Abstract

Background: A hepatitis B vaccine adjuvanted with a Toll-like receptor 9 agonist achieves higher seroprotection rates than the conventional alum-adjuvanted vaccine in fewer doses, and is licensed on that basis. Seroprotection is defined as an anti-HBs concentration at or above 10 mIU/mL, a binary threshold. Objectives: To express the comparative benefit in absolute terms and by population; to identify where the reported comparisons are and are not like for like; and to examine what the reverse cumulative frequency data show about the distribution of antibody concentrations that a threshold endpoint cannot capture. Methods: Randomised comparisons of CpG-adjuvanted against alum-adjuvanted hepatitis B vaccine in adults reporting seroprotection rates were eligible. Absolute differences, confidence intervals and numbers needed to vaccinate were derived from reported proportions. Reported descriptions of the antibody concentration distribution were extracted and are presented schematically. Analyses were performed in Python 3. Results: Across four randomised trials comprising 7,056 recipients of the two-dose CpG-adjuvanted vaccine and 3,214 recipients of the three-dose alum-adjuvanted vaccine, seroprotection rates were 90.0% to 100.0% against 70.5% to 90.2%. In adults aged 18 to 55, the CpG vaccine reached 88.4% at week 8 against 26.7% for the comparator at the same visit—a contrast at which the comparator had received only two of three doses—and against its peak of 81.3% at week 28 the difference was 7.1% points (number needed to vaccinate 14.1). In adults aged 60 to 70 with type 2 diabetes the difference was 27.3% points (85.8% against 58.5%, 95% CI 18.0-36.8; number needed to vaccinate 3.7). Geometric mean concentration at week 8 was 83.1 against 6.5 mIU/mL and median time to seroprotection was five months earlier. Reverse cumulative frequency data show that more CpG recipients crossed the threshold and occupied the 10 to 1,000 mIU/mL band, while more alum recipients exceeded 1,000 mIU/mL and more failed to seroprotect at all. Conclusions: The CpG-adjuvanted vaccine confers a substantial and clinically meaningful advantage, largest in exactly the populations for whom conventional vaccination fails most often. The distributional data indicate that the adjuvant compresses the antibody response—raising the floor and lowering the ceiling—which a threshold endpoint cannot detect. Whether the smaller high-titre tail bears on duration of protection is a question the seroprotection endpoint is not constructed to answer, and it warrants longterm follow-up rather than inference.

Keywords: Hepatitis B vaccine, CpG 1018, Toll-like receptor 9, Seroprotection, Adjuvant, Antibody distribution