Browsing by Author "Ukkonen, Aino"

The human papilloma virus (HPV) is the main cause of cervical cancer in women and HPV is the most common sexually transmitted infection. To prevent women from developing cervical cancer there are two methods in use, the primary method being vaccination against HPV and the secondary being screening. Some long term effects of screening and vaccination will not be observed during the first decades of vaccination, and therefore predictive mathematical models serve as an indicator of what to anticipate in the future. In this thesis we studied two types of cervical cancer, adenocarcinoma and squamous cell carcinoma. We determined how vaccination against a virus, associated with squamous cell carcinoma, together with screening programs affects adenocarcinomas caused by non-vaccine type HPV. A precancerous adenocarcinoma lesion, which is located deeper in the glandular cervical tissue, is difficult to detect directly in screening but can be found indirectly by uncovering a common type of cervical cancer, squamous cell carcinoma, which is easier to detect in screening. These two cancers are mostly associated with two different strains of HPV. When vaccinating against the strain that is found in squamous cell carcinomas the elimination of the precancerous stages of squamous cell carcinomas mean that the detection method for adenocarcinoma is impaired, which allows for a possible increase in adenocarcinoma prevalence. In this thesis we studied this possible increase. To predict the effect vaccination has on adenocarcinoma, we constructed a mathematical model of two HPV types, which were associated with squamous cell carcinoma and adenocarcinoma respectively. We modeled a vaccine that protects against the HPV types associated with squamous cell carcinoma but not adenocarcinoma. The model included the simplified natural history of HPV, the progression of the disease, the vaccination program and the screening program. For the two virus infections we developed a deterministic compartmental progression model. In the computations we modeled a cohort of women through their lifetime and studied the precancerous findings in screenings and the number of cancer cases in the cohort. To understand which model components contributed to the adenocarcinoma incidence a sensitivity analysis was conducted by varying the screening sensitivity, the force of infection and the recovery rates. The model yielded an increase in adenocarcinomas when vaccinating against the virus associated with squamous cell carcinoma. The incidence of adenocarcinoma correlated with the non-vaccine type HPV infections that would otherwise be found in screenings without vaccination. The increase in adenocarcinomas was more prominent in a sexually active population. Compared to the reduction in squamous cell carcinoma provided by the vaccine, the adenocarcinoma increase was, although positive, very minor.