Protection duration after vaccination or infection

In addition to antibody responses, T cell responses can be detected in convalescents that focus on the recognition of Spike, Membrane, Nucleoprotein, nsp3 and ORF3a of SARS-CoV-2 ^3,9. Spike, Membrane, Nucleoprotein and ORF3a specific CD8⁺ T cell responses decay with a half-life of 120-190 days ^2,3, and Spike, Membrane, Nucleoprotein, ORF3a and nsp3 specific CD4⁺ T cell responses slightly faster with a half-life of 140-150 days ^2,3. These responses complement B cell responses in preventing severe disease after SARS-CoV-2 reinfection, but most likely cannot prevent SARS-CoV-2 infection and transmission as effectively as (neutralizing) antibody responses.

2. Estimated kinetics of protection against re-infection with mild or severe symptoms

Although a true correlate of protection from SARS-CoV-2 infection after vaccination or after recovering from a first infection has not yet been established, it is highly likely that neutralizing IgG are substantially involved. By correlating the neutralizing antibody titers after vaccination or after SARS-CoV-2 infection with (re-)infection outcomes, the level of antibody titers associated with 50% protection from asymptomatic/mild infection versus severe infection were estimated and compared to SARS-CoV-2 naïve and unvaccinated individuals, respectively ¹. Khoury et al propose that a level of 20% of the initial average antibody titer after convalescence is associated with 50% protection against asymptomatic/mild infection and 3% of the initial titer is required for 50% protection against severe infection¹. The respective study has a number of caveats that are also acknowledged by the authors. Namely, neutralizing antibody titers are used as protective correlate, while some individuals might clear SARS-CoV-2 infection primarily with T cell responses. Furthermore, neutralizing antibody responses were measured with different assays in the included studies. Thirdly, classification into re-infection with mild to moderate or severe symptoms uses different clinical symptoms in the analyzed studies. Fourthly, starting titers after vaccination might be higher after vaccination than after infection (2 to 4 fold¹⁰) and lower after convalescence of individuals with more advanced age if stratified according to disease severity ^10,11. Finally, SARS-CoV-2 variants with a relevant transmission advantage and/or partial antibody immune escape phenotype might infect vaccinees or convalescents at higher antibody levels and thus earlier after the initial infection episode or immunization ^12-14. Nevertheless, with keeping these confounding factors in mind, the antibody response to SARS-CoV-2 might decline to 20% of the initial level associated with 50% protection against mildly symptomatic re-infection with possible transmission after 240 days (8 months) in individuals <65 years of age. In covalescents above 65 years of age, protection against mild re-infection was shown to be more short-lived and dropped 50% within 3-6 months ¹⁰.  Assuming an antibody decline with a stable half-life of 100 days for 1.5 years reaching  a steady-state with a half-life of >10 years thereafter ¹, the threshold of 3% of the initial antibody titer associated with 50% protection against severe infection would be reached in convalescents <65 years of age after 16 to 18 months and in those above 65 years after 10 to 12 months, unless natural virus circulation or vaccination would boost immunity before. 

3. Estimated kinetics of mRNA-vaccine induced protection against infection with mild or severe symptoms

After vaccination with highly immunogenic mRNA vaccines initial neutralizing antibody titers are 2-4-fold higher than after natural infection and short-term protective immunity is at 95% in those populations enrolled in phase 2/3 clinical trials¹⁵. Population data from Israel have demonstrated short term BNT/Pfizer vaccine effectiveness of 94% against any SARS-CoV-2 infection and of 97% against severe infection also in people >85 years of age ¹⁶. Only for asymptomatic SARS-CoV-2 infection, there was a trend of reduced vaccine effectiveness according to age (83% in >85 years; 88% in 65-84 years; 91% in 45-64 years; 94% in 16-45 years). Together with recent data on some age dependence of vaccine induced neutralizing IgG titers¹⁷ this indicates that induction of protective immunity after mRNA COVID-19 vaccines is reduced in older individuals above 70-75 years of age although this seems to be much less pronounced than for other vaccines like influenza ¹⁸. Moreover, data on mid-term maintenance of protection of convalescents against SARS-CoV-2 re-infections indicate some age dependence as well: Above 65 years of age, protection decayed to 47% within 3-7 months whereas younger individuals maintained protection at >80% ¹⁰. Therefore, it is very likely that potential protective thresholds against severe COVID-19 are reached at earlier time points in people >70-75 years of age not only after recovering from natural infection but also after mRNA vaccination. According to the modelling of Khoury et al., individuals below 65 years of age should maintain >50% protection against mild infection for 500 days (16 months) or even longer after two doses of mRNA vaccines (Figure 3c, lower panels, starting efficacy 90-95% ¹) and against severe infection at >80% for more than 3 years as long as circulating virus variants do not exhibit significant antibody immune escape phenotypes. For severe infection outcomes we anticipate that a potential protective threshold of 50% is not sufficient from a public health perspective, since 50% of infected individuals would be at risk for hospital admission despite vaccination. We therefore extrapolated the data of Khoury et al. to provide assumptions of 70-80% protection for severe outcomes ¹. Assuming that antibody decay kinetics are largely age independent, the situation is likely to be different and comparable to starting efficacies of 80-90% in people >70-75 years of age (Figure 3c¹). Protection against mild infection would fall below the 50%-threshold within 200-300 days (7-10 months) and against severe infection below a 70% protection threshold within 450 to 750 days (15-24 months). However, after two doses of mRNA vaccines a 50% protection level against severe COVID-19 may also be maintained for more than 3 years in the elderly and very elderly. Therefore, the level of desired protective immunity will dictate at which time point vaccine booster doses may be required as long as natural virus circulation is very limited and circulation of (partial) immune escape variants does not warrant for earlier boosts with adequately adapted vaccines.

Specifically, if circulating virus variants exhibit significant antibody immune escape phenotypes the initial level of protection and its duration will be substantially reduced. For a 5-fold decreased potency for variant neutralization as estimated for B.1.617.2 (or delta variant)¹⁷, efficacy of mRNA vaccines starts at 77% instead of 95% protection and the 20% neutralizing capacity associated with 50% protection against mild to moderate infection is already reached after 200 days¹. More importantly, the threshold of 70% protection against severe infection may be reached within 300-500 days (10-18 months) in individuals <65 years age and even earlier those above 70-75 years. In this case, booster vaccinations with adapted mRNA vaccines containing the Spike protein sequence of relevant variants should be considered within one year and before anticipated epidemiologic acceleration during the cold season. Booster vaccinations with variant sequences are likely to induce neutralizing antibodies both against the variant and the original virus as shown after natural SARS-CoV-2 infection, where convalescent sera from patients infected with the immune escape variant B.1.351 contained broadly neutralizing IgG with comparable activity against B.1.351, the original virus and other SARS-CoV-2 variants¹⁹.