Herd Protection against Influenza W P Glezen, P A Piedra, M J Gaglani

Houston, Texas and SCOTT & WHITE Clinic Temple, Texas

Herd Protection against Influenza

“…it is apparent that progress in the control of influenza has not been impressive.

A reassessment of the basic assumptions upon which the program was developed is warranted.”

A. D. Langmuir et al 1964; Am J Public Health 54:563-71

Annual Rates of Influenza-Associated Hospitalizations, U.S. † H3 Other

250 200 182.9

195.7

183.9

196.6

150 100 66.7

82.4

50 37.3

105 96 139.8

88.2

84.7

110.5

111.4

87.8

128.2

118.6

124.3

109.9

111.8

50.1

112.9

0 1979 1980 1980 1981 1981 1982 1982 1983 1983 1984 1984 1985 1985 1986 1986 1987 1987 1988 1988 1989 1989 1990

Year

1990 1991 Thompson etal JAMA September 15, 2004;292:11:1337 1991 1992 1992 1993 1993 1994 1994 1995 1995 1996 1996 1997 1997 1998 1998 1999 1999 2000 2000 2001

Problems With Targeting High Risk Patients

 

High risk patients are not easily accessible for vaccination Many high risk patients are debilitated or immunocompromized and fail to respond optimally to vaccine

Risk Based Strategies Have Failed

Universal Recommendations Generally Are More Successful e.g.: > 65 yr, 6-23 mo., Ontario program The most vulnerable persons – elderly and infants – have poor immune responses to vaccines and are at the end of the transmission chain = inefficient use of vaccine.

School Children, Preschool children and Working adults have the:

   

highest attack rates for influenza, are the spreaders in the community and the introducers into the household.

They also are most accessible for rapid deployment of vaccine.

Shift in Age Distribution of Persons with Culture Positive Illness Presenting to Sentinel Clinics during Influenza Epidemics, Houston, 1974-1981 Age (yr) Early (%) < 5 5-19 > 20 236(18.4) 687(53.6) 359(28.0) Total 1,282 Epidemic stage Peak (%) 489(24.3) 741(36.8) 785(39.0) 2,015 Late (%) 248(24.5) 356(35.2) 407(40.3) 1,011

60 50 40

Epidemic Period

30 20 10 0 60 50 40 30 20 10 0

Epidemic Period

Early Early

After School Holiday

Peak Late

Interrupted by School Holiday

Peak Late

Years

<5 5-19 >19 <5 5-19 >19

Influenza vaccines generate optimal immune responses in healthy schoolchildren and working adults.

Immunization of these groups has the potential for establishing indirect protection of the vulnerable [HERD IMMUNITY or HERD PROTECTION] = efficient use of influenza vaccine.

Examples of Herd Protection by Currently Used Vaccines

• •

Rubella vaccine – infant immunization protects pregnant women Hemophilus influenzae type b (Hib) vaccines – in The Gambia (Adegbola et al. Lancet 2005;366:144)

• •

Pneumococcal conjugate vaccine (7-valent) – in infants (Poehling et al. JAMA 2006;295:1668; in adults (Metlay et al. Vaccine 2006; 24:468; Hammitt et al. JID 2006;193: 1487; Flannery et al. Ann Int Med 2006;144: 1-9.) Hepatitis A vaccine – in US adults (Wasley et al. JAMA 2005;294:194; in Israel (Dagan et al 2005;294:202.).

Herd Protection by Influenza Vaccines

• • • • •

Tecumseh MI study (Monto et al. Bull WHO 1969;41:537) children. – 67% reduction in adult illness rates by single dose of TIV in school Northern Territory, Australia (Warburton et al. Med J Aust 1972;2:67) – reduction in attack rate in communities with variable vaccine coverage compared to those with no vaccine. Novgorod, Russia schoolchildren study (Rudenko et al. J Infect Dis 1993;168:881) – reduction in attack rate in staff where LAIV given to students.

San Diego (Hurwitz et al. JAMA 2000;284:1677) – TIV for daycare toddlers reduced ILI in older siblings and parents. Moscow, Russia (Ghendon et al. Epidemiol Infect 2006;134:71) TIV in 57% of preschool and 72% of school children reduced illness and complications in unvaccinated, non institutionalized adults.

Herd Protection – Proof of Concept

Japanese School Children Program (Reichert et al. N Engl J Med 2001;344:889) In the decade, 1977-1987, influenza vaccine was mandatory for school attendance. Two doses of inactivated vaccine were recommended each year. Influenza was not recommended for elderly or high risk patients.

The Japanese policy makers were unaware of the indirect effectiveness until Reichert analyzed wintertime excess mortality from 1959-1998.

Coincident with the school program, he found that influenza related excess mortality was reduced by 35,000 to 47,000 lives per year. When the program was discontinued, excess mortality rose to pre-program levels.

Herd Protection – Proof of Concept Continued

Reichert also showed that the summertime baseline mortality trends for Japan were very different than the wintertime excess mortality, separating the school influenza program from the economic recovery of Japan after WWII. (Reichert TA. Seminars Pediatr Infect Dis 2002;13:104)

.

Control of Epidemic Influenza: Study Design

• An open-label, non-randomized, community-based trial of annual influenza immunization of school-age children to effect herd immunity

MAARI Rates in the Intervention and Comparison Sites during Influenza Outbreaks for SWHP Members > 35 years old MAARI rates per 100 person-season Overall effectiveness (1-RR) Year Intervention site: T-B Comparison site: W/B-CS 95% CI 1997-98; baseline 9.3

9.2

0 -0.03-0.05

1998-99; Year 1 12.0

13.1

0.08

0.04-0.13

1999-00; Year 2 2000-01; Year 3 12.2

15.1

15.0

17.8

0.18

0.15

0.14-0.22

0.12-0.19

Piedra et al: Vaccine 2005;23:1540-8

CAIV-T FIELD TRIAL Summary 1.

2.

3.

Safe-side effects do not increase direct medical costs.

Direct Effectiveness a. Protection inversely related to age (VEadj 0.70-0.91) b. Persists through two seasons c. Heterovariant d. Single dose is sufficient Indirect Effectiveness (Herd Immunity) – For proportion vaccinated compatible with Longini Model.