outcome would seem warranted.
Increased rates of runny nose/nasal congestion have been reported in other CAIV trials. 24 In our study of medically attended events, there was an increased risk of URI in only 3 of 41 analyses. URI was most frequent for 18 to 35 month olds in the combined setting for an incidence rate of 88.9 visits/1000 person-months in the vaccine group vs. 68.6 visits/1000 person-months in the placebo group. This observation in this age group is consistent with reports by Vesikari et al. 23 which demonstrated that in contrast to the very short lived replication of CAIV in older individuals and adults, viral replication may continue for 1 week or more in young children. We hypothesize that this longer replication could be associated with the URI symptoms observed.
Large studies such as this one, where medical visits recorded in health care utilization databases are the outcomes of interest, can be a very effective means of evaluating vaccine safety. The clinical significance of medical events with increased risk and biologic plausibility for association with vaccination, however, is most appropriately considered in the context of an assessment of overall risk and benefit. In addition large studies of this type serve to generate hypotheses for evaluation in subsequent follow-up studies. Influenza vaccine is the most effective strategy presently available to prevent influenza. As we try to improve our vaccine coverage of already identified high risk groups, target additional high risk groups and promote vaccination in all segments of the population, an intranasal, even potentially self-administered influenza vaccine would be of great benefit. The CAIV appears to be safe in children >3 years of age and in adolescents. The significance of a possible increased risk of reactive airway disease observed in our study awaits further investigation.
1. Glezen WP, Paredes A, Taber LH. Influenza
in children: relationship to other respiratory agents. JAMA 1980; 243: 1345–9.
2. Neuzil KM, Mellen BG, Wright PF, Mitchel EF, Griffin MR. The effects of influenza
on hospitalization, outpatient visits, and courses of antibiotics in children. N Engl J Med 2000; 342: 225–31.
3. Izurieta HS, Thompson WW, Kramarz P, et al. Influenza
and the rates of hospitalization for respiratory disease among infants and young children. N Engl J Med 2000; 342: 232–9.
4. Longini IM Jr, Koopman JS, Monto AS, Fox JP. Estimating household and community transmission parameters for influenza
. Am J Epidemiol 1982; 115: 736–51.
5. Glezen WP, Couch RB, Interpandemic influenza
in the Houston area, 1974–76. N Engl J Med 1978; 298: 587–92.
6. Fox JP, Cooney Mk, Hall CE, Foy HM. Influenzavirus infections in Seattle families, 1975–1979: II. Pattern of infection in invaded households and relation of age and prior antibody to occurrence of infection and related illness. Am J Epidemiol 1982; 116: 228–42.
7. Bridges CB, Fukuda K, Uyeki TM, Cox NJ, Singleton JA. Prevention and control of influenza
: recommendations of the Advisory Committee on Immunization Practices. MMWR Recomm Rep 2002; 51: 1–31.
8. Maassab HF, Heilman CA, Herlocher ML. Cold-adapted influenza
viruses for use as live vaccines for man. Adv Biotechnol Processes 1990; 14: 203–42.
9. Couch RB, Advances in influenza
virus vaccine research. Ann NY Acad Sci 1993; 686: 803–12.
10. Clancy RL. Mucosal vaccines for the prevention of influenza
. Drugs 1995; 50: 587–94.
11. Belshe RB A review of attenuation of influenza
viruses by genetic manipulation. Am J Respir Crit Care Med 1995; 152: S72–5.
12. 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.
13. 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: 168–75.
14. Maassab HF, DeBorde DC. Characterization of an influenza
A host range mutant. Virology 1983; 130: 342–50.
15. Maassab HF, Kendal AP, Abrams GD, Monto AS. Evaluation of a cold-recombinant influenza
virus vaccine in ferrets. J Infect Dis 1982; 146: 780–90.
16. Donabedian AM, DeBorde DC, Maassab HF. Genetics of cold-adapted B/AnnArbor/1/66 influenza
virus reassortants: the acidic polymerase (PA) protein gene confers temperature sensitivity and attenuated virulence. Microb Pathog 1987; 3: 97–108.
17. Guess HA, Lydick EG, Small RD, Miller LP. Epidemiologic programs for computers and calculators. Exact binomial confidence intervals for the relative risk in follow-up studies with sparsely stratified incidence density data. Am J Epidemiol 1987; 125: 340–7.
18. Clements ML, Makhene MK, Karron RA, et al. Effective immunization with live attenuated influenza
A virus can be achieved in early infancy. J Infect Dis 1996; 173: 44–51.
19. Edwards KM, Dupont WD, Westrich MK, et al. A randomized controlled trial of cold-adapted and inactivated vaccines for the prevention of influenza
A disease. J Infect Dis 1994; 169: 68–76.
20. Gruber WC, Belshe RB, King JC, et al. Evaluation of live attenuated influenza
vaccines in children 6–18 months of age: safety, immunogenicity, and efficacy. J Infect Dis 1996; 173: 1313–9.
21. Swierkosz EM, Newman FK, Anderson EL, Nugent SL, Mills GB, Belshe RB. Multidose, live attenuated, cold-recombinant, trivalent influenza
vaccine in infants and young children. J Infect Dis 1994; 169: 1121–4.
22. Piedra PA, Yan L, Kotloff K, Zangwill K, et al. Safety of the trivalent, cold-adapted influenza
vaccine in preschool-aged children. Pediatrics 2002; 110: 662–72.
23. Vesikari T, Karvonen A, Korhonen T, et al. A randomized, double blind, placebo-controlled trial of the safety, transmissibility and phenotype stability of a live attenuated cold-adapted influenza
virus vaccine (CAIV-T) in children attending day care [Abstract G-450]. In: 41st Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, December 16 to 19, 2001. Washington, DC: American Society for Microbiology, 2001.
24. Mendelman PM, Cordova J, Cho I. Safety, efficacy and effectiveness of the influenza
virus vaccine, trivalent, types A and B, live, cold adapted (CAIV-T) in healthy children and healthy adults. Vaccine 2001; 19: 2221–6.