Human papillomavirus (HPV): space-time epidemiology and issues concerning laboratory diagnosis

Human papilloma virus (HPV) is considered the most common sexually transmitted agent in the world. It is the main responsible of cervical cancer, the second more frequent cancer in women (Catsellsague et al., 2007, De Sanjosé et al., 2007). In particular, each year 529,409 women are diagnosed with cervical cancer, while women with older ages (65 years and over) have the greater mortality (WHO 2010, Catsellsague et al., 2007). For these reasons, HPV worldwide impact is consistent. Introduction of Papanicolau cytological test permitted a decrease of incidence and mortality rates of HPV-related cervical cancer. However, it can fails to reduce cervical cancer rates, especially in developing countries (Agorastos et al., 2010). For this reason, virological test is a promising tool to reveal HPV infection in early stages. Despite the importance of HPV testing and genotyping, actual tests are not uniform with regard to genotype-specific detection rates (Klug et al., 2008). Therefore, diagnostic methods for HPV genotyping must be implemented and improved. For the above mentioned reasons, the goal of the present thesis was to deepen the information on the diagnostic approaches and epidemiology of HPV infection. Two methods (Manos+GP PCRs, SPF/InnoLipa [SPF], ProDect ®HPV Typing [BCS]) for the diagnosis of HPV infection were compared. Differences in the identification of HPV infected samples and infectious genotypes identified by the two methods were evaluated. In the second part of the thesis, an analysis of the geographical distribution of HPV cases in Rome and surroundings was performed to evaluate a potential presence of HPV infection clusters and HPV genotypes clusters. Finally, a time-series analysis was conducted to evaluate the trend of HPV infection over time and potential of seasonal regularities of HPV cases. The evaluation of SPF and the BCS methods revealed that they are not perfectly concordant on HPV genotyping, as showed by the significant differences in the prevalence found for HPV73, 59, 56, 53, 52, 51, 35 and 31. On the other hand, the overall concordance between the two tests was 86.1%; in particular concordance for the detection of high-risk genotypes was 87.0%. Moreover, SPF and BCS were highly concordant in the detection of HPV18 and HPV16, the HPV genotypes most related with cervical cancer. Such result is particularly interesting in the era of HPV vaccination for the selection of the ideal candidates for prophylactic vaccination. Geostatistical analysis showed geographical differences of HPV infections and HPV genotypes. In particular, a different distribution of HPV16 and HPV18 in the different municipalities in the Latium region was observed. By time-series analysis, no seasonal component was found. Moreover, no growing trend was observed. Finally, a specific ARIMAX model was built for forecasting. The addition of birth rates as covariate in the model improved the quality. Such result could be useful to make short or medium term predictions about the prevalence of HPV infections. In conclusion, the findings obtained in the present thesis highlight the importance to have good diagnostic methods and the lead a careful epidemiological surveillance in infectious diseases, such as HPV infection, to correctly evaluate the impact of a specific health intervention.