(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Long-term vaccination strategies to mitigate the impact of SARS-CoV-2 transmission: A modelling study [1]

['Alexandra B. Hogan', 'School Of Population Health', 'Faculty Of Medicine', 'Health', 'University Of New South Wales', 'Sydney', 'Mrc Centre For Global Infectious Disease Analysis', 'Jameel Institute', 'School Of Public Health', 'Imperial College London']

Date: 2023-12

Abstract Background Vaccines have reduced severe disease and death from Coronavirus Disease 2019 (COVID-19). However, with evidence of waning efficacy coupled with continued evolution of the virus, health programmes need to evaluate the requirement for regular booster doses, considering their impact and cost-effectiveness in the face of ongoing transmission and substantial infection-induced immunity. Methods and findings We developed a combined immunological-transmission model parameterised with data on transmissibility, severity, and vaccine effectiveness. We simulated Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) transmission and vaccine rollout in characteristic global settings with different population age-structures, contact patterns, health system capacities, prior transmission, and vaccine uptake. We quantified the impact of future vaccine booster dose strategies with both ancestral and variant-adapted vaccine products, while considering the potential future emergence of new variants with modified transmission, immune escape, and severity properties. We found that regular boosting of the oldest age group (75+) is an efficient strategy, although large numbers of hospitalisations and deaths could be averted by extending vaccination to younger age groups. In countries with low vaccine coverage and high infection-derived immunity, boosting older at-risk groups was more effective than continuing primary vaccination into younger ages in our model. Our study is limited by uncertainty in key parameters, including the long-term durability of vaccine and infection-induced immunity as well as uncertainty in the future evolution of the virus. Conclusions Our modelling suggests that regular boosting of the high-risk population remains an important tool to reduce morbidity and mortality from current and future SARS-CoV-2 variants. Our results suggest that focusing vaccination in the highest-risk cohorts will be the most efficient (and hence cost-effective) strategy to reduce morbidity and mortality.

Author summary Why was this study done? Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), the virus causing Coronavirus Disease 2019 (COVID-19), is now endemic globally.

Vaccination remains important to reduce hospitalisations and deaths.

Health authorities need to consider how frequently booster doses will be required and in which age groups.

They also need to know the additional value of switching to vaccines that have been adapted to match more recently circulating variants of concern. What did the researchers do and find? We developed a mathematical model that captures the continued circulation and evolution of the SARS-CoV-2 virus in the presence of widespread infection-induced immunity from past exposure as well as vaccine-induced immunity from primary vaccination campaigns in 2021/2022.

We used the model to explore different strategies for continued vaccination including the age group targeted, the frequency of boosting, and whether the vaccine was adapted to match more recent variants.

In both high-income and low-middle-income settings, regardless of whether there was a high level of transmission in 2020/21 or a zero-COVID policy, we found that the most cost-effective vaccination strategy was to boost those at highest risk.

We found that the current variant-adapted vaccines could avert nearly twice as many hospitalisations and deaths compared to the ancestral vaccines, and that updating these vaccines each year—as is done for seasonal influenza—could avert a further 30% of hospitalisations and deaths. What do these findings mean? Continued booster vaccinations will remain important to reduce both hospitalisations and deaths but should be targeted towards the highest risk groups.

The estimates provided here can help to inform discussions around value for money by comparing the cost-effectiveness of COVID-19 vaccination to other health programmes.

The precise values are limited by uncertainty as to whether the virus will continue to evolve, whether any new variants may emerge, and the additional protection provided by further booster doses given the widespread exposure of the population to infection.

Citation: Hogan AB, Wu SL, Toor J, Olivera Mesa D, Doohan P, Watson OJ, et al. (2023) Long-term vaccination strategies to mitigate the impact of SARS-CoV-2 transmission: A modelling study. PLoS Med 20(11): e1004195. https://doi.org/10.1371/journal.pmed.1004195 Academic Editor: Mirjam E. E. Kretzschmar, Universitair Medisch Centrum Utrecht, NETHERLANDS Received: January 17, 2023; Accepted: October 25, 2023; Published: November 28, 2023 Copyright: © 2023 Hogan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: The model code is open access at https://mrc-ide.github.io/safir. All analysis code is available at https://github.com/mrc-ide/covid_booster_strategies. Funding: This work was supported by a grant from World Health Organisation to ACG. ABH acknowledges support from an NHMRC Investigator Grant and Imperial College Research Fellowship. PW is supported by an Imperial College Research Fellowship. OJW is supported by a Schmidt Science Fellowship in partnership with the Rhodes Trust. GC and ACG acknowledge support from The Wellcome Trust. NMF and ACG acknowledge funding from the MRC Centre for Global Infectious Disease Analysis (reference MR/X020258/1), jointly funded by the UK Medical Research Council (MRC) and the UK Foreign, Commonwealth & Development Office (FCDO), under the MRC/FCDO Concordat agreement and is also part of the EDCTP2 programme supported by the European Union. This work was additionally supported by the NIHR Health Protection Unit for Modelling and Health Economics (NMF: [NIHR200908]); and philanthropic funding from Community Jameel (PD, NMF). WHO provided feedback and relevant data on early iterations of the scenarios presented here but did not endorse the outputs or play any other role in the study. The other funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: ACG was previously a non-renumerated member of a scientific advisory board for Moderna, has received consultancy funding from GSK and Sanofi related to COVID-19 vaccination, and is a member of the CEPI scientific advisory board. She has received grant funding from Gavi for COVID-19 related work. ABH, PW and ACG have previously received consultancy payments from WHO for COVID-19 related work. ABH provides COVID-19 modelling advice to the New South Wales Ministry of Health, Australia. ABH was previously engaged by Pfizer Inc to advise on modelling RSV vaccination strategies for which she received no financial compensation. EMR is a non-remunerated member of the UK Vaccines Network, the UKRI COVID-19 taskforce and the British Society for Immunology Covid-19 taskforce. SLW is currently employed by Merck (USA/CAN), MSD (outside USA/CAN). None of this work was done while the author was an employee and none of his work for Merck is related in any way to the contents of this paper. NMF declares grants from UK MRC, UK NIHR, Bill and Melinda Gates Foundation, Gavi, Wellcome Trust. NMF is Member of UK government and WHO advisory committees. Abbreviations: COVID-19, Coronavirus Disease 2019; HIC, high-income country; ICU, intensive care unit; IL, immunity level; LMIC, lower-middle-income country; SARS-CoV-2, Severe Acute Respiratory Syndrome Coronavirus 2; VFR, variant fold reduction; WHO, World Health Organization

Introduction The rapid development and delivery of vaccines to protect against Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection and Coronavirus Disease 2019 (COVID-19) dramatically altered the course of the pandemic, saving an estimated 19.8 million lives in the first year of vaccination alone [1]. However, the effectiveness of COVID-19 vaccines wanes, with considerable declines against infection but slower declines against severe disease and death [2–8]. Thus, it is likely that continued booster programmes will be needed to maintain high effectiveness against severe disease and death, particularly in those at highest risk of more severe outcomes [9]. In addition, vaccine booster programmes have been successful in partially restoring effectiveness against severe disease and death when levels of existing vaccine protection have been eroded by the emergence of new variants that have resulted in immunological escape [10–12]. Many countries are continuing to evaluate how best to schedule regular boosting to protect against ongoing endemic circulation of the virus as well as against future epidemic waves with new variants. This is reflected in the World Health Organization (WHO) roadmap which, in the most recent update (March 2023), recommended different vaccination schedules for high, medium, and low priority use groups, with boosting restricted to the highest risk group (older adults, younger adults with significant comorbidities or severe obesity, adults with moderate to severe immunocompromising conditions, pregnant people, and frontline health workers) [9]. The benefit of such strategies in any given population will depend on the current stage of the vaccine programme, including the supply of vaccine doses, and the extent to which these doses are matched to the current circulating strains. It will also depend on the extent of infection-acquired immunity and the additional protection that this provides. In general, both cohort and test negative case-control studies have suggested an additional impact of past infection in addition to vaccine-induced immunity in providing protection against severe outcomes, although these studies were not designed to control for the timing of exposure versus vaccination [13–15]. However, in a population-based cohort with frequent access to COVID-19 testing that enabled the timing of exposures to be included in the analysis, hybrid immunity was demonstrated to provide higher levels of neutralising antibodies over time [16]. Lastly, the benefits of booster vaccination will depend on the extent to which any future variant replaces the current Omicron variant and whether it further evades existing immunity. Throughout the COVID-19 pandemic, mathematical and computational modelling has been a key component of longer-term planning and has been widely used to inform decisions on future vaccine strategies [17]. A number of studies have focussed on either country-specific projections of epidemic progression and vaccine impact, or on allocation or prioritisation of the limited supply of doses (particularly in the early stages of vaccine rollout) [18–22]. More recently, models have been developed to consider longer-term strategies for continued vaccination. Data-driven approaches have used the relationship between neutralising antibody titres and protection to estimate individual-level protection [23–25], while transmission models have incorporated profiles for vaccine-induced and infection-induced immunity over time to estimate the direct and indirect impact of different vaccination and dosing strategies [26–28]. Planning for these scenarios, particularly considering potential future variant characteristics, has been identified as a global public health priority [29,30]. The aim of our study was to estimate the infections, hospitalisations, and deaths averted at the population level of a range of future COVID-19 vaccination strategies compared to a baseline of no further boosting. In contrast to other transmission modelling studies, we developed a transmission model that explicitly incorporates hybrid immunity (immunity induced by both infection and vaccines), in order to capture the interactions between past exposure and vaccination. We did this by embedding an existing within-host model of underlying immunity dynamics and protection against infection and severe disease (similar to that presented in Khoury and colleagues [24]), which has been previously fitted to vaccine effectiveness data from England [31] within a population-based virus transmission model for SARS-CoV-2. We used this model to explore the impact of different targeted booster strategies—evaluating age-based targeting, different frequencies of boosting, and the value of using variant-adapted vaccines (both the 2022 bivalent products and theoretical yearly updated vaccines)—under the assumption that the Omicron variant continues to gradually evolve. We additionally consider the potential impact of the emergence of a new variant and the likelihood that vaccination could sufficiently mitigate its impact.

Discussion Assuming continued circulation of the Omicron variant with a degree of virus drift, our simulations mimic continuation of the pattern of low levels of SARS-CoV-2 circulation from late 2022 onwards, with waves of infection, hospitalisation, and deaths driven by the continued evolution of the virus. Our results indicate that the greatest impact on this endemic prevalence can be achieved through regular boosting of the 10+ years population. However, the efficiency and cost-effectiveness of a boosting programme depends on the outcome measure; a strategy targeting only 75+ years averts the largest number of deaths and hospitalisations per dose, whereas a strategy targeting 10+ has the largest reduction in infections but is relatively inefficient in reducing severe outcomes. Similar patterns were obtained regardless of whether the country has previously experienced large waves of infection (and therefore has considerable infection-induced immunity) or whether the country had previously pursued a zero-COVID policy. However, in the zero-COVID policy setting we generally estimate higher numbers of hospitalisations and deaths compared to settings with both high prior transmission and vaccine coverage. The health burden is higher because vaccine-induced immunity alone is estimated to be less protective than the combination of vaccine- and infection-induced immunity (or hybrid immunity) [33,50,51]. Cost-effectiveness will likely be the metric driving future vaccine strategies. Our results suggest that across all settings considered, targeting the highest risk group is likely to be the most cost-effective strategy as judged by the cost of preventing either a hospitalisation or a death. With variant-adapted vaccines now routine in high-income settings and also being used in some lower-income settings, we estimate that switching to such a variant-adapted vaccine could reduce the cost of preventing a hospitalisation or a death by around half. Switching to a yearly updated variant-adapted vaccine—as indicated by recently released United States Food and Drug Administration (FDA) guidance for manufacturers [52]—is projected to increase the cost-effectiveness by approximately 30%. It should be noted that our estimates of variant-adapted vaccine effectiveness are based on immunogenicity studies and will therefore be sensitive to our fitted relationship between the underlying immunological mechanism and protection. The results will also be sensitive to our assumed continued level of antigenic drift. Furthermore, to capture the full cost-effectiveness further information is needed on variant-adapted and yearly updated vaccine effectiveness and the comparative unit price of new products. Importantly, we found that even continuing administration of the ancestral vaccine products (or existing variant-adapted vaccines where they have been introduced) is likely to reduce infections and severe outcomes in all settings. Furthermore, while estimating cost-effectiveness based on reductions in hospitalisations and deaths is relatively straightforward, such analyses do not account for the impact of high infection levels on long COVID incidence. COVID-19 hospitalisations and deaths in HICs have been concentrated in elderly populations; in contrast, long COVID is reported across a wider age range [53,54]. Comprehensive cost-effectiveness analyses therefore need to consider the potential longer-term effects of this illness on quality of life and future productivity. Our analysis is caveated by the uncertainty in the timing and impact of any new variant. By definition, any variant that can replace the currently circulating Omicron variant will either need to be more transmissible or exhibit significant immune escape. Given that antigenic mapping studies suggest that, to date, there is no clear pattern of antigenic drift [38–40], our assumptions should be regarded as plausible but illustrative rather than predictive. Furthermore, translating the antigenic cartography into its associated impact on transmissibility and/or immune escape remains difficult and further research is needed to better quantify this relationship. In addition, there is concern that a new variant could exhibit the increased severity seen with Delta. Our results illustrate that, under a worst-case scenario, an epidemic wave of similar magnitude to those experienced in the first year of the pandemic could occur, even with regular boosting to the highest risk age groups using variant-adapted vaccines. This ongoing uncertainty provides a further challenge in valuing vaccination programmes; while widespread boosting could mitigate the impact of a new variant and would be substantially more cost-effective if it did arise, such a boosting strategy is inefficient and therefore unlikely to be cost-effective if such a variant does not emerge. It will therefore be important for countries to consider other mitigation strategies such as timely provision of antivirals. Our study has several limitations. First, the timing and magnitude of waves of SARS-CoV-2, the dominant circulating variant during these waves (particularly over the past 12 months), the timing and stringency of non-pharmaceutical interventions, and the vaccination response, has varied widely between countries. Our results are therefore illustrative and more detailed country-specific modelling will likely be required. Second, our immunological model is necessarily a simplification of the complex underlying immune response. The quality and durability of this response will likely vary by age; however, there are currently insufficient data to explore the impact of age on waning efficacy or immune escape from booster doses due to the shorter follow-up in younger populations. Furthermore, the degree to which prior immunity protects against future variants (including the currently circulating Omicron subtypes that are antigenically distinct from BA.1 estimates in the data used here [38,40]) remains uncertain. The durability of infection-induced immunity compared to vaccine-induced immunity remains uncertain; while studies using antibody data as a surrogate of protection have suggested that vaccine-induced immunity can be more durable [23,55], there is a growing body of evidence demonstrating the importance of cell-mediated immunity in providing longer-term protection for both vaccine-induced and infection-induced immunity [56–58]. It has been suggested that immune imprinting could reduce the effectiveness of continued boosting across different populations [59]. This effect was not included in our model and would lower the public health impact and cost-effectiveness of the vaccination strategies that we considered. Finally, we only provide illustrative costing metrics as a first step towards broader cost-effectiveness analyses. Such analyses will depend on longer-term follow-up of the quality of life and persistence of disability following both mild infections and hospitalisation. Our analyses illustrate the importance of continued booster doses as part of the wider public health response to ongoing endemic transmission of SARS-CoV-2 [23,60]. Prioritising boosters to high-risk and older populations is an efficient strategy in terms of reducing hospitalisations and death, while managing finite healthcare resources, but further data are required to understand the cost-effectiveness of vaccinating a wider age group to protect against the consequences of long COVID.

Supporting information S1 Text. Supplementary details including further mathematical description of the model and parameters. Fig A. Schematic diagram illustrating the COVID-19 vaccine allocation algorithm. Fig B. Modelled trajectories of the reproduction number R t over time. Fig C. Exemplar demographic patterns for each of the 2 income settings. Fig D. Impact of vaccination in a high-income country setting with substantial prior transmission and high vaccine access, assuming no additional variant emergence beyond Omicron (i.e., constant transmission and no additional immune escape, or no “drift”). Fig E. Impact of vaccination in a lower-middle-income country setting with substantial prior transmission and moderate vaccine access, assuming no additional variant emergence beyond Omicron (i.e., constant transmission and no additional immune escape, or no “drift”). Fig F. Impact of vaccination in a high-income country setting with minimal prior transmission and high vaccine access (Category 3). We assume mRNA-1273 is implemented for the first 2 doses and the first booster (dose 3) and a variant-adapted vaccine for subsequent booster doses. Fig G. Impact of vaccination in a low-middle-income country setting with substantial prior transmission and low vaccine access, where individuals 40+ years are initially targeted. Fig H. Impact of vaccination in a lower-middle-income country setting with substantial prior transmission and moderate vaccine access (Category 2), assuming WHO coverage targets. Fig I. Comparison of vaccine impact for different ancestral and variant-adapted vaccine scenarios for the lower-middle-income country setting with substantial prior transmission (Category 2). Fig J. Comparison of vaccine impact for different ancestral and variant-adapted vaccine scenarios for the high-income country setting with minimal prior transmission (Category 3). Fig K. Impact of vaccination in future scenarios where an additional variant of concern emerges from 1 October 2022, in a high-income setting with minimal prior transmission and high vaccine access (Category 3). We assume a variant-adapted vaccine is implemented from dose 4. Fig L. Sensitivity of the model output to the level of “drift” in a high-income setting with substantial prior transmission (Category 1). Fig M. Sensitivity of model results to assumptions regarding the level of protection afforded by infection in a high-income setting with substantial prior transmission (Category 1). Fig N. Sensitivity of model results to the decay rate of protection following recovery from SARS-CoV-2 infection in a high-income setting with substantial prior transmission (Category 1). Table A. Prior and posterior parameter estimates for the immunological model. Table B. Transmission model state transitions. Table C. Transmission parameter description and values. Table D. Default scenarios and assumptions for vaccine uptake within targeted age groups for each of the 2 income settings. Table E. Default scenarios and assumptions for the 3 broad modelled categories of country and epidemiological state. Table F. Total doses delivered, infections, hospitalisations, and deaths for each of the Category 1 and 3 settings, for a range of vaccine dose strategies and variant scenarios. We assume mRNA-1273 is implemented for the first 2 doses and the first booster (dose 3), and a variant-adapted vaccine for subsequent booster doses. Table G. Total doses delivered, infections, hospitalisations, and deaths for each of the Category 1 and 3 settings, for a range of vaccine dose strategies and variant scenarios. We assume no additional variant emergence beyond Omicron (i.e., constant transmission and no additional immune escape, or no “drift”). Table H. Total doses delivered, infections, hospitalisations, and deaths for the Category 2 setting, for a range of vaccine dose strategies and variant scenarios. We assume AZD1222 is implemented for the first 2 doses, mRNA-1273 for the first booster (dose 3), and a variant-adapted vaccine for subsequent booster doses (doses 4 and 5). Table I. Total doses delivered, infections, hospitalisations, and deaths for the Category 2 setting, for a range of vaccine dose strategies and variant scenarios. We assume no additional variant emergence beyond Omicron (i.e., constant transmission and no additional immune escape, or no “drift”). Table J. Total doses delivered, infections, hospitalisations, and deaths for the Category 2 setting, for a range of vaccine dose strategies and variant scenarios, assuming the WHO coverage target assumption. We assume AZD1222 is implemented for the first 2 doses, mRNA-1273 for the first booster (dose 3), and a variant-adapted vaccine for subsequent booster doses (doses 4 and 5). Table K. Total additional infections, hospitalisations, and deaths averted, and total additional vaccine doses delivered for the Category 1 and 3 settings. We assume no additional variant emergence beyond Omicron (i.e., constant transmission and no additional immune escape, or no “drift”). Table L. Total additional infections, hospitalisations, and deaths averted, and total additional vaccine doses delivered for the Category 2 setting. We assume no additional variant emergence beyond Omicron (i.e., constant transmission and no additional immune escape, or no “drift”). Table M. Total doses delivered, infections, hospitalisations, and deaths for the Category 2 setting, for a range of vaccine dose strategies, for the scenario where individuals 40+ years are initially targeted. We assume AZD1222 is administered for all doses. Table N. Total doses delivered, infections, hospitalisations, and deaths for different ancestral, variant-adapted, and yearly updated vaccine scenarios. Table O. Impact of vaccination in future scenarios where an additional variant of concern emerges from 1 October 2023. https://doi.org/10.1371/journal.pmed.1004195.s001 (DOCX)

Acknowledgments We thank Bob Verity, Nick Grassly, Sarah Pallas, and the WHO SAGE working group on COVID vaccines for helpful comments and suggestions on earlier versions of this work. For the purpose of open access, the author has applied a “Creative Commons Attribution” (CC BY) licence (where permitted by UKRI, “Open Government Licence” or “Creative Commons Attribution No-derivatives (CC-BY-ND) licence” may be stated instead) to any Author Accepted Manuscript version arising.

[END]
---
[1] Url: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004195

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.