TY - JOUR
T1 - Direct comparison of HPV16 serological assays used to define HPV-naïve women in HPV vaccine trials
AU - Safaeian, Mahboobeh
AU - Ghosh, Arpita
AU - Porras, Carolina
AU - Lin, Shih Wen
AU - Rodriguez, Ana Cecilia
AU - Schiffman, Mark
AU - Wacholder, Sholom
AU - Kemp, Troy
AU - Gonzalez, Paula
AU - Wentzensen, Nicolas
AU - Esser, Mark
AU - Meuree, Ariane
AU - Matys, Katie
AU - Quint, Wim
AU - Van Doorn, Leen Jan
AU - Sherman, Mark E.
AU - Herrero, Rolando
AU - Pinto, Ligia A.
AU - Hildesheim, Allan
PY - 2012/9
Y1 - 2012/9
N2 - Background: Two HPV serological assays, the competitive Luminex immunoassay (cLIA), and an enzyme-linked immunoassay (ELISA) against HPV16 have been used to define HPV-naïve subcohorts within large HPV vaccination trials. Some of the variation in estimated vaccine efficacies may be due to the differences in these assays used to define the HPV-naïve subgroups. To guide the interpretation of published results, we compared these assays. Methods: Replicate enrollment sera from a stratified sample of 388 unvaccinated women from the control arm of the Costa Rica HPV 16/18 Vaccine Trial were measured for antibodies against HPV16 using cLIA and ELISA. Agreement between the assays was estimated using standard and alternative assay cutoffs. Results: Using laboratory-determined seropositivity cutoffs, sampling-adjusted HPV16 seropositivity was 24.8% by ELISA and 7.2% by cLIA. Comparing cLIA and ELISA antibody levels based on the standard cutoffs, overall agreement was 53% (positive-agreement = 49%). The poor agreement was mainly driven by the higher sensitivity of the ELISA than cLIA, resulting in 30% of the ELISA-positive sample that were cLIA-negative (none of the ELISA-negatives were cLIA-positive). Increasing ELISA cutoff to 54 ELISA units (EU)/mL (the level which maximized agreement with cLIA; ELISA standard cutoff is 8 EU/mL) resulted in higher agreement (overall agreement = 91%; positive agreement = 78%). Conclusions: ELISA and cLIA are different from each other based on the laboratory-determined cutoff. Increasing ELISA cutoff increased agreement with cLIA, which could facilitate comparisons among studies that use different assays. Impact: Keeping cLIA at the laboratory-determined cutoff but altering ELISA cutoff for seropositivity might facilitate vaccine efficacy comparisons in the naïve cohorts defined by cLIA.
AB - Background: Two HPV serological assays, the competitive Luminex immunoassay (cLIA), and an enzyme-linked immunoassay (ELISA) against HPV16 have been used to define HPV-naïve subcohorts within large HPV vaccination trials. Some of the variation in estimated vaccine efficacies may be due to the differences in these assays used to define the HPV-naïve subgroups. To guide the interpretation of published results, we compared these assays. Methods: Replicate enrollment sera from a stratified sample of 388 unvaccinated women from the control arm of the Costa Rica HPV 16/18 Vaccine Trial were measured for antibodies against HPV16 using cLIA and ELISA. Agreement between the assays was estimated using standard and alternative assay cutoffs. Results: Using laboratory-determined seropositivity cutoffs, sampling-adjusted HPV16 seropositivity was 24.8% by ELISA and 7.2% by cLIA. Comparing cLIA and ELISA antibody levels based on the standard cutoffs, overall agreement was 53% (positive-agreement = 49%). The poor agreement was mainly driven by the higher sensitivity of the ELISA than cLIA, resulting in 30% of the ELISA-positive sample that were cLIA-negative (none of the ELISA-negatives were cLIA-positive). Increasing ELISA cutoff to 54 ELISA units (EU)/mL (the level which maximized agreement with cLIA; ELISA standard cutoff is 8 EU/mL) resulted in higher agreement (overall agreement = 91%; positive agreement = 78%). Conclusions: ELISA and cLIA are different from each other based on the laboratory-determined cutoff. Increasing ELISA cutoff increased agreement with cLIA, which could facilitate comparisons among studies that use different assays. Impact: Keeping cLIA at the laboratory-determined cutoff but altering ELISA cutoff for seropositivity might facilitate vaccine efficacy comparisons in the naïve cohorts defined by cLIA.
UR - http://www.scopus.com/inward/record.url?scp=84866153670&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84866153670&partnerID=8YFLogxK
U2 - 10.1158/1055-9965.EPI-12-0558
DO - 10.1158/1055-9965.EPI-12-0558
M3 - Article
C2 - 22752082
AN - SCOPUS:84866153670
SN - 1055-9965
VL - 21
SP - 1547
EP - 1554
JO - Cancer Epidemiology Biomarkers and Prevention
JF - Cancer Epidemiology Biomarkers and Prevention
IS - 9
ER -