Rapidly shifting immunologic landscape and severity of SARS-CoV-2 in the Omicron era in South Africa

Abstract

South Africa was among the first countries to detect the SARS-CoV-2 Omicron variant. Propelled by increased transmissibility and immune escape properties, Omicron displaced other globally circulating variants within 3 months of its emergence. Due to limited testing, Omicron's attenuated clinical severity, and an increased risk of reinfection, the size of the Omicron BA.1 and BA.2 subvariants (BA.1/2) wave remains poorly understood in South Africa and in many other countries. Using South African data from urban and rural cohorts closely monitored since the beginning of the pandemic, we analyzed sequential serum samples collected before, during, and after the Omicron BA.1/2 wave to infer infection rates and monitor changes in the immune histories of participants over time. Omicron BA.1/2 infection attack rates reached 65% (95% CI, 60% - 69%) in the rural cohort and 58% (95% CI, 61% - 74%) in the urban cohort, with repeat infections and vaccine breakthroughs accounting for >60% of all infections at both sites. Combined with previously collected data on pre-Omicron variant infections within the same cohorts, we identified 14 distinct categories of SARS-CoV-2 antigen exposure histories in the aftermath of the Omicron BA.1/2 wave, indicating a particularly fragmented immunologic landscape. Few individuals (<6%) remained naïve to SARS-CoV-2 and no exposure history category represented over 25% of the population at either cohort site. Further, cohort participants were more than twice as likely to get infected during the Omicron BA.1/2 wave, compared to the Delta wave. Prior infection with the ancestral strain (with D614G mutation), Beta, and Delta variants provided 13% (95% CI, -21% - 37%), 34% (95% CI, 17% - 48%), and 51% (95% CI, 39% - 60%) protection against Omicron BA.1/2 infection, respectively. Hybrid immunity (prior infection and vaccination) and repeated prior infections (without vaccination) reduced the risks of Omicron BA.1/2 infection by 60% (95% CI, 42% - 72%) and 85% (95% CI, 76% - 92%) respectively. Reinfections and vaccine breakthroughs had 41% (95% CI, 26% - 53%) lower risk of onward transmission than primary infections. Our study sheds light on a rapidly shifting landscape of population immunity, along with the changing characteristics of SARS-CoV-2, and how these factors interact to shape the success of emerging variants. Our findings are especially relevant to populations similar to South Africa with low SARS-CoV-2 vaccine coverage and a dominant contribution of immunity from prior infection. Looking forward, the study provides context for anticipating the long-term circulation of SARS-CoV-2 in populations no longer naïve to the virus.

Fig. 1.. PHIRST-C study June 2020 – April 2022, SARS-CoV-2 serology and epidemiologic curve in the two study sites.

(A) Serum samples and epidemiologic curve in the rural site. Dots represent the Roche Elecsys Anti-SARS-CoV-2 nucleocapsid assay cutoff index (COI) at different timepoints of the serum specimen collection; Each dot represents one serum specimen collection, with dot color denoting blood draw collection time, from blue (early) to red (late). The shaded curve at the bottom represents the daily incidence of SARS-CoV-2 cases in routine surveillance data collected by the Ehlanzeni District, Mpumalanga Province. Colors of the shaded curve represent different variant types. Here, blood draw (BD) 10 was collected at the end of the first Omicron wave. Since in South Africa, Omicron BA.4 and BA.5 only started to rise at April, 2022 (8), we assume the Omicron wave prior to BD 10 were BA.1 and BA.2 subvariants. The hatched area represents the period of intense follow-up of the PHIRST-C cohort, when nasal swabs were collected and tested on rRT-PCR at twice-a-week frequency. (B) Same as (A) but for the urban site. Abbreviation: BD stands for blood draw.

Fig. 2.. SARS-CoV-2 infection attack rates and shifts in immunologic landscape.

(A) Infection attack rates in the rural site by variant type (left) and the cumulative number of infection episodes per capita after each epidemic wave (right). The end of the 1^st wave is marked by blood draw 2, 2^nd wave is marked by blood draw 5, 3^rd is marked by blood draw 8, 4^th is marked by blood draw 10. (B) Same as (A) but for the urban site. (C) Sankey diagram demonstrating the distribution of different type of immunologic exposures (including vaccination and infection) in the population of the rural site after each epidemic wave and the transition of immunologic exposures in-between waves. Gray color represents SARS-CoV-2 immunologic naïve individuals; blue shades represent non-Omicron exposures; red shades represent Omicron exposures; darker colors represent repeat exposures while transparent shading represents primary exposures. (D) same as (C) but for the urban site. (E) The distribution of the population by the number of SARS-CoV-2 exposures experienced after the 3^rd epidemic wave (left) or after the 4^th epidemic wave (right). Each infection/vaccine dose is counted as 1 exposure. Gray bars represent naïve individuals; blue bars represent individuals who have been infected by pre-Omicron variants or received SARS-CoV-2 vaccinations; red bars represent individuals who have been infected by Omicron. *In additional to Delta, here also includes other less frequent lineages including other lineages including Alpha and C.1.2 variants.

Fig. 3.. Risk factors associated with SARS-CoV-2 Omicron BA.1/2 and Delta infection.

Odds ratios (adjusted after controlling for other risk factors, see Material and Methods Section 4 for details) were estimated by a chain binomial model where the infection status was inferred from serology. Abbreviation: PLWH: persons living with HIV. Category “Unknown” for “HIV infection status” and category “Rest” in “Prior exposure” (Table 1) were included in the model but omitted here due to small sample size in the strata. *0–4 age group have odds ratio point estimate less than 0.01, thus not shown in the figure. ^#Non-primary infections represent repeat/breakthrough infections. ^†Household size denotes the number of household members within a household and is analyzed as a continuous variable.

Figure 4:. The infection fatality ratios and the factors associated with SARS-CoV-2 disease severity for different epidemic waves in the urban site’s district.

(A) the estimated infection fatality ratio for each epidemic wave; (B) the mortality burden of each epidemic wave measured by the cumulative rate of in-hospital deaths per 100 individuals; (C) the infection attack rate of each epidemic wave in the North West based on the PHIRST-C urban cohort, assuming that the urban cohort population is representative of the population of the North West Province; (D) The wave-specific distribution of infection types based on prior exposure histories, including primary infection, vaccine breakthroughs (1 or 2 doses of vaccines), reinfections (infection after one prior infection), and multiple prior exposures (infection with two or more prior infections or a mixture of prior infection and vaccination). (E) The wave-specific age distribution of infections. (F) The wave-specific distribution of variant type among infections. Panel B-F share the same axis on the right. *For the 4^th wave, we could not confirm variant type by variant-specific rRT-PCR or sequencing, however, judging from the timing of emergence and dominance of Omicron in South Africa in late November 2021, we assumed here that all infections during the 4^th wave were due to Omicron BA.1/2 variants.