Maria Luisa Sardu (1), Anna Largajolli (1), Nele Plock (1), Emilie Schindler (1), Carolyn Cho (2), Seth Robey (2), Jos Lommerse (1), Li Qin (1), S. Y. Amy Cheung (1), Jeffrey R. Sachs (2)
(1) Certara; Princeton, NJ, USA (2) Merck & Co., Inc.; Kenilworth, NJ, USA
Introduction: Shingles, also known as herpes zoster (HZ), is a painful illness resulting from varicella-zoster virus (VZV) reactivation. It is characterized by a painful, blistering dermatomal rash, with a small proportion of patients developing post-herpetic neuralgia. Currently available HZ vaccines include zoster vaccine live [1] and zoster vaccine recombinant adjuvanted [2]. To support the development of improved vaccine candidates and provide an integrated understanding of their emerging data, knowledge about the relationship between vaccine efficacy and immunogenicity (such as titers of anti-glycoprotein (gp) antibody (Ab)) is desirable. A model relating fold-change titer to efficacy was previously published [3]. In addition, previous modeling work on varicella zoster using subject level data indicated that absolute titers may be predictive of the likelihood of breakthrough infections after vaccination [4]. To our knowledge however, a model integrating vaccine efficacy data across different vaccines, relating absolute titer to efficacy, is not publicly available. A Model-Based Meta-Analysis (MBMA) integrating literature data on zoster vaccine live and zoster vaccine recombinant adjuvanted was therefore performed to provide a quantitative framework for future decisions.
Objectives:
- Development of a clinical outcomes database for publicly available HZ clinical trial data.
- Development of a HZ incidence rate (IR) versus post-vaccination anti-gp Ab absolute titers correlates-of protection model
Methods: A systematic search in Pubmed, www.clinicaltrials.gov, company registries, regulatory reviews from FDA websites, and clinical study reports was performed to identify open-label trials, randomized controlled trials, or non-randomized prospective trials on vaccines against VZV, focusing on trials reporting HZ episode IR and anti-gp or anti-gE antibody as efficacy and immunogenicity endpoints, respectively. An MBMA was performed, focusing on confirmed HZ episodes and on anti-gp antibody absolute titers at 1- and 3-months post-vaccination. Age-stratified information was included in the analysis whenever available.
Results: Initial assessment of the analysis data indicated a very limited amount of informative non-zero IR data (<10 data points). A major component of the analysis was therefore the development and implementation of data enrichment strategies. These included updates of the time point selection, harmonization of age strata, Ab concentration unit conversion (based on Ab concentrations at baseline), and imputation of anti-gp Ab concentrations based on anti-gE Ab vs anti-gp Ab relationship: these were added to trial strata for which efficacy and anti-gE Ab information was available. A Poisson regression model expressing HZ IR as a function of anti-gp Ab concentrations was successfully fit to the enriched data. A suppression function (sigmoid Emax model) driven by anti-gp Ab titers and the log-ratio between the minimum IR (IRmin) and the maximum IR (IRmax) was assumed. Real-world data was added to support the estimation of IRmax per age strata. The analysis yielded IC50 as the key parameter to describe anti-gp Ab effect on HZ IR and was estimated to be 11.7 log2 units/mL anti-gp Ab concentration. A hill coefficient parameter was also identified, with an estimated value of 10.
Conclusions: Starting from an initially very sparse database, a data enrichment strategy yielded an informative dataset supporting MBMA model development for HZ. The MBMA model may be used to predict vaccine efficacy based on early Phase 1 vaccine immunogenicity data and enables quantitatively informed support of go/no-go decisions through integration of heterogeneous comparator data. Model and method are supported by comparability of resulting quantitative titer-protection with that from previous vaccine-specific work.
References:
[1] Zostavax https://www.cdc.gov/vaccines/vpd/shingles/hcp/zostavax/references-resources.html
[2] https://www.ema.europa.eu/en/medicines/human/EPAR/shingrix
[3] Dudášová, J, Laube, R, Valiathan, C, et al., NPJ Vaccines 6, 133 (2021). https://doi.org/10.1038/s41541-021-00377-6
[4] Chan, IS, Li, S, Matthews, H, Chan, C, Vessey, R, Sadoff, J, Heyse, J., Stat Med. 2002 Nov 30;21(22):3411-30. https://doi.org/10.1002/sim.1268
Reference: PAGE 30 (2022) Abstr 10108 [www.page-meeting.org/?abstract=10108]
Poster: Drug/Disease Modelling - Infection