Modelling screening of HPV vaccinated birth cohorts. The infection transmission system

Transcription

1 Modelling screening of HPV vaccinated birth cohorts Dr. Iacopo Baussano Imperial College, London & UPO/CPO-Piemonte, Italy. The infection transmission system Infectious Agent in case of CC: men in case of CC: women in case of CC: carcinogenic HPV Biological Vector Human Host Environment Dynamic transmission systems: mechanistic models based on the assumption that infectious agent transmission results from the dynamic interaction between populations (IA, HH, BV) within a specific and changing environment.

2 Modelling ID natural history Exposure Disease Virus Immune response Individuals Years Weeks Decades Susceptible (S) transmission Clearance Immune (R) I N HPV infection (I) Susceptibles: Infected: Recovered: ds dt di dt dr dt S, S I I, Cases (n) Populations 1 2 Days Infectious Susceptible Recovered Modelling ID control - Vaccination Incidence of infection is directly proportional to exposure (characterized by contact rate, probability of transmission, duration of infectiousness) and fraction of susceptible individuals Vaccination acts on Incidence of infection by decreasing the fraction of susceptible S I R Cases (n) vaccination Infectious Optimal Coverage Time Cases (n) vaccination Infectious Sub-optimal Coverage Time

3 Behaviour Changes Impact on cervical cancer incidence - Changes in sexual behavior - Changes in HPV16 incidence - Changes in attitude towards screening Age group targeted Coverage, % of the female population Rate per 1 women-year Predicted Observed Visits Invitations ( Geographical Heterogeneity Florence - HC2 prevalence(%) Bologna - HC2 prevalence(%) Imola - HC2 prevalence(%) Age (years ) Viterbo - HC2 prevalence(%) Verona - HC2 prevalence(%) Trento - HC2 prevalence(%) 2 1 Observed prevalence Model s estimate Age (years ) Padua - HC2 prevalence(%) Turin - HC2 prevalence(%) Ravenna - HC2 prevalence(%) Age (years ) Baussano et al., 211.In progress

4 Modelling Cervical Cancer Control HPV infection and Cervical Cancer Natural History Estimation of the probability of transmission of HPV-16 infection. Impact of vaccination or screening (separately) Impact of vaccination as a function of patterns of clearance of HPV-16 infection. Impact of screening as a function of changes in sexual behaviors over time. Investigation of integration of vaccination and screening Identification of optimal screening round intervals as a function of vaccination coverage. Sample size definition for a RCT combining HPV vaccination & screening strategies. HPV and Cervical Cancer Model Infectious cases Prevention Unvaccinated Susceptible Birth Vaccination Vaccination Vaccinated all causes mortality rate Clearance (SIR) Clearance (SIS) HPV infection Loss of vaccine induced immunity Immune progression Lesion Clearance Cancer mortality CIN2+ Lesion progression Cervical Cancer Screen &Treat Screening programme Baussano et al., Epidemics 21; 2:

5 Transmission probability & assortativeness HPV-16 Prevalence (%).% 4.% 3.% 2.% 1.%.% Observed HPV-16 prevalence Model's estimate (best fitting combination of parameter values) Age Classes Range explored (best estimate) Probability of HPV-16 transm. 1 (.4) CIN2+ incidence 2.E E E -3 8.E -4 4.E -4.E+ Observed CIN2+ incidence (HPV-16+)- Model's estimate (best fitting combination of parameter values) Assort. Age 1 (.4) Assort. CSA 1 (.4) Fully assortative Random choice. Baussano et al., Epidemics 21; 2: Age Classes Burchell et al., (.36.47) per partnership Patterns of clearance and vaccination SIS SIR Endemic prevalence 6% % 4% 3% 2% 1% % Taira, 24. (~%) Transmission Probability SIS = 16% Susceptible Infected % X 1 X 2 1% Effective coverage of vaccine Susceptible Infected Immune Transmission Probability SIR = 4% Baussano et al., Vaccine 211; 29:

6 Patterns of clearance and Herd Effect Women only vaccination Men and women vaccination Fraction of Cervical Cancers (HPV-16+) prevented through vaccination (%) 1% 8% 6% 4% 2% % % 2% 4% 6% 8% 1% 1% 8% 6% 4% 2% SIR model Indirect Protection Quinquennial birth cohorts Vaccinated Cohorts (2-22) Partially Vaccinated Cohorts (199-2) Unvaccinated Cohorts ( ) Vaccinated Cohorts (2-22) Partially Vaccinated Cohorts (199-2) 1% 8% 6% 4% 2% SIS model % % 2% 4% 6% 8% 1% 1% % Unvaccinated Cohorts % % 2% 4% 6% 8% 1% ( ) % 2% 4% 6% 8% 1% Vaccination Effective Coverage % (before sexual debut); VEC = CoverageVaccineEfficacy (9%) 8% 6% 4% 2% Baussano et al., Vaccine 211; 29: Sexual Behaviour & Impact of Screening Cervical cancer incidence (per 1, women-year) Year opportunistic screening opportunistic screening organized screening organized screening Cervical cancer incidence as reported by cancer registry Turin Florence Parma

7 Integrated cancer control programmes Vaccination coverage (%) unscreened women yrs yrs 6yrs 7yrs 11yrs 1yrs 9yrs 8yrs 2yrs 19yrs 18yrs 17yrs 16yrs yrs 14yrs 13yrs 12yrs Vaccination coverage (%) screened women Baussano et al., 211. in progress Phase IV HPV16/18 vaccination trial 33 communities (11 communities/arm): A-arm: HPV16/18 boys&girls B-arm: HPV16/18 girls, boys HBV C-arm: HBV boys&girls Birth-cohorts enrolled: 1992/93 in 27/8 1994/9 in 28/9/1 Total eligible: 8 early adolescents Enrolled: vaccin./ volunteers 1992/93 born 16 / /9 born 16 2 / 18 Total: 32 2 / 3 Follow-up: 6% HPV-tested at 18 years of age End-points: Significant reduction of hrhpv prevalence by birth-cohort in 18-year old overall and in unvaccinated by (A vs. C, B vs. C, A vs. B)

8 Discussion: dynamic perspective Opportunities Explicit theoretical framework Support study design e.g. to estimate sample size for community RCT. Cost-effectiveness analyses. Unifying approach to investigate the epidemiology of cancers caused by infections Combining dynamic and carcinogenic models Limitations Overwhelming complexity and outputs are sensitive to assumptions and initial conditions Limit the interpretation of the results (uncertainty and sensitivity analyses) New skills and tools needed Computing methods and high-performance clusters Acknowledgments Prof. Geoff Garnett. Bill & Melinda Gates Foundation, Seattle, US. Dr. Guglielmo Ronco. CPO-Piemonte, Turin, Italy Prof. Paolo Vineis. Imperial College, London, UK Prof. Matti Lehtinen. University of Tampere, Finland. Funding FIRMS-Turin, Italy PreHdict project FP7, EU.

9 Thank you for your attention