Journal of Medical Virology 86:1017–1025 (2014)
Influenza Vaccine Effectiveness During the 2012 Influenza Season in Victoria, Australia: Influences of Waning Immunity and Vaccine Match Sheena G. Sullivan,1* Naomi Komadina,1 Kristina Grant,2 Lauren Jelley,1 Georgina Papadakis,2 and Heath Kelly2,3 1
WHO Collaborating Centre for Reference and Research on Influenza, Melbourne, Victoria, Australia Victorian Infectious Diseases Reference Laboratory, Melbourne, Victoria, Australia 3 National Centre for Epidemiology and Population Health, The Australian National University, Canberra, Australian Capital Territory, Australia 2
Vaccine effectiveness may wane with increasing time since vaccination. This analysis used the Victorian sentinel general practitioner (GP) network to estimate vaccine effectiveness for trivalent inactivated vaccines in the 2012 season. A test-negative design was used where patients presenting to GPs with influenza-like illness who tested positive for influenza were cases and noncases were those who tested negative. Vaccination status was recorded by GPs. Vaccine effectiveness was calculated as (1-odds ratio) 100%. Estimates were compared early versus late in the season and by time since vaccination. Virus isolates were assessed antigenically by hemagglutination inhibition assay in a selection of positive samples and viruses from healthy adults who experienced a vaccine breakthrough were analyzed genetically. The adjusted vaccine effectiveness estimate for any type of influenza was 45% (95% CI: 8,66) and for influenza A(H3) was 35% (95% CI: 11,62). A non-significant effect of waning effectiveness by time since vaccination was observed for A(H3). For those vaccinated 37 [WHO Collaborating Centre for influenza at CDC Atlanta, 2009] that failed to grow in culture were considered negative. Samples that did not register a CT within 45 amplification cycles or did not have an exponential amplification profile were considered to be negative for influenza. Influenza-positive specimens were forwarded to the WHO Collaborating Centre for Reference and Research on Influenza in Melbourne where they were further characterized to identify the virus strain using the hemagglutination inhibition (HI) assay [Hobson et al., 1972]. Virus isolates were first obtained by inoculating original clinical samples into MDCK (Madin-Darby Canine Kidney) cells. HI assays were performed as previously described [Hobson et al., 1972]. Isolates were identified as antigenically similar to the reference strain if the test samples had a titre that was less than an 8-fold difference compared with the homologous reference strain. Results were reported against reference sera for A/California/07/2009(H1N1pdm09)-like, A/ Perth/16/2009(H3N2)-like, B/Wisconsin/01/2010-like (Yamagata lineage), and B/Brisbane/60/2008-like (Victoria lineage) viruses. A(H3) viruses obtained from healthy (i.e., no predisposing condition for severe influenza), vaccinated adults were selected for sequencing of the HA gene to identify mutations potentially contributing to vaccine failures. For comparison, healthy unvaccinated adults matched as best as possible for age (within 5 years) date of symptom onset (within 1 week) and gender were also selected. RNA was extracted from the original clinical specimens using the QIAGEN Xtractor Gene robot, followed by reverse transcription PCR using the BIOLINE MYTAQ one step reverse transcription PCR kit according to the manufacturer’s recommendations and gene specific HA primers (primer sequences available on request: whoflu@ influenzacentre.org). Conventional Sanger sequencing was carried out on PCR product using an Applied
Influenza vaccine Effectiveness in Victoria, 2012
Biosystems 3500 XL sequencer, and sequence assembly performed using the Seqman Pro Module of DNASTAR Lasergene version 9.1.0 software. (DNAStar, Madison, WI). Phylogenetic analysis was carried out using MEGA 5 [Tamura et al., 2011]. For GISAID Accession numbers refer to Supplementary material. Statistical Analysis
Negative Positive
0
10
Number of consultations 20 30
40
The analysis was limited to the fortnights when both cases (influenza A or B) and noncases were identified (see Fig. 1). Exploratory analyses, including differences between those who tested positive or negative for influenza, and between the vaccinated and unvaccinated, were compared by odds ratio and chi-square test for categorical variables and t-test for continuous variables. All P-values were two-sided. Crude odds ratios were estimated for each influenza type/subtype. Adjusted estimates were calculated using a similar model as used in Europe [Kissling et al., 2012], which included age group (10-year bands), gender, month of onset and the presence of any predisposing conditions (yes/no). Information on hospitalizations or GP visits was not collected so could not be included in the models. Because some observations were missing information on some variables, they were dropped from the complete case analysis. However in order not to lose their information, missing values were imputed by multiple imputation using chained equations. Details of this type of imputation procedure are explained, step-by-step, in Azur et al. [2011]. Ten imputed datasets were generated, and imputations were generated using all the variables in the dataset, including the outcome. Results are reported for both the complete case analysis and the imputed analysis. Additionally, sensitivity analyses were performed for A(H3): (1) comparing the results obtained using
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
Week
Fig. 1. Consultations by influenza status. All data shown, including excluded patients. Week 18 was the week beginning 30 April 2012, Week 44 was the week ending November 2, 2012. Only Weeks 18–42 were used for overall vaccine effectiveness estimates; Weeks 18–39 for A(H3).
1019
I-MOVE’s logistic model with the model used to evaluate Victoria’s data in 2011 [Fielding et al., 2012]; (2) limiting patients to those in a target group for vaccination (defined as those eligible for free vaccination under the government’s programme; i.e. health care workers, people with predisposing conditions to severe influenza, those aged 65 years or older [Australian Government Department of Health and Ageing, 2012]); (3) limiting the analysis to working-age adults (20–64 years), who tend to be overrepresented in this surveillance system [Kelly et al., 2013]; (4) conducting separate estimates for the first half of the surveillance period (Weeks 18–30) versus the latter half (Weeks 31–44); and 5) vaccination categorised by time since vaccination (
Influenza Vaccine Effectiveness During the 2012 Influenza Season in Victoria, Australia: Influences of Waning Immunity and Vaccine Match Sheena G. Sullivan,1* Naomi Komadina,1 Kristina Grant,2 Lauren Jelley,1 Georgina Papadakis,2 and Heath Kelly2,3 1
WHO Collaborating Centre for Reference and Research on Influenza, Melbourne, Victoria, Australia Victorian Infectious Diseases Reference Laboratory, Melbourne, Victoria, Australia 3 National Centre for Epidemiology and Population Health, The Australian National University, Canberra, Australian Capital Territory, Australia 2
Vaccine effectiveness may wane with increasing time since vaccination. This analysis used the Victorian sentinel general practitioner (GP) network to estimate vaccine effectiveness for trivalent inactivated vaccines in the 2012 season. A test-negative design was used where patients presenting to GPs with influenza-like illness who tested positive for influenza were cases and noncases were those who tested negative. Vaccination status was recorded by GPs. Vaccine effectiveness was calculated as (1-odds ratio) 100%. Estimates were compared early versus late in the season and by time since vaccination. Virus isolates were assessed antigenically by hemagglutination inhibition assay in a selection of positive samples and viruses from healthy adults who experienced a vaccine breakthrough were analyzed genetically. The adjusted vaccine effectiveness estimate for any type of influenza was 45% (95% CI: 8,66) and for influenza A(H3) was 35% (95% CI: 11,62). A non-significant effect of waning effectiveness by time since vaccination was observed for A(H3). For those vaccinated 37 [WHO Collaborating Centre for influenza at CDC Atlanta, 2009] that failed to grow in culture were considered negative. Samples that did not register a CT within 45 amplification cycles or did not have an exponential amplification profile were considered to be negative for influenza. Influenza-positive specimens were forwarded to the WHO Collaborating Centre for Reference and Research on Influenza in Melbourne where they were further characterized to identify the virus strain using the hemagglutination inhibition (HI) assay [Hobson et al., 1972]. Virus isolates were first obtained by inoculating original clinical samples into MDCK (Madin-Darby Canine Kidney) cells. HI assays were performed as previously described [Hobson et al., 1972]. Isolates were identified as antigenically similar to the reference strain if the test samples had a titre that was less than an 8-fold difference compared with the homologous reference strain. Results were reported against reference sera for A/California/07/2009(H1N1pdm09)-like, A/ Perth/16/2009(H3N2)-like, B/Wisconsin/01/2010-like (Yamagata lineage), and B/Brisbane/60/2008-like (Victoria lineage) viruses. A(H3) viruses obtained from healthy (i.e., no predisposing condition for severe influenza), vaccinated adults were selected for sequencing of the HA gene to identify mutations potentially contributing to vaccine failures. For comparison, healthy unvaccinated adults matched as best as possible for age (within 5 years) date of symptom onset (within 1 week) and gender were also selected. RNA was extracted from the original clinical specimens using the QIAGEN Xtractor Gene robot, followed by reverse transcription PCR using the BIOLINE MYTAQ one step reverse transcription PCR kit according to the manufacturer’s recommendations and gene specific HA primers (primer sequences available on request: whoflu@ influenzacentre.org). Conventional Sanger sequencing was carried out on PCR product using an Applied
Influenza vaccine Effectiveness in Victoria, 2012
Biosystems 3500 XL sequencer, and sequence assembly performed using the Seqman Pro Module of DNASTAR Lasergene version 9.1.0 software. (DNAStar, Madison, WI). Phylogenetic analysis was carried out using MEGA 5 [Tamura et al., 2011]. For GISAID Accession numbers refer to Supplementary material. Statistical Analysis
Negative Positive
0
10
Number of consultations 20 30
40
The analysis was limited to the fortnights when both cases (influenza A or B) and noncases were identified (see Fig. 1). Exploratory analyses, including differences between those who tested positive or negative for influenza, and between the vaccinated and unvaccinated, were compared by odds ratio and chi-square test for categorical variables and t-test for continuous variables. All P-values were two-sided. Crude odds ratios were estimated for each influenza type/subtype. Adjusted estimates were calculated using a similar model as used in Europe [Kissling et al., 2012], which included age group (10-year bands), gender, month of onset and the presence of any predisposing conditions (yes/no). Information on hospitalizations or GP visits was not collected so could not be included in the models. Because some observations were missing information on some variables, they were dropped from the complete case analysis. However in order not to lose their information, missing values were imputed by multiple imputation using chained equations. Details of this type of imputation procedure are explained, step-by-step, in Azur et al. [2011]. Ten imputed datasets were generated, and imputations were generated using all the variables in the dataset, including the outcome. Results are reported for both the complete case analysis and the imputed analysis. Additionally, sensitivity analyses were performed for A(H3): (1) comparing the results obtained using
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
Week
Fig. 1. Consultations by influenza status. All data shown, including excluded patients. Week 18 was the week beginning 30 April 2012, Week 44 was the week ending November 2, 2012. Only Weeks 18–42 were used for overall vaccine effectiveness estimates; Weeks 18–39 for A(H3).
1019
I-MOVE’s logistic model with the model used to evaluate Victoria’s data in 2011 [Fielding et al., 2012]; (2) limiting patients to those in a target group for vaccination (defined as those eligible for free vaccination under the government’s programme; i.e. health care workers, people with predisposing conditions to severe influenza, those aged 65 years or older [Australian Government Department of Health and Ageing, 2012]); (3) limiting the analysis to working-age adults (20–64 years), who tend to be overrepresented in this surveillance system [Kelly et al., 2013]; (4) conducting separate estimates for the first half of the surveillance period (Weeks 18–30) versus the latter half (Weeks 31–44); and 5) vaccination categorised by time since vaccination (
