key: cord-0980209-wfkprxx5
authors: Hunziker, P.
title: Minimizing loss of life in Covid-19 in a 100 day period in the U.S.A. by age-tailored dosing and distribution of a limited vaccine supply
date: 2021-02-01
journal: nan
DOI: 10.1101/2021.01.30.21250834
sha: 470fcccdcdd7ef406bdf57a3e882132c9b1242a0
doc_id: 980209
cord_uid: wfkprxx5

Background: We aimed at minimizing loss of lives in the Covid-19 pandemic in the USA by identifying optimal vaccination strategies during a 100-day period. While lethality is highest in the elderly, transmission and case numbers are highest in the young. Due to limited vaccine availability, a strategy of first vaccinating the elderly is widely used. Vaccines elicit enhanced immune responses in younger persons compared to the elderly, but even in the latter, mRNA vaccines are effective, implying that in the young, lower vaccine doses might suffice for protection, thereby increasing the number of persons that can be vaccinated. Methods: Using published immunogenicity data of the Moderna mRNA-1273 vaccine, we examined the value of tailored-dose vaccination strategies, using a modeling approach that incorporates age-related vaccine immunogenicity, social contact patterns, population structure, Covid-19 case and death rates in the U.S. in late January 2021. Reduced dose vaccination in the young multiplies the number of individuals vaccinated early. Results: Age-tailored dosing strategies led to faster case reduction and thereby shortened the pandemic, reducing the delay to reaching <100'000/day from 64 to 30 days and avoiding 25'000 deaths within 100 days in the USA. While regular vaccination of the elderly can continue, the main protection of the elderly resulted from stopping the pandemic by vaccinating a large number of young persons by age-adapted dosing. Conclusion: Rapid reduction of Covid-19 case and death rate in the U.S. can be best achieved by using highly effective vaccines in an age-tailored dosing regimen. Protecting the vulnerable is most effectively achieved by dose tailored vaccination of all population segments, while an 'elderly first' approach costs more lives, even in the elderly.

Vaccines against SARS-Cov2 have been developed at warp speed, and mRNA vaccines like the Pfizer BNT162b2 1 (Tozinameran) and the Moderna mRNA-1273 2 have shown strong immunogenicity, safety and efficacy against disease. Emerging data also show protection against virus infection (and thus, transmission) by the immune response to natural infection 3 and, presumably, vaccination 4 . While global case numbers and deaths reached new peaks in January 2021, manufacturing lines are not currently capable to cover the huge global demand, calling for optimally effective strategies for their use. While phase I-II data have confirmed the old knowledge that elderly people have a decreased immune response to vaccines, the vaccines have been dose-optimized such that they achieve an optimal immunity even in the elderly, implying that for the young, the vaccine dose may be redundant. To achieve proof of efficacy quickly, pivotal trials were then also performed at this single dose level. This raises the question, if in the younger, sufficient immunity can be reached with lower vaccine doses 5 . Faced with the severe availability issues of the vaccines in winter 2021, a lower dose per vaccination could translate into a significantly larger number of people receiving the vaccine early. As the young, due to their frequent social contacts, drive the pandemic to a large degree, vaccinating them early may prove to be a game changer for stopping the pandemic and reducing the number of cases and deaths invariably associated with an excessive pandemic duration.

Moderna vaccine has been performed towards optimal protection of the elderly, exploring doses of 25, 50, 100 and 250ug. Laboratory-assessed immunity levels typically exceeded those seen in the plasma of reconvalescent patients who have a protection of 83% for at least 5 months 6 . Immunogenicity in the young is even higher than in the elderly. Vaccination study populations are protected against infection at least four months despite some decline of the measured immunity parameter in the elderly. We noted that in the young, a 25µg dose of the Moderna vaccine elicited an immune response level at day 57 that was comparable to the immune response seen in patients older than 71 years at day 119 (Table) , a group in which the vaccine achieves >86% protection. The interpretation that good immunogenicity translates into good protection is very plausible, 7 ,. We therefore hypothesized that exploiting age-tailored vaccination dosing may allow multiplying the proportion of persons early and thereby may lead to improved pandemic control.

Data reported by Moderna on phase I 8 , 9 , 10 , 11 , 12 II and III 13 studies were analyzed. Population size and age structure was modeled according to the U.S.A with a population of 332'599'000, split into a cohort of 54'303'000, "old" persons > 64 years and of 278'296'000 "young" persons ≤64, according to U.S. government data 14 . Covid-19 case numbers were from the U.S. Center for Disease Control 15 , the Johns Hopkins University CSSE dataset 16 and the Oxford university "our world in data" repository 17 and were used to initialize the model to 193'717 cases per day as per January 20, 2021. Newly infected persons <ds infectious from day 1 to day 7. One million vaccine doses per day are available. Protection by vaccination occurrs from day 10 on. In the main analysis, protective efficacy against reinfection the 100µg vaccine dose was set to 95·6% in the young and 86·2% in the elderly as published; the vaccine efficacy of a 25µg dose in the young was set to 86·2% based on the levels of immunogenicity achieved in the young compared to the immune response in the elderly vaccinated with 100µg as shown in the table. In a further exploratory analysis, protective efficacy of natural infection and vaccination was set to 30% only. Based on known differences in social contact behavior between age groups 18 , young persons were set to have 80% of their social contacts with the "young" and 20% with the "old", while for the old, contacts to other elderly and the young were each set to be 50%. Using risk contacts at study start (derived from the number of non-immune, the number of infectious, the number of cases in each age segment) as reference, we used the change in contacts to compute the daily transmissions in each age group, whereby encounters of non-immune persons with infectious persons were considered "risk contacts", while encounters of immune persons with infectious persons were considered as "semi-risk contacts", that were weighted according to the protection afforded by the vaccine in that age segment. Deaths on a given day were computed from the daily case count 14 days before, using the case fatality rate in the U.S. in January 2021, approximately 1·6%, according to the relatively stable case and death counts in this period. The age-dependent Covid-19 death distribution was derived from the Center for Disease Prevention and Control data 19 , indicating a case fatality rate of 0.35% for the young and 8.9% for the elderly and supported by other analyses 20 . The following scenarios were tested: "elderly first": starting with regular dose vaccination until 80% of the elderly are covered, then vaccinating the young at regular dose.

"young first": starting with regular dose vaccination in the young, leaving the elderly aside during the first 100 days "adaptive": in parallel, using half the stock for each, vaccinate the elderly at full dose and the young at quarter dose "adaptive, the young first": starting with quarter dose vaccination in the young, leaving the elderly aside until 80% of the young are vaccinated, then vaccinating the elderly at full dose.

Ethics: The relevant Ethical Kommittee Nordwestschweiz declared that computer modeling studies do not fall under the jurisdiction of Ethical Committees.

Baseline results in the "elderly first" strategy, using standard vaccine dosing, predict a cumulative death count of 153'000 cumulative deaths over 100 days. Case numbers fall below 100'000/day on day 64 and the daily death rate falls below 1'000/day on day 55, as shown in Figure. In the elderly, death rates initially fall fastest compared to other scenarios, but later on, the significant exposure of the elderly to the large number of infectious persons, when vaccination has not yet achieved the larger, younger, socially active population segments, and taking into account the protective efficacy of the vaccine in the elderly being less than 100%, leads to higher mortality in the elderly during the later phases of the pandemic, compared to other scenarios.

In an alternative scenario where the young are vaccinated first, at standard dose vaccine, case numbers fall faster in the young, but at the expense of higher mortality in the unprotected elderly cohort throughout most of the study period, yielding a higher overall death count of 184'000, with case rates falling below 100'000/day on day 42 and death rates falling below 1000/day on day 70.

In the "adaptive" strategy, with half the vaccine stock used for the elderly at full dose, and the other half used for vaccinating a four-fold larger number of persons ≤64 years at a quarter dose, case numbers fall faster than with either prior strategy, and while mortality in the elderly is initially only slightly higher than in the "elderly first" strategy, mortality, even in the elderly, falls below the one observed in the other strategies, resulting in markedly lower cumulative deaths of 128'000 in 100 days. The milestones of <100'000 deaths/day are reached on day 30, and of <1000 deaths/day on day 49, significantly faster than in the prior scenarios. This improvement is due to the strong reduction in risk contacts, as well as the "semi-risk" contacts (a partially immune meets an infectious person) in the young as well as in the elderly, compared to the prior scenarios, as shown in Figure Panels C-E.

Limiting the vaccine campaign to the young and vaccinating them at quarter dose leads to an even faster reduction of risk contacts and case numbers, but with a similar cumulative number of deaths: 121'000 in 100 days, achieving <100'000 cases/day in 22 days and <1000 death/day in 45 days. However, this approach, as effective as it is, might convey a sense of injustice.

Sensitivity of the results were tested in particular for a varying effectiveness parameter because protective effects of the vaccines against virus transmission are still sparse: assuming a protective efficacy against infection (and transmission) of quarter dose vaccination and of natural infection of only 30%, while full dose vaccination yielded infection protection levels equal to the published disease protection levels, the adaptive strategy yielded similar cumulative death numbers (162'000 vs 163'000) and a shortened time to <100'000 cases (50 vs 82 days) compared to the "elderly first" strategy. At protection

The Covid-19 pandemic calls for decisive action to minimize excess deaths and long-term sequelae of the disease, protect healthcare facilities, and minimize the damage to the economy. Vaccination is a cornerstone for mastering the pandemic but identifying the optimal vaccination policies is a prodigious task. Here, we use the demographics and current epidemiologic data from the United States together with age-dependent social interaction patterns to build a predictive country-scale model and combine it with age-dependent immunity responses observed in the early clinical studies of the Moderna mRNA vaccine, similar to separate analyses 21 focusing on Europe. While the most frequently chosen policy is "protect the vulnerable" implemented as "vaccinate the elderly first", we find that the vulnerable are best protected by protecting society as a whole through a broad vaccination strategy. This was achievable by exploiting the excellent immunogenic properties of the available mRNA vaccine through fractional vaccine dosing in the young, an approach that proved preferrable even when the lower efficacy of the reduced dose against infection and transmission is factored in.

Two of the studied scenarios, namely "adaptive: split the vaccine stock to vaccinate the elderly at full dose and use it at quarter dose to vaccinate as many young as possible", as well as the scenario "Vaccinate the young first, at quarter dose" excelled in terms of shorting the pandemic, minimizing the number of cases and of deaths. As the "justice" of the latter strategy may be difficult to convey to the public because the "vulnerable" seem to be left out (although they even benefit from the approach), the adaptive approach seems to be a preferrable policy.

The strong efficacy of the Moderna vaccine already at moderate immune response level, evident by its effectiveness in preventing disease within at 10-14 days after the first standard dose, before full immune response is achieved, supports these findings. Also, prior infection significantly protects against reinfection 22 , despite antibody levels in reconvalescents that are lower than those reported with quarter dose vaccination 23 . Emerging data from Israel indicate that the immune response elicited in the elderly, elicited by a single dose of Pfizer's mRNA vaccine mediates a degree of protection against infection and transmission 24 , although the age and the fact of having received only a single dose in this time window imply less than optimal antibody levels. "Fractional" dose vaccination has proven beneficial in viral poverty diseases 25 , 26 further supporting the findings of this study.

The social interaction patterns have a significant impact on "risk contacts" in Covid-19, as known from literature 27 , 28 , 29 and again shown in this study. Even imperfect immunity conveyed to a significant proportion of the age groups that fire the pandemic most is therefore highly desirable.

Mutant viruses are projected to represent the next challenges. As it is probable that mutants arise in persons who are unable to eliminate the virus, but infection coinciding with a not-yet full immune response in the healthy leads to strong subsequent immunity, we believe that using reduced-dose vaccination will not lead to more mutants; in contrast, fastest mastering of the pandemic probably is the keystone to prevent further mutants from emerging. As the Moderna vaccine preserves activity to the prevalent variant of concern B.1.1.7 30 we believe that the strategy delineated here is reasonable at this time. As further mutants that might evade the immune response from prior infection or by the current vaccines are expected soon 31 and may require additional shots with modified vaccines, such an iterative "booster" again encoding the spike protein will further enhance the protection conveyed by prior fractional dose vaccinations even if the target molecule is slightly modified. Less strain on the currently overloaded production lines by a fractional dose approach may even free some resources to produce new, optimized vaccine batches that cover variants of concern, faster.

Study limitations include assumptions that stem from phase I and II studies of limited size, including the extrapolation on clinical efficacy based on comparing measured immune titers. Preferably, the findings of this study are scrutinized by well-designed trials although such trials would also need to be performed at "warp speed". The most straightforward way is to allocate cities within a country to the adaptive approach proposed and use daily cases, deaths, hospital and ICU occupancy as continuously available endpoints, allowing to draw conclusions and adapt policies early. The "off-label use" character of this approach needs to be mentioned, calling for suited permits before its clinical application. While we base our analysis on the Moderna vaccine, it seems reasonable that the excellent efficacy, and All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted February 1, 2021. ;  thus amenability to fractional dosing, is a class effect of the currently available mRNA vaccines, potentially allowing extension of the findings to other mRNA vaccines, e.g., the Pfizer Tozinameran vaccine, although this will require more detailed analyses including access to detailed immunogenicity data from the trials performed up to now.

This modeling study demonstrates that adaptive vaccination strategies that rely on an age-tailored vaccine dosing of a highly effective mRNA vaccine, applied to all population segments, may markedly outperform standard dose regimens that are initially focused on the elderly. By multiplying the number of persons that can be vaccinated early, this approach limits society-wide transmission earlier, and shortens pandemic duration and markedly lowers case counts and death rates, even in the vulnerable elderly.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in

The copyright holder for this this version posted February 1, 2021. ;

Covid-19 case numbers and deaths, and "risk contact" propensity (i.e., a susceptible meets an infectious person) during 100 days of a vaccination campaign modeled for the U.S.A. "elderly first" uses full dose vaccination in the elderly until 80% of the elderly are vaccinated, then switchin to the young at full dose. "young first" vaccinates only the young, at full dose. "adaptive" uses full dose in the elderly and in parallel, quarter dose in the young. "young quarter" only vaccinates the young, at quarter dose, until 80% of the young are vaccinated. Panel A: Covid-19 cases per day, comparing the different scenarios. Note that using a reduced (quarter) dose vaccine reaches a larger proportion of the population faster, and even with a potentially reduced efficacy per individual person, the societal impact, measured as case number reduction, is best for an adaptive strategy incorporating reduced dose vaccination, and for the strategy that initially focusses a reduced dose vaccination solely on the young. Panel B: Death number reduction is initially faster by an "elderly first", and to some degree by the "adaptive" strategy, while later on, the broadest approach, namely vaccinating primarily the young, has the largest impact because it stops the pandemic the fastest. Panel C, D, E: Propensities of risk contacts (i.e., a non-immune person meets an infectious person) and "semi-risk contacts" (i.e., a vaccinated or previously infected person meets an infective person, acknowledging that the protection provided is less than 100%) compared to day 1. As in the strategy "elderly first" an unchecked virus propagation in the other population segment occurs, there is a rapid decay of "risk contacts" for the elderly, but the immunized elderly will encounter large numbers of infected young persons, maintaining a residual risk. In contrast, the strategies that include the young from the beginning lead to significant reduction in "risk" and "semi-risk" encounters, in particular if the quarter dose vaccination is part of the strategy. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in

The copyright holder for this this version posted February 1, 2021. ; https://doi.org/10.1101/2021.01.30.21250834 doi: medRxiv preprint

Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine

Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine

Antibody status and incidence of SARS-CoV-2 infection in health care workers

Are COVID vaccination programmes working? Scientists seek first clues

Looking beyond COVID-19 vaccine phase 3 trials

Antibody status and incidence of SARS-CoV-2 infection in health care workers

Antibodies to SARS-CoV-2 are associated with protection against reinfection

An mRNA vaccine against SARS-CoV-2-preliminary report

Safety and immunogenicity of SARS-CoV-2 mRNA-1273 vaccine in older adults

Durability of responses after SARS-CoV-2 mRNA-1273 vaccination

An mRNA vaccine against SARS-CoV-2 -preliminary report

mRNA-1273 Clinical Development Program. Moderna presentation

Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine

Coronavirus Pandemic (COVID-19). 2020. Published online at OurWorldInData.org. Retrieved from

Using data on social contacts to estimate age-specific transmission parameters for respiratory-spread infectious agents

Estimating the infection-fatality risk of SARS-CoV-2 in New York City during the spring 2020 pandemic wave: a model-based analysis. The Lancet Infectious Diseases

Vaccination strategies for minimizing loss of life in Covid-19 in a Europe lacking vaccines

Antibody status and incidence of SARS-CoV-2 infection in health care workers

An mRNA vaccine against SARS-CoV-2-preliminary report

Are COVID vaccination programmes working? Scientists seek first clues

Boosting immune responses following fractional-dose inactivated poliovirus vaccine: a randomized, controlled trial. The Journal of infectious diseases

Duration of post-vaccination immunity to yellow fever in volunteers eight years after a dose-response study

Vaccine optimization for COVID-19: who to vaccinate first? medRxiv 2020

Age-targeted dose allocation can halve COVID-19 vaccine requirements. medRxiv2020

Determinants of SARS-CoV-2 transmission to guide vaccination strategy in a city

mRNA-1273 vaccine induces neutralizing antibodies against spike mutants from global SARS-CoV-2 variants

Genomic characterisation of an emergent SARS-CoV-2 lineage in Manaus: preliminary findings

An mRNA vaccine against SARS-CoV-2-preliminary report

Durability of responses after SARS-CoV-2 mRNA-1273 vaccination

Vaccination strategies for minimizing loss of life in Covid-19 in a Europe lacking vaccines