Journal club summaries - 2015


A 9-valent HPV vaccine against infection and intraepithelial neoplasia in women

Joura EA, Giuliano AR, Iversen OE, et al.
New England Journal of Medicine 2015;372(8):711-23
Link to abstract  

Human papillomavirus (HPV) infection is extremely common, infecting about 80% of the population at some point in their lives. Globally HPV types 16 and 18 are responsible for most cervical cancers and HPV types 6 and 11 have been associated with 90% of cases of genital warts. Other important HPV types associated with cervical disease include types 31, 33, 45, 52 and 58. There are currently two HPV vaccines, the bivalent vaccine (2vHPV, HPV 16 and 18) and the quadrivalent vaccine (4vHPV, HPV 16, 18, 6 and 11). In clinical trials, both vaccines were efficacious against infection and cervical dysplasia associated with HPV types 16 and 18, and the quadrivalent vaccine was efficacious against genital warts related to HPV types 6 and 11. Post-licensure reports from countries with established HPV vaccination programs indicate a beneficial effect at the population level, including decreases in the incidence of high-grade cervical abnormalities and the incidence of genital warts.

This publication presents results of a study of the efficacy and immunogenicity of a 9-valent HPV (9vHPV) vaccine (HPV types 6, 11, 16, 18, 31, 33, 45, 52 and 58) in women aged 16–26 years. An international randomised clinical trial of the 9vHPV vaccine was performed in 14,215 women. Participants received the 9vHPV vaccine or the 4vHPV vaccine in a series of three injections on day 1 and at months 2 and 6. Serum was collected for analysis of antibody responses and swabs of anogenital tissues were obtained regularly through the study and used for HPV DNA testing. The rate of high-grade cervical, vulvar or vaginal disease related to the additional HPV types in the 9vHPV vaccine (31, 33, 45, 52 and 58) was 0.1 per 1000 person-years in the 9vHPV group and 1.6 per 1000 person-years in the 4vHPV group (efficacy of the 9vHPV vaccine, 96.7%; 95% CI 80.9 to 99.8%). Antibody responses to HPV 6, 11, 16 and 18 were non-inferior to those generated by the 4vHPV vaccine. 

Both the bivalent and quadrivalent HPV vaccines are currently registered in Australia. Australia introduced funded HPV vaccination in 2007 (using the quadrivalent vaccine) for girls aged 12–13 years with catch-up to 26 years of age. The program was expanded to adolescent boys in 2013. The 9vHPV vaccine will likely be registered in Australia shortly and could be used to provide additional protection for HPV types not in the currently available vaccines. 

Presented by Dr Helen Quinn, Senior Research Fellow, NCIRS

Assessment of herd immunity and cross-protection after a human papillomavirus vaccination programme in Australia: a repeat cross-sectional study

Tabrizi SN, Brotherton JM, Kaldor JM, et al.
Lancet Infectious Diseases 2014;14(10):958–66
Link to abstract  

This repeat cross-sectional study measured the prevalence of vaccine-targeted and related HPV types for assessing protection (direct and cross-protection) as well as any herd immunity effect after the introduction of the human papillomavirus (HPV) vaccination program in Australia. 

The pre-vaccine sample (n=202) included women aged 18–24 years who attended Pap screening between October 2005 and July 2007, in three major metropolitan areas of Australia (Sydney, Melbourne and Perth). The post-vaccine sample (n=1058) included women in the same age group from the catchment areas from August 2010 to November 2012. Prevalence was measured in the vaccine-targeted HPV genotypes (6, 11, 16 or 18) and also in any high-risk HPV genotype (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68); any high-risk HPV genotype excluding the vaccine genotypes 16 and 18; HPV types genetically related to the vaccine-targeted HPV types (i.e. 45 [related to 18], and 31 and 33 [related to 16]) and prevalence individually for HPV types 52 and 58, which are less closely related to HPV 16. 

Prevalence of vaccine-targeted HPV genotypes was significantly lower in the post-vaccine than in the pre-vaccine sample but there was no significant declines observed for non-vaccine targeted HPV genotypes. In addition, a lower prevalence of vaccine-targeted types was seen in unvaccinated women, suggesting herd immunity. In vaccinated women, a likely cross-protective effect against HPV types related to the vaccine-targeted types was also observed. 

In Australia, this study has implications for use of the 9-valent human papillomavirus (HPV) vaccine (9vHPV) (Gardasil 9, Merck and Co. Inc.) that protects against HPV types 6, 11, 16, 18, 31, 33, 45, 52 and 58. In the USA, the FDA approved the use of Gardasil 9 in December 2014 and the Advisory Committee on Immunization Practices (ACIP) in its February meeting recommended Gardasil 9 as one of three HPV vaccines for routine vaccination in the USA. In Australia, the recommendation for use of Gardasil 9 is currently being considered by the Australian Technical Advisory Group on Immunisation. We await future news in this area.  

Presented by Dr Aditi Dey, Manager Surveillance, NCIRS

The Vaccination Confidence Scale: a brief measure of parents’ vaccination beliefs

Gilkey MB, Magnus BE, Reiter PL, et al.
Vaccine 2014;32:6259-65
Link to abstract

This US-based study used national population-based telephone survey data to develop a brief vaccine confidence scale for parents of adolescents. Although much work has been done in this area to measure confidence in parents of younger children, there are fewer tools and datasets relating to parents of adolescents. 

Eleven beliefs were measured using an 11-point scale (0 = “strongly disagree” to 10 = “strongly agree”). These items were developed based on the Health Belief Model, and responses were subject to a two-phase factor analysis, beginning with an exploratory factor analysis using a randomised sample of half the data. In this phase, three items were dropped as they did not meaningfully load into any factor. A confirmatory factor analysis was then undertaken on the remaining 8-item set, using a model with one factor, and another with three, “Trust”, “Benefits” and “Harm”. 

While the one-factor model showed a “marginally acceptable” fit with the data, the three-factor model gave “good” fit. 

The resulting 8-item Vaccination Confidence Scale offers a reasonable measure of parental adolescent vaccination beliefs, and when divided into three factors fit the data best while aligning well with Health Belief Model constructs. 

There is potential to test these items in the Australian context, not only with parents of adolescents, but with other groups in the population. It was shown with the American sample to be consistent across a number of demographic groups, and has the potential to be used to routinely measure vaccine confidence over time. The information garnered could inform targeted intervention strategies. 

Presented by Ms Kerrie Wiley, Research Fellow – Social Research, NCIRS

Impact of repeated vaccination on vaccine effectiveness against influenza A(H3N2) and B during 8 seasons

McLean HQ, Thompson MG, Sundaram ME, et al.
Clinical Infectious Diseases 2014;59:1375-85
Link to abstract 

Recent studies of influenza vaccine effectiveness (VE) have suggested that a person’s previous season influenza vaccination status may have some impact on the current season VE. There have been some studies which have suggested a decrease in VE if previously vaccinated. These studies have often been limited to 1 or 2 seasons and have had inconsistent results. This study aimed to explore the effect of interactions between different combinations of previous season and current season vaccination status as well as current season and 5-year vaccination history, across a single population over eight seasons. 

The authors performed vaccine effectiveness calculations, using a test-negative design, on 7315 subjects ≥9 years old who presented for acute respiratory illness and had influenza testing. One model investigated differences in VE in subjects vaccinated in the current season only, in current/previous season, in a previous season only, and those not vaccinated. A second model compared interactions between current season vaccinated or not vaccinated, each subdivided according to 5-year previous vaccination history into “frequent vaccinee”, “infrequent vaccinee”, and “not vaccinated” groups. 

The authors found that for both A/H3N2 and B influenza, vaccination in the current season only, in current/previous season, or in previous season only was effective and had similar VE, compared to the reference non-vaccinated group. However, in the second model, they found that subjects who were unvaccinated during the previous 5 years had a significantly higher VE with current season vaccination than subjects who were frequently vaccinated during the previous 5 years.

Vaccine interference may be an explanation for the observed effect. The authors discussed the phenomenon of “original antigenic sin”, whereby on meeting a new variant influenza strain, there may be a larger immune response to historical strains than the current de novo strain. Unmeasured confounding must be borne in mind as potentially contributing to the effects. No data was provided on whether vaccine and circulating strains matched. In the two-season analysis, current and previous season vaccination were similarly effective with no detriment from previous season vaccination. More research examining vaccine interference in randomised controlled trials are needed.

Presented by Dr Jean Li-Kim-Moy, Research Fellow – Clinical Research, NCIRS