A Simulation Study of Using Composite HPV Genotyping Assay Results to Monitor Human Papillomavirus Infections

Background:  Researchers often group various HPV types into composite measures based on vaccine subtypes, oncogenic potential, or phylogenetic position.  Composite prevalence estimates based on various genotyping assay results have been calculated to assess the burden of HPV infections and to monitor HPV vaccine effectiveness.  While these prevalence estimates have been used to assist development of prevention strategies, research on how well these estimates measure the true underlying infection burden is limited. 

Methods:  A simulation study was conducted to evaluate use of composite genotyping assay results to monitor HPV infections when underlying infection rates change or assays with different performance are used.  Data were generated based on mathematical algorithms with pre-specified type-specific infection rates, assay sensitivity and specificity, and correlations between various HPV types.  Estimated-to-true infection rate ratios and percent reduction of vaccine types were calculated. 

Results:  When the true underlying type-specific prevalence was specified as the reported prevalence between 2003-2006 in the US and genotyping assays with high sensitivity and specificity (0.95, 0.95) were used, estimated-to-true rate ratios were 2.31, 2.17, 2.12 and 1.62, for composite measures with 2 vaccine high-risk types, 4 vaccine types, 14 high-risk types and 37 HPV types, respectively.  For a given composite measure, estimated-to-true rate ratios increased when the true underlying infection rates or assay specificity declined.  When underlying type-specific infection rates were reduced by 50%, the true composite infection rate of any of the 4 vaccine types decreased by 47%, but the estimated rate only decreased by 15%.

Conclusions:  Composite infection rate estimates calculated based on panels of genotyping assay results generally over-estimate the true infection burden and could under-estimate effectiveness.   The impact of assay specificity is as or more important than sensitivity and should be considered in selecting a genotyping assay to monitor HPV.