Dynamics of competing SARS-CoV-2 variants during the Omicron epidemic in England

The SARS-CoV-2 pandemic has been characterized by the successive emergence of genomic variants. High levels of natural and vaccine-induced immunity create selective pressure favoring variants capable of evading neutralizing antibodies. The Omicron variant, first detected in November 2021, demonstrated a high degree of immune evasion, leading to a global surge in infection rates. However, initial estimates of Omicron's impact relied heavily on routine testing data, which are susceptible to various biases. This research aimed to provide a more accurate assessment of the Omicron wave in England by utilizing data from the REACT-1 study, a series of cross-sectional surveys designed to evaluate SARS-CoV-2 infection prevalence. The study's importance lies in its potential to provide a more robust understanding of Omicron's dynamics, particularly its transmission rate and the impact of immune evasion, moving beyond potentially flawed routine testing data. This detailed understanding is crucial for informing public health strategies and predicting future pandemic waves.

The study references several previous works highlighting the regular emergence of SARS-CoV-2 variants of concern (VOCs) and their impact on the pandemic's trajectory. Specific mentions include the high transmissibility and immune evasion capabilities of Omicron, its milder disease compared to previous variants, and the reduction in vaccine effectiveness against Omicron infection and reinfection. The literature review underscores the need for accurate epidemiological data to effectively model and manage the pandemic's ongoing evolution.

The study employed data from rounds 14-18 of the REACT-1 study (September 9, 2021 – March 1, 2022), a large-scale, cross-sectional survey of the English population involving random sampling and self-administered throat and nose swabs. Samples with sufficiently low N-gene Ct values underwent next-generation sequencing to determine lineages using the Pangolin algorithm. The researchers used mixed-effects Bayesian P-spline models to estimate the daily prevalence of Delta and Omicron variants separately, and later BA.2 and non-BA.2 Omicron sub-lineages. Bayesian logistic regression models were used to assess growth rates. Phylogenetic analysis, using RAxML and IQTree, was conducted to investigate the relationships between variants and their geographic spread. Apple mobility data provided insights into the correlation between social contacts and infection rates. Statistical analyses included the Wilson method for calculating confidence intervals and logistic regression for analyzing symptom data. The methodology emphasizes a rigorous statistical approach to analyze a large dataset for a more accurate representation of SARS-CoV-2 variant dynamics in England.

Key findings include: 1. Omicron's rapid rise to dominance in England, reaching over 90% of cases by late December 2021. 2. An initial peak in national Omicron prevalence of 6.89% (95% credible interval: 5.34%, 10.61%) in January 2022. 3. A subsequent resurgence driven by the more transmissible BA.2 sub-lineage. 4. The daily growth rate advantage of Omicron over Delta declined over time, potentially due to factors such as shorter generation time and initial high transmission in younger, more socially active groups. 5. Consistent growth advantages across regions and age groups, although the timing of peak prevalence varied. 6. A significant decline in Delta's reproduction number (Rt) as Omicron became prevalent. 7. Omicron's Rt initially exceeded 1 even with high vaccination coverage. 8. BA.2 demonstrated a higher transmission rate than BA.1, contributing to the prolonged Omicron wave. 9. A higher proportion of BA.2-infected individuals reported diverse symptoms, suggesting potential improvements in symptom-based surveillance. 10. Heterogeneity in BA.2 growth advantages across regions, potentially due to early introductions in areas with high international travel. 11. High BA.2-specific Rt values in late February/early March 2022, even in older age groups with high vaccination rates, indicating limited vaccine effectiveness against Omicron.

The study's findings highlight the limitations of relying solely on routine testing data for assessing pandemic waves and the importance of utilizing more robust, less biased data like those from REACT-1. The observed dynamics, including the initial Omicron peak and the subsequent resurgence driven by BA.2, emphasize the challenges posed by immune evasion and the continuous evolution of SARS-CoV-2. The findings also underscore the complexity of variant competition and the role of transmission rates and immune landscapes in shaping epidemic trajectories. The variation in peak prevalence timing across regions and age groups reflects variations in prior immunity and other factors influencing transmission dynamics. The study's implications are significant for public health planning and policy, emphasizing the need for continuous surveillance, vaccination strategies, and potential vaccine updates to minimize the impact of future waves.

This study provides a detailed analysis of the Omicron wave in England, demonstrating the value of unbiased, prevalence-based data for understanding pandemic dynamics. The findings confirm the rapid dominance of Omicron and its subsequent resurgence due to the increased transmissibility of BA.2. The limitations of current vaccines against Omicron infection are highlighted, underscoring the need for ongoing surveillance, booster vaccinations, and potential vaccine updates. The authors suggest that recurring waves of infection of comparable magnitude may be expected given the virus's ability to evolve to evade immunity.

The study acknowledges limitations, including the discrete sampling periods in the REACT-1 study resulting in gaps in data and wider credible intervals during key periods. The reliance on N-gene Ct values for detection and the potential biases associated with differences in Ct values between lineages are also acknowledged. Further, the study's focus on prevalence rather than incidence and the assumptions made in region and age-group specific Rt estimates must be considered.