In humans, viruses contribute to the development of about 15% of all malignancies. Cancers of the liver and cervix make up approximately 80% of all viral associate malignancies.

Human papillomavirus (HPV) is seen in almost all cervical cancers. The epidemiologic evidence that HPV is the main causative agent of cervical cancers is overwhelming. For instance, when patients with disease are compared with population-based controls odds ratios of >200 have been observed.

Papillomaviruses are DNA viruses consisting of an outer protein coat which surrounds a genome of double stranded DNA containing 8000 base pairs. HPV DNA is divided into "early (E) and "late" (L) genes. Early genes are responsible for DNA replication, transcriptional regulation and transformation. Late genes control the formation of the capsid coat. Early gene products E6 and E7 encode the major transforming proteins which are capable of inducing cell proliferation and immortalisation.

In humans, HPV infections occur only in epithelial surfaces such as skin and mucous membranes. Minor trauma, as occurs with sexual intercourse, allows access to the basal cells of cervical epithelium. HPV infection involves coordinated expression of early viral proteins in lower epithelial layers with a switch to late gene expression as viral replication takes place leading to koilocytosis, nuclear enlargement and, in some cases intraepithelial neoplasia. Viruses multiply exclusively in the nuclei of infected cells.

To date over 100 types of HPV have been identified. Over 30 types infect cervical mucosa. There are at least 15 types of HPV associated with high grade cervical intraepithelial neoplasia (CIN) and cervical carcinoma.

HPV infection is very common in sexually active young women, with a prevalence of between 20 to 46%. The majority of these infective episodes are of short duration, and the prevalence of high risk HPV infection drops to 3-7% in women aged 30 and over. The median duration of a new infection is 8 months. By 12 and 24 months, 70% and 91% of women are no longer infected.

The frequency of integration of the viral genome into the host chromosome correlates with lesion grade, being rare in CIN 1 and common in CIN 3. Viral integration occurs in all carcinomas.

HPV and its oncogenic genes/proteins are only capable of inducing cellular immortalisation with transformation to the malignant phenotype requiring activation of cellular oncogenes and/or loss of tumor suppressor genes. Several studies have identified loss of heterozygositiy on chromosome 3p and gain of chromosome 3q with possible importance in progression from CIN to invasive disease. There is also evidence that activation of oncogenes, such as c-Ha-ras, is important in progression of malignancy.

Cigarette smoking, both passive and active, has been associated with an increased risk of CIN and invasive cervical cancer. Smoking acts in various ways including modulating the action of the immune system resulting in a reduction of Langerhan's cells in the cervix.

The vast majority (over 80%) of HPV infections are transient, being cleared by the immune system within a few months. The remaining 20% persist and go on to promote the development of CIN. In older women, persistence and subsequent progression to a high grade lesion are more frequent. Women with CIN who are unable to clear a high risk HPV infection will then be at risk of further genetic damage and progression to invasive cancer. The risk of progression of HPV related lesions correlates with viral type.

There are three different methods for measuring HPV in clinical specimens; non-amplification methods (e.g. Southern blotting), signal amplification (e.g. hybrid capture), and DNA amplification (e.g. polymerase chain reaction). In general, non-amplification methods have fallen out of favor because of a lack of sensitivity and specificity. Polymerase chain reaction is a flexible tool that can be used to analyze fresh tissues specimens, paraffin embedded tissue and cytological samples for defined parts of the HPV genome (particularly the L1 gene). Because it is such a sensitive technique, however, polymerase chain reaction studies are susceptible to cross contamination. At present there are no commercially available kits for polymerase chain reaction studies.

Hybrid capture techniques analyze cells obtained from a brush or swab. The viral capsid is disrupted and the released HPV DNA is bound to a specific RNA probe. The RNA:DNA hybrid is tagged with a chemi-luminecent substrate which then permits measurement. This is currently the only type of HPV assay commercially available. The sensitivity of the assay is high in all populations studied. However, as high risk HPV's are also found in low grade disease and normal pap smears, the specificity and positive predictive value for high grade disease and cervical cancers depends on the prevalence of HPV infection in the population.

Screening for high risk HPV in young sexually active women will do more harm than good because of the high prevalence of HPV and the tendency of the the majority of these lesions to regress. HPV screening for high risk HPV in older women will have a higher chance of finding persistent HPV infection with a high progression rate.

Although testing for HPV in detection, monitoring and preventing invasive disease is important, the ultimate aim must be eradication of high risk HPV by vaccination. The immune system controls both the primary infection and progression. The cell-mediated immune system seems to be more important in the prevention of HPV infection than the humoral system.

HPV's are viewed as infective agent of low immunogenicity; this combined with the large number of types makes the development of vaccines a potentially complex task. The most attractive vaccines will be those that are multivalent, cheap, easily administered and have both prophylactic and therapeutic potential. It should be noted that genetic engineering methods will be required, as these viruses cannot be grown in conventional cell cultures.

Condensed from: Kirwin et al. Human papillomavirus and cervical cancer: where are we now? Br J Obstet Gynecol 2001;108:1204-1213