Background Human papilloma viruses (HPV) 16/18 are known to cause approximately 70% of cervical cancers. Phase III clinical trials of HPV vaccination have demonstrated >95% efficacy against persistent HPV type 16/18 infections and associated cervical intraepithelial neoplasia (CIN) grade 2+ lesions, and up to 90% efficacy against all CIN3+ lesions. A long-term follow-up is, however, needed to confirm the protective efficacy against cervical carcinoma.
Methods Phase III clinical trial (FUTURE II) consisted of intensive clinical 4-year follow-up including health education and counselling. The intervention potentially affects the incidence of neoplasia also in the placebo group. To increase power of the long-term follow-up and to determine the impact of the clinical intervention as such, a population based reference cohort of similarly aged women not exposed to any intervention was enrolled at the same time from the same communities. The HPV vaccine cohort and placebo vaccine cohort of 16–17-year-old women from the Finnish FUTURE II trial (N=1749) and a reference cohort of 18–19-year-old women (N=15 744) were linked with the Finnish Cancer Registry to determine the incidence of CIN3 and cervical cancer (CIN3+) during the passive follow-up, starting 6 months after the clinical follow-up of the phase III trial was completed.
Results & Conclusions Currently the incidence of CIN3+ at the age of 20–24 years is 95 per 100 000 person years in Finland (http://www.cancer.fi). The incidence doubles in 5 to 10 years as the cohorts age. Thus, in less than 10 years the cumulative incidence yields 80% power to demonstrate 90% vaccine efficacy against cervical CIN3+. During the first 2 years this passive registry-based follow-up identified no CIN3+ cases in the HPV vaccine cohort, two cases in the placebo vaccine cohort, and 21 cases in the unvaccinated reference cohort suggesting that the vaccine efficacy translates into efficacy against cervical cancer. The passive follow-up continues and new cases emerging in future will be monitored by redoing linkage with the population-based cancer register at specific time intervals in the future, which will effectively add up person years to our follow-up study. In conclusion, valid comparisons between the vaccine and placebo recipients (excluding cross-vaccinated placebo vaccine recipients) and the reference cohort not exposed to intervention are feasible, and will be critical to define more definitively the long-term protection provided by HPV vaccination against the hard endpoints.