Effectiveness of bivalent and quadrivalent human papillomavirus vaccination in Luxembourg

In Luxembourg, the human papillomavirus (HPV) vaccination program introduced in 2008, provided either bivalent (BV) or quadrivalent (QV) vaccines to girls aged 12-17 years. Here, we estimate the effectiveness of BV and QV vaccines combined and separately in reducing type-specific HPV prevalence eight years after the introduction of the vaccination program.

A cross-sectional prevalence study was conducted among women aged 18-29 years in 2015-2017. Seven hundred sixteen participants were recruited at family planning centres or private gynaecology practices in Luxembourg. Vaccination records were verified in the social security database. Cervical samples were tested using the Anyplex II HPV28 assay. Vaccine effectiveness was estimated using logistic regression.

In total, 363/716 (50.7%) participants were HPV positive with any HPV and 209/716 (29.2%) with carcinogenic HPV genotypes. HPV vaccination offered high protection against HPV16/18 (adjusted odds ratio (AOR) = 0.13; 95% CI 0.03-0.63), HPV6/11 (AOR = 0.16; 95% CI 0.05-0.48) and cross-protection against HPV31/33/45 (AOR = 0.41; 95% CI 0.18-0.94). The AORs were generally enhanced when only considering vaccination before sexual debut corresponding to AORs: 0.05 (95% CI 0.00-0.88), 0.08 (95% CI 0.02-0.36) and 0.20 (0.06-0.65) against HPV16/18, HPV6/11 and HPV31/33/45, respectively. We observed significant protection against carcinogenic genotypes included in nonavalent vaccine for BV (AOR = 0.29; 95% CI 0.13-0.67), but not for QV (AOR = 0.81; 95% CI 0.47-1.40) (heterogeneity Chi² P = 0.04).

Our study suggests high effectiveness of HPV vaccination against HPV6/11, HPV16/18 and a cross-protection against HPV31/33/45. Vaccination effectiveness was slightly higher for women vaccinated before sexual debut.