If the diagnosis rate (incidence) increases, and people aren't dying rapidly, then the prevalence would surely increase as well. Unless you're using prevalence in the context of total people affected by a disease, regardless of whether or not they were diagnosed.