Long-Term Changes in Vaccination Protection Against Severe Courses of COVID-19 by Non-adapted MRNA-Based Vaccines

A retrospective observational study in elderly persons during the XBB wave

In Germany, the monovalent mRNA-based vaccines against older SARS-CoV-2 variants were replaced in autumn 2022 by bivalent vaccines targeting the Omicron strains BA.1 and BA.4/BA.5 . In February 2023, however, a new variant—XBB and its sublineages—became the leading cause of infection, and it was not until November 2023 that XBB again caused fewer than 25% of infections. However, XBB-adapted vaccines were not licensed until 18 September 2023. Although non-adapted vaccines are acutely effective against XBB, the durability of their efficacy against XBB is still uncertain. In the research presented here we analyzed the extent to which non-adapted vaccines provided long-term protection against severe courses of COVID-19 (relative to the degree of protection immediately after vaccination) during the XBB wave.

Methods

We extracted the data from all Bavarian cases of SARS-CoV-2 infection between 1 February and 30 November 2023 that were registered in accordance with the German Infection Protection Act (Infektionsschutzgesetz, IfSG). The following exclusion criteria were applied: not fully vaccinated, not reported to have received homologous mRNA-based vaccination schemes, vaccination after 18 September 2023. We analyzed only cases in persons aged 60–99 years because of their increased risk of severe COVID-19. Comparisons with incompletely vaccinated/unvaccinated cases were part of a previous study (1).

The outcome variable was time to hospitalization or death up to day 60 after the date of infection (resulting in an observation period from 1 February 2023 to 1 February 2024). The infection date was defined as the earlier of the two dates on which either the infection was reported or the first symptoms appeared. By means of Cox regression we calculated adjusted hazard ratios (aHR) with 95% confidence intervals (CI) for the risk of a severe outcome depending on the time from the onset of immunity immediately after vaccination to the date of infection. The methods and their limitations are described in detail elsewhere (1). All analyses were performed using the program R 4.3.1. The analysis code and further results can be found at https://osf.io/7s6jy/.

Results

After application of the exclusion criteria, 2125 out of 128 400 registered cases were available for analysis (1.7%). A severe outcome was observed in 359 cases (16.9%) (Table). With adjustment for potential confounders, the risk of a severe course increased only slowly with increasing time since immunity onset; for example, the aHR after 18 months was 1.43, 95% CI: [1.14; 1.80] (Figure). A later infection date was associated with an increasing risk of a severe course (compared with the start of the observation period). Increasing age (80 versus 60 years: aHR 12.39 [9.99; 15.37]), male sex (1.53 [1.24; 1.89]), and a known previous SARS-CoV-2 infection (1.29 [1.01; 1.65]) were associated with an increased risk, while a boosted level of immunity was associated with a decreased risk (0.69 [0.51; 0.95]).

Figure

Relative change in immune protection

Table

Demographic characteristics and proportions of severe COVID-19 courses*

Discussion

To our knowledge, this is the first observational study evaluating homologous, non-adapted mRNA vaccination programs and their potential loss of effectiveness over time in preventing severe COVID-19 courses during the Omicron XBB wave. Our results suggest that even non-adapted mRNA-based vaccination programs can induce long-lasting protection against severe manifestations of COVID-19 caused by the XBB variant (including sublineages), likely based on long-term changes in the T-cell system (recognition of the XBB variant and its sublineages).

In contrast to our results, a recently conducted European multicenter test-negative case–control study in persons aged ≥ 60 years reported that during the XBB wave the relative vaccine effectiveness of non-adapted vaccines against hospitalization declined from 80% in the first 89 days to 15% after 90–179 days and that there was no protection at all after 270–359 days (2). The differences from our results may be explained by possible weaknesses in the design of the European study (e.g., dependence on the sensitivity and specificity of the SARS-CoV-2 tests and/or selection bias) (3). In addition, there was no adjustment for previous SARS-CoV-2 infections and no differentiation between homologous and heterologous vaccination schemes. The study therefore included persons vaccinated with vector vaccines, which have been found to provide less protection than mRNA vaccines against hospitalization and death more than 120 days after the first vaccination (4). On the other hand, our results are consistent with a study from Singapore that showed, in a setting similar to ours, stable protection against COVID-19-related hospitalization even 360 days and more after receipt of booster doses (5).

One limitation of our study is that a small proportion of cases could have been caused by SARS-CoV-2 variants other than XBB, possibly leading to overestimation of vaccine efficacy. Furthermore, cases with a high intrinsic risk for a severe course (e.g. those suffering from significant comorbidity) were over-represented in the IfSG data (and this over-representation increased during the course of 2023). However, our adjustment for infection date, known previous SARS-CoV-2 infections and age, should have minimized the bias caused by this over-representation. Even if we had not adjusted sufficiently for severe comorbidities, our results would still mean that protection is broadly stable, at least in high-risk cases.

Andreas Beyerlein, Katharina Katz, Helmut Küchenhoff*, Wolfgang Hartl*

Conflict of interest statement
The authors declare that no conflict of interest exists.

Submitted on 21 March 2024, revised version accepted on 26 July 2024

Cite this as:
Beyerlein A, Katz K, Küchenhoff H, Hartl W: Long-term changes in vaccination protection against severe courses of COVID-19 by non-adapted mRNA-based vaccines—a retrospective observational study in elderly persons during the XBB wave. Dtsch Arztebl Int 2024; 121: 814–5. DOI: 10.3238/arztebl.m2024.0179

References

1. Beyerlein A, Weigert M, Katz K, Küchenhoff H, Hartl W: Long-term trends in the protection against severe courses of COVID-19 by vaccination— a retrospective observational study in older persons during the Omicron wave. Dtsch Arztebl Int 2023; 120: 873–8.

2. Antunes L, Mazagatos C, Martinez-Baz I, et al.: Effectiveness of the adapted bivalent mRNA COVID-19 vaccines against hospitalisation in individuals aged ≥ 60 years during the Omicron XBB lineage-predominant period: VEBIS SARI VE network, Europe, February to August, 2023. Euro Surveill 2024; 29: 2300708.

3. Sullivan SG, Khvorov A, Huang X, et al.: The need for a clinical case definition in test-negative design studies estimating vaccine effectiveness. NPJ Vaccines 2023; 8: 118.

4. Black B, Thaw DB: Vaccinating against a novel pathogen: a critical review of COVID-19 vaccine effectiveness evidence. Microorganisms 2023; 12: 89.

5. Wee LE, Pang D, Chiew C, et al.: Long-term real-world protection afforded by third mRNA doses against symptomatic severe acute respiratory syndrome coronavirus 2 infections, coronavirus disease 19-related emergency attendances and hospitalizations amongst older singaporeans during an omicron XBB wave. Clin Infect Dis 2023; 77: 1111–9.