Influenza Vaccine Effectiveness: Considerations for Prioritization of Pandemic Influenza Vaccine

1
Influenza Vaccine Effectiveness Considerations
for Prioritization of Pandemic Influenza Vaccine

• Carolyn B. Bridges, MD
• National immunization Program, CDC
• for
• NVAC/ACIP Influenza Vaccine Working Group
• April 20, 2005

2
Influenza Vaccine Effectiveness (VE)

• Reviewed by ACIP Influenza Working Group January
  2005 meeting on inter-pandemic prioritization
• Kristin Nichol
• John Treanor
• Wendy Keitel
• Lone Simonsen
• Litjen Tan
• Niranjan Bhat
• Guillermo Herrera
• Raymond Strikas

3
Outline

• VE TIV during inter-pandemic years
• By age group and chronic condition
• VE LAIV
• 1 versus 2 doses in immunologically naïve
  populations
• TIV in children
• LAIV in children
• Swine flu and Russian flu H1N1 vaccines
• Studies of H5 vaccines

4
Vaccine Effectiveness

• Varies by age group, risk group, and antigenic
  match
• Variety of outcomes/methods in literature
• Influenza-like illness
• Laboratory-confirmed influenza
• Influenza-related hospitalization and death
  largely based on modeling
• Herd immunity effects may also be considered

5
Healthy Adults lt 65 Yrs

• Key literature reviewed
• US military trials
• Cochrane review (updated 2004)
• 8 clinical trial papers published since 1988

6

• Efficacy of Influenza Vaccine in Healthy Young
  Adults

US Army Field Trials US Air Force
Field Trials
Adapted from Davenport, Med J Aust 1973
(suppl) 33-8. Adapted from Meiklejohn. J
Infect Dis 1983 148 775-84.
7
VE Healthy Adults

• VE estimates (Cochrane review for IV)
• Lab-C illness 70 (56 - 80)
• Clinical ILI 25 (13 - 35)
• Work loss reduction
• 0.16 days per person vaccinated (0.04 0.29)
• Other studies generally similar
• Vaccination also associated with reductions in
  health care provider visits antibiotic use due
  to URI / ILI
• Insufficient data on serious complications

8
VE During Pregnancy (1 cohort study)

• Black SB, Am J Perinatology 2004
• Subjects 49,585 women with live births Nov thru
  Feb 97-98 thru 00-01 (KPNC)
• Results
• Hospitalizations very rare
• Outpt visits for resp illness HR 1.15 (p .09)
• Note vaccination rates low (4.7 - 11.9)

9
Community Dwelling Elderly

• 4 clinical trials published from 1994 on
• Numerous observational studies
• 1 meta analysis

10
VE in Community Dwelling Elderly Persons (meta
analysis)
Vu T, et al. Vaccine 2002 20 1831.
11
VE by Risk Status
12
VE Elderly in Nursing Homes

• 1 meta analysis
• Several observational studies

13
VE Nursing Home Residents
14
VE Poor Match Yrs
15
Summary of VE in Adults

• Healthy adults
• Vaccination reduces illness / work loss
• Elderly
• Vaccination reduces illness serious
  complications of influenza
• Vaccination provides benefits for healthy high
  risk elderly for community dwelling NH
  residents
• VE with mismatch is variably reduced
• Even with lower VE, NNT must be considered

16
Influenza VE Studies in Children

• Data more limited compared with adults
• Hoberman, et al JAMA 2003
• 2 dose TIV vs placebo among children 6-24 months
• 66 VE in year 1 and 0 in year 2 with low
  incidence influenza
• No VE versus otitis media
• Good immune responses to vaccine

17
Pediatric VE Summary TIV

• Influenza vaccine is efficacious in children
• 21-76 for ILI
• 30-95 for lab-confirmed influenza
• 32-36 for otitis media
• Generally similar results for healthy and high
  risk

18
Pediatric VE Summary, cont

• Vaccine efficacy in children increases with age
• Limited data in children aged 6-23 months

19
LAIV vs. TIV
20
Live and inactivated vaccines

• Theoretical considerations
• Live vaccines must replicate
• Level of replication depends on the host
• Children gt adults gt elderly
• Live vaccine stimulate mucosal immunity
• May be more effective at limiting shedding
• No well standardized immune correlates

21
Live and inactivated vaccines

• Theoretical considerations
• Inactivated vaccines do not replicate
• Level of immunity depends on host priming
• Adults gt children gt elderly
• Inactivated vaccines stimulate serum antibody
• Well standardized immune correlate

22
Live and inactivated vaccines

• There are few direct comparisons
• Indirect comparisons can be difficult to
  interpret
• Randomized direct comparisons
• Pediatric
• Adult
• Elderly
• More definitive comparisons in adult and
  pediatric populations are underway

23
Pooled results of experimental infection studies
in adults Vaccine 18899 (2000)
Virus shedding
Infection
Influenza illness
0.36
0.14
0.18
TIV
0.64
0.35
0.10
CAIV
0
1
0
1
0
1
Pooled Odds Ratio (95 CI) compared to placebo
24
Natural infection in adults

• Edwards (1994) J Infect Dis 16968-76
• Multiple years, subjects remain in group
• Control vaccines monovalent B/allantoic fluid
• Children did not receive 2-dose schedule
• LAIV given by drops
• Outcomes included ILI, serologic and
  culture-confirmed illness

25
Evaluation of the protective efficacy of LAIV in
adults
Rate per 1,000 subjects
H1N1
H3N2
26
Pediatric subgroup analysisNeuzil (2001) Pediatr
Infect Dis J 20733

• Analysis restricted to children younger than 16
  at the time of immunization
• 474 age 1 to 5
• 744 age 6 to 10
• 591 age 11 to 15
• Two outcomes
• Culture positive illness
• Seroconversion

27
Evaluation of the protective efficacy of LAIV in
kids
Per 1,000
Per 100
Rate per 1,000 or per 100 subjects
Cx positive
Ab positive
28
Elderly

• LAIV is not infectious less than 10 have
  detectable shedding by culture
• Low antibody response rates, even in subjects
  with low prevaccination antibody
• Combined vaccine may provide additional
  protection
• Nursing home yes
• COPD - no

29
One Versus 2 Doses In Naïve Populations
30
1976 Swine Flu and 1977 Russian Flu Vaccine
Trials

• For persons born before H1N1 viruses last
  circulated (circa 1957)
• 2 doses needed for best antibody response
• 7 µg doses gave comparable responses with 2
  doses
• Shallow doses response with 1 dose with gt50 µg
  needed
• Whole virus vaccine more immunogenic, but more
  reactogenic at high doses

31
VE of LAIV in children 15-71 monthsBelshe RB, et
al. JAMA 19983381405-12

• Trivalent LAIV versus placebo
• N532 received placebo
• N1070 received 1 or 2 doses
• VE against culture-confirmed influenza
• 89 with 1 dose
• 94 with 2 doses

32
VE of TIV with 1 versus 2 dosesRitzwoller DR.
Pediatrics 2005 (in press)

• Retrospective cohort study children 6-23 months
  enrolled Kaiser Colorado
• 2003-04 when suboptimal antigenic match
• gt5000 in cohort
• Controlled for high risk conditions using
  administrative data
• No laboratory confirmation
• VE NS with 1 dose, 25 for ILI and 49 for PI
  with 2 doses

33
LAIV vs Inactivated Vaccine

• One dose alone of inactivated vaccine in
  immunologically naïve persons
• Less likely to provide protective immune response
  unless use high doses
• Not protective in young children
• 2 doses inactivated vaccine likely to provide
  protective immune response at lower antigenic
  content
• LAIV may provide better protection with 1 dose
• Immune-correlate less well defined, so assessment
  of probable efficacy based on immune response
  difficult

34
H5 vaccines
35
Response to Recombinant H5 VaccineTreanor JJ, et
al Vaccine 2001191732-7.

• Placebo-controlled trial
• 2 doses at 21, 28 or 42 day intervals
• 25, 45, or 90µg x 2 doses or 90 then 10µg
• Serum collected days 0, 14 days after 1st dose,
  dose 2 day 0, the 1,2,3,4 weeks after dose 2
• 21-45 with ELISA immune response
• 17-52 with micro-neutralization response
• Dose-response, highest at 90 µg x 2
• No significant effect by dosing interval

36
Use of adjuvant MF59-H5N3 vaccineStephenson I,
et al JID 20051911201-5and Stephenson I, et al
Vaccine 2003211687-93

• Adults 18-45 yrs given 7.5, 15 or 30µg
  A/duck/Singapore/97 (H5N3)
• IM injection with and without MF-59
• 2 doses, 21 days apart
• 7-14 with MF-59 versus 0-9 without
• 3rd dose 16 months later to subset
• Serum tested by micro-neutralization against HPAI
  H5N1 1997-2004 strains
• No dose response detected
• Seroconversion 43-100 with MF-59 and 0-27
  without
• No 3rd dose booster effect with non-MF-59
• Conclusions
• 3 doses and adjuvant needed to improve response
• No difference among doses used, but small numbers

37
Overall Summary

• Responses to inter-pandemic influenza vaccines
• Varies by age and chronic condition
• Within an age group, generally higher VE against
  complications than influenza illness
• LAIV and TIV options for children and adults lt65
• Immunologically naïve persons need 2 doses to
  reach protective immune response for
  inactivated vaccines
• May be able to achieve with high single dose
• May need only 1 dose for live attenuated
  vaccines?
• Testing of H5 needed to assess safety and
  immunogenicity
• Clinical studies H5 vaccines to date suggest
  lower immunogenicity without adjuvant
• Further trials pending