Prophylactic vaccination against human papillomavirus infection and disease in women: a systematic review of randomized controlled trials.(Research)(Clinical report)

[infinity] See related articles pages 456, 462, 464, 480 and 484

ABSTRACT

Background: Human papillomavirus (HPV) is now known to be a necessary cause of cervical cancer, and prophylactic HPV vaccines aimed at preventing genital warts, precancerous cervical lesions and cervical cancer are now available. To gauge the potential impact on disease burden, we performed a systematic review of the evidence from randomized controlled trials.

Methods: We conducted a systematic search of the literature to identify all randomized controlled trials of prophylactic HPV vaccination. Reports in 5 electronic databases covering 1950 to June 2007 (MEDLINE, MEDLINE in process, EMBASE, the Cochrane Central Registry of Controlled Trials and the Cochrane Library), bibliographies of all included studies and of narrative reviews (2006-2007), clinical trial registries, Google Scholar, public health announcements, selected conference proceedings (2004-2007) and manufacturers' information on unpublished data or ongoing trials were screened against predefined eligibility criteria by 2 independent reviewers. Vaccines had to contain coverage against at least 1 oncogenic HPV strain. The primary outcome of interest was the frequency of high-grade cervical lesions (high-grade squamous intraepithelial lesion, or grade 2 or 3 cervical intraepithelial neoplasia). The secondary outcomes were persistent HPV infection, low-grade cervical lesions (low-grade squamous intraepithelial lesion or grade 1 cervical intraepithelial neoplasia), external genital lesions, adverse events and death. Meta-analysis of the data was done in all cases where adequate clinical and methodological homogeneity existed.

Results: Of 456 screened reports, 9 were included in the review (6 were reports of randomized controlled trials and 3 were follow-up reports of initial trials). Findings from the meta-analysis showed that prophylactic HPV vaccination was associated with a reduction in the frequency of high-grade cervical lesions caused by vaccine-type HPV strains compared with control groups: Peto odds ratio 0.14 (95% confidence interval [CI] 0.09-0.21) from combined per-protocol analyses, and 0.52 (95% CI 0.43-0.63) from modified intention-to-treat analyses. Vaccination was also highly efficacious in preventing other HPV-related infection and disease outcomes, including persistent HPV infection, low-grade lesions and genital warts. The majority of adverse events were minor. The incidence of serious adverse events and death were balanced between the vaccine and control groups.

Interpretation: Among women aged 15-25 years not previously infected with vaccine-type HPV strains, prophylactic HPV vaccination appears to be highly efficacious in preventing HPV infection and precancerous cervical disease. Long-term follow-up is needed to substantiate reductions in cervical cancer incidence and mortality.

Une version francaise de ce resume est disponible a l'adresse www.cmaj.ca/cgi/content/full/177/5/469/DC1

CMAJ2007;177(5):469-79

Published at www.cmaj.ca on Aug. 1, 2007.

Cervical cancer is an important cause of preventable cancer-related death among women. Because of the overwhelming burden of this disease in developing countries, cervical cancer is the second most common cause of cancer among women worldwide. (1) It primarily affects women between 30 and 45 years of age, thereby representing in an important source of potential years of life lost.

With the obligate link between HPV and cervical cancer now established, prophylactic HPV vaccination represents a potential means of reducing the burden of cervical cancer and its precursor lesions. Two prophylactic HPV vaccines are now available. Both target HPV types 16 and 18, and one of the vaccines also targets HPV types 6 and 11. There are over 100 known subtypes of HPV, but types 16 and 18 are the most prevalent oncogenic strains of the virus, accounting for an estimated 70% of cervical cancers worldwide. (1,2) Non-oncogenic strains, such as HPV types 6 and 11, are associated with the development of external genital disease, including genital warts. Most sexually active women will become infected with HPV in their lifetime, and over 50% of girls will acquire HPV within 48 months of becoming sexually active. (3) Infection with an oncogenic strain of HPV does not guarantee that cervical cancer or HPV-related disease will develop. HPV-associated precancerous changes in the cervix either resolve spontaneously or may be identified through screening surveillance and treated.

Although 2 combined analyses of randomized trials of the quadrivalent vaccine were recently published, (4,5) in this systematic review we summarize the body of existing evidence from randomized trials regarding the value of prophylactic vaccination against HPV as assessed by surrogate outcomes related to persistent HPV infection and precancerous lesions that lead to cervical cancer in women. We explicitly sought to determine whether women who receive prophylactic HPV vaccination have a lower incidence of persistent HPV infection and precancerous cervical lesions than women who are not vaccinated.

Methods

We conducted this systematic review based on a protocol developed a priori (the protocol is available from the corresponding author upon request). The QUOROM (Quality Reporting of Meta-analyses) statement was used to guide the content and reporting of the review. (6)

Search

The methods used for conducting and reporting the literature search followed the approach proposed by the STARLITE investigators. (9) Using the OVID interface, we conducted electronic subset searches of 5 databases: MEDLINE (1950-2007 week 20), MEDLINE in process and other nonindexed citations (to June 20, 2007), EMBASE (1980-2007 week 21), the Cochrane Central Registry of Controlled Trials (to first quarter 2007) and the Cochrane Library. We also reviewed bibliographies of all included studies and of narrative reviews published in the last quarter of 2006 to May 2007, clinical trial registries, Google Scholar, public health announcements, selected conference proceedings (2004-2007) and information from vaccine manufacturers regarding any unpublished data or ongoing randomized controlled trials. For each electronic database, we developed an independent search strategy using relevant MeSH terms and terms identified from other reviews on HPV-related topics. Published search filters for identifying randomized trial evidence were applied to further refine the searches in both the MEDLINE and EMBASE databases. (10,11) Subsets of records retrieved both with and without application of the search filters were examined to ensure that relevant records were not being missed. The electronic search strategy used for the MEDLINE database is in Appendix 1 (available online at www.cmaj.ca/cgi/content/full/177/5/469/DC2).

Selection

We considered eligible all published and unpublished reports of randomized controlled trials, with no exclusion on the basis of language or year of publication. Only studies involving women were included, with no exclusion on the basis of age or other demographic characteristics of the women enrolled. Interventions included any vaccine against HPV, as long as it contained activity against at least 1 oncogenic strain of the virus and was being administered with prophylactic intent. Reports of therapeutic vaccination were excluded. Any dosing regimen was considered acceptable. Comparators had to be either placebo or a "no HPV vaccination" group. Studies not designed to address outcomes related to vaccine efficacy against oncogenic HPV strains were excluded. The ultimate goal of HPV vaccination is to prevent death from cervical cancer. However, because cervical cancer is a rare event, even among women with persistent infection with an oncogenic HPV strain, and because following patients to a cancer…