Impact of screening for human papillomavirus-positive oropharyngeal cancers: a microsimulation-based modeling study.

Abstract

We estimated the impact of screening on morbidity and mortality of human papillomavirus (HPV)16-positive oropharyngeal cancer among US men aged 45-79 years. We developed an individual-level, state-transition natural history microsimulation model to estimate the impact of screening using oral HPV16 detection, HPV16-E6 antibody detection, and transcervical ultrasound of neck/oropharynx. We compared clinical detection to counterfactual screen detection for cancer stage, single- vs multiple-modality treatment, and survival. Screening scenarios encompassed 4 progression speeds across cancer stages (very slow, slow, fast, and very fast) and 4 screening frequencies. Among US men aged 45-79 years in 2021 (n = 54 881 311), 163 958 clinically diagnosed human papillomavirus (HPV)-positive oropharyngeal cancers and 32 009 deaths would occur through age 84 in the absence of screening. Assuming very-fast progression, 4%, 20%, 31%, and 60% of these cancers would be detected by one-off, 5-yearly, 3-yearly, and annual screening. Annual screening (very-fast progression) could reduce the number of cancers diagnosed at advanced stages (American Joint Committee on Cancer, 7th edition, stages III/IV: 90.0% with no screening vs 59.1%) and treated by multiple modalities (80.6% with no screening vs 50.6%). Cancer mortality would also be reduced by 36.2%, with a gain of 106 000 life-years. Annual screening would have a number needed to screen (NNS) of 561 per screen-detected cancer, 1118 per additional cancer treated by single modality, 4740 per death prevented, and 520 per life-year gained; such high NNS reflect potential inefficiency of population-level screening. If proven efficacious in randomized trials and cost-effective, screening for HPV-positive oropharyngeal cancers could provide considerable population-level reductions in advanced stage cancers, treatment-related morbidities, and mortality.