Nearly one in ten cancers is a consequence of a viral infection. Among them, the majority are caused by human papillomavirus (HPVs). These cancers are a privileged target in public health because they can also be fought with the "classic" arsenal used against infectious diseases. Unfortunately, despite the release of safe and efficient vaccines since 2006, more than 6,000 new cases of HPV-induced cancers were diagnosed in France in 2015, about half of which affect the cervix, a quarter the anal region and a quarter the oropharyngeal region. The main difficulty in studying these cancers lies in the discrepancy between the immense prevalence of HPVs (more than 80% of sexually active adults will be infected before they turn 45) and their low virulence per infection (even for HPV16, the most virulent type, less than 2% of infections cause invasive cancers). Finally, very little data is available for non-persistent infections because they are generally benign and asymptomatic.
We propose to develop mathematical models to make the most of the limited data available on non-persistent HPV infections and incorporate them into a common conceptual framework. In particular, we will study the role played by stochasticity, and therefore rare events, on two open questions: 1) Why is the majority of infections eliminated? and 2) Why does a minority of chronic infections lead to cancers? These models will be based on both so-called viral kinetics approaches, based on differential equations (deterministic), and branching processes (stochastic). The models can be calibrated and tested using data from the PAPCLEAR clinical study conducted by the team, as well as data from a collaboration with the National Reference Centre in Besançon.
Murall C. L. & Alizon S. (2019) Modelling the evolution of viral oncogenesis. Phil Trans R Soc Lond B 374(1773):20180302For further details, see our publications »