Protection of Omicron sub-lineage infection against reinfection with another Omicron sub-lineage

There is significant genetic distance between SARS-CoV-2 Omicron (B.1.1.529) variant BA.1 and BA.2 sub-lineages. This study investigates immune protection of infection with one sub-lineage against reinfection with the other sub-lineage in Qatar during a large BA.1 and BA.2 Omicron wave, from December 19, 2021 to March 21, 2022. Two national matched, retrospective cohort studies are conducted to estimate effectiveness of BA.1 infection against reinfection with BA.2 (N = 20,994; BA.1-against-BA.2 study), and effectiveness of BA.2 infection against reinfection with BA.1 (N = 110,315; BA.2-against-BA.1 study). Associations are estimated using Cox proportional-hazards regression models after multiple imputation to assign a sub-lineage status for cases with no sub-lineage status (using probabilities based on the test date). Effectiveness of BA.1 infection against reinfection with BA.2 is estimated at 94.2% (95% CI: 89.2–96.9%). Effectiveness of BA.2 infection against reinfection with BA.1 is estimated at 80.9% (95% CI: 73.1–86.4%). Infection with the BA.1 sub-lineage appears to induce strong, but not full immune protection against reinfection with the BA.2 sub-lineage, and vice versa, for at least several weeks after the initial infection.


Introduction
(WHO) guidelines, 13,14 in addition to sex, age, and nationality information retrieved from the national registry. Further description of these national databases can be found in previous publications. 5,[15][16][17][18] During this study, from December 19, 2021 to February 21, 2022, nearly all infection incidence Informed by viral genome sequencing and RT-qPCR genotyping, a SARS-CoV-2 infection with the BA.1 sub-lineage was proxied as an S-gene "target failure" (SGTF) case using the TaqPath COVID-19 Combo Kit (Thermo Fisher Scientific, USA). 19 Conversely, an infection with the BA.2 sub-lineage was proxied as a non-SGTF case using the same assay.
We assessed effectiveness of BA.1 infection against reinfection with BA.2 (denoted as (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted February 25, 2022.

Laboratory methods
Laboratory methods for the RT-qPCR testing, rapid antigen testing, and viral genome sequencing are found in Section S1.

Statistical analysis
Frequency distributions and measures of central tendency were used to describe full and matched cohorts. Group comparisons were performed using standardized mean differences (SMDs), with an SMD <0.1 indicating adequate matching. 29 Cumulative incidence of infection was defined as the proportion of individuals at risk whose primary endpoint was an incident infection during follow-up, and was estimated in each cohort using the Kaplan-Meier estimator method. 30 Incidence rate of infection in each cohort, which was defined as the number of identified All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Oversight
Hamad Medical Corporation and Weill Cornell Medicine-Qatar Institutional Review Boards approved this retrospective study with a waiver of informed consent. The study was reported following STROBE guidelines. The STROBE checklist is found in Table S1. Figure 2 shows the population selection process for the BA.1-against-BA.2 study. Table 1 shows baseline characteristics of full and matched cohorts. The study was based on the total population of Qatar and is broadly representative of the diverse (international), but young and predominantly male, total population of Qatar (Table S2).
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  Figure 2 shows the population selection process for the BA.2-against-BA.1 study. Table 1 shows baseline characteristics of full and matched cohorts. The study population was representative of the population of Qatar (Table S2).

BA.2-against-BA.1 study
The median time of follow-up was 12 days (IQR, 8-15 days) for both the BA.2-infected cohort and the uninfected-control cohort ( Figure 3). The proportion of individuals who had a PCR or All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted February 25, 2022.
However, it is remarkable that incidence of reinfection, regardless of sub-lineage, was much lower in the BA.1-infected and BA.2-infected cohorts than incidence of infection in the corresponding uninfected-control cohorts (Figure 2), consistent with strong protection against reinfection regardless of sub-lineage. Our findings indicate that infection with an Omicron sublineage appears to elicit strong protection against reinfection with the other sub-lineage at an effectiveness that exceeds 85%, similar to the protection observed for infection with original virus or earlier variants (Alpha, Beta, or Delta) against reinfection with original virus or earlier variants. 2,[5][6][7][8][9]20,24 These findings, in the context of broader evidence for natural immunity, 2,5-9,20-22 suggest that natural immunity of SARS-CoV-2 variants cluster into two groups: early non-Omicron variants, and Omicron BA.1 and BA.2 sub-lineages. Within each group, there appears to be strong protection against reinfection with an effectiveness that exceeds 85%. However, across groups, the protection may not exceed 60%, as was observed recently. 7 This conclusion is also supported by evidence on sensitivity of variants to SARS-CoV-2 antibodies. [2][3][4]31,32 This study has limitations. Since the Omicron wave was initially dominated with BA.1 (  All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted February 25, 2022. ;

12
Effectiveness against reinfection was estimated for only few weeks after the primary infection. A longer duration of follow-up may identify differences not yet seen given the recency of the Omicron wave. However, evidence has been consistent that natural immunity, unlike vaccine immunity, wanes slowly with minimal waning for at least several months after primary infection. 2,[5][6][7][8][9][20][21][22] BA.1 and BA.2 ascertainment was based on proxy criteria, presence or absence of SGTF using the TaqPath PCR assay, but this method of ascertainment is well established not only for Omicron sub-lineages, but also for other variants such as Alpha. 6 Some Omicron infections may have been misclassified Delta infections, but this is not likely, as Delta incidence was limited during the time of follow-up (Section S1). With the recency of the Omicron wave, we had to use a short interval of 35 days to define reinfection, perhaps introducing bias due to misclassification of prolonged infections as reinfections. However, such potential bias is less likely to affect . (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
As an observational study, the investigated cohorts were neither blinded nor randomized, so unmeasured or uncontrolled confounding cannot be excluded. While matching was done for sex, age, and nationality, this was not possible for other factors, such as comorbidities, occupation, or geography, as such data were not available. However, matching was done to control for factors that affect infection exposure in Qatar. 15,25-28 Matching by age may have reduced potential bias due to comorbidities. The number of individuals with severe chronic conditions is also small in Qatar's young population. 15,34 Matching by nationality may have partially controlled for differences in occupational risk or socio-economic status, given the association between nationality and occupation in Qatar. 15,25-28 Qatar is essentially a city state and infection incidence was broadly distributed across neighborhoods/areas; that is, geography is not likely to have been a confounding factor. Lastly, matching by the considered factors has been shown to provide adequate control of bias in studies that used control groups in Qatar. 16,[35][36][37][38] These control groups included unvaccinated cohorts versus vaccinated cohorts within two weeks of the first dose, 16,[35][36][37] when vaccine protection is negligible, 39,40 and mRNA-1273-versus BNT162b2-vaccinated cohorts, also in the first two weeks after the first dose. 38 A strength of this study is exclusion of those with a documented prior infection before the Omicron wave, to minimize potential confounding introduced by natural immunity due to earlier variants.
In conclusion, infection with an Omicron sub-lineage appears to induce strong, but not full protection against reinfection with the other sub-lineage, for at least several weeks after the initial infection.
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Competing interests
Dr. Butt has received institutional grant funding from Gilead Sciences unrelated to the work presented in this paper. Otherwise, we declare no competing interests. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted February 25, 2022. ;

Figure 1. Proportion of BA.1 (versus BA.2) Omicron infections in PCR-positive tests assessed using TaqPath COVID-19 Combo Kit during the study period.
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.    All rights reserved. No reuse allowed without permission.

Supplementary Appendix
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted February 25, 2022. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Classification of infections by variant type
Surveillance for SARS-CoV-2 variants in Qatar is mainly based on viral genome sequencing and multiplex RT-qPCR variant screening 1  able to assign a genotype in 129 samples. The agreement between RT-qPCR genotyping and sequencing was 100% for Delta (n=82), 100% for Omicron BA.1 (n=18), and 93% for Omicron All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. sequencing, and the remaining 4 failed sequencing). All the variant RT-qPCR genotyping was conducted at the Sidra Medicine Laboratory following standardized protocols.
The large Omicron-wave exponential-growth phase in Qatar started on December 19, 2021 and peaked in mid-January, 2022. 4,10,11 The study duration coincided with the intense Omicron wave where Delta incidence was limited. Accordingly, any PCR or rapid antigen positive test during the study duration, between December 19, 2021 and February 21, 2022, was assumed to be an infection with the BA.2 sub-lineage was proxied as a non-SGTF case using the TaqPath assay.
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.   (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  The effectiveness estimates were derived by comparing 1) incidence of BA.2 infection in the BA.1-infected cohort and uninfected-control cohort, and 2) incidence of BA.1 infection in the BA.2-infected cohort and uninfected-control cohort. Cohorts were exact-matched by sex to control for potential differences in the risk of exposure to SARS-CoV-2 infection by sex. Age Cohorts were exact-matched by 10-year age group to control for potential differences in the risk of exposure to SARS-CoV-2 infection by age. Nonetheless, with the young population of Qatar, our findings may not be generalizable to other countries where elderly citizens constitute a larger proportion of the total population. Race or ethnicity group Cohorts were exact-matched by nationality to control for potential differences in the risk of exposure to SARS-CoV-2 infection by nationality. Nationality is associated with race and ethnicity in the population of Qatar. Geography Individual-level data on geography were not available, but Qatar is essentially a city state and infection incidence was broadly distributed across the country's neighborhoods/areas. Cohorts were exact-matched by nationality to control for potential differences in the risk of exposure to SARS-CoV-2 infection by nationality. Qatar has unusually diverse demographics in that 89% of the population are international expatriate residents coming from over 150 countries from all world regions.

Other considerations
Individual-level data on co-morbid conditions were not available, but only a small proportion of the study population may have had serious co-morbid conditions. Only 9% of the population of Qatar are ≥50 years of age (older age as proxy for co-morbidities). The national list of persons prioritized to receive the vaccine during the first phase of vaccine roll-out included only 19,800 individuals of all age groups with serious co-morbid conditions. Individual-level data on occupation were not available but matching by nationality may have (partially) controlled the differences in occupational risk, given the association between nationality and occupation in Qatar. Overall representativeness of this study The study was based on the total population of Qatar and thus the study population is broadly representative of the diverse, by national background, but young and predominantly male, total population of Qatar. While there could be differences in the risk of exposure to SARS-CoV-2 infection by sex, age, and nationality, cohorts were exact-matched by these factors to control for their potential impact on our estimates for effectiveness of prior infection with a SARS-CoV-2 Omicron sub-lineage (BA.1 or BA.2) against reinfection with the other sub-lineage. Given that only 9% of the population of Qatar are ≥50 years of age and the limited proportion of the population with significant co-morbidities, our estimates of effectiveness may not be generalizable to other countries where elderly citizens constitute a larger proportion of the total population or where comorbid conditions are prevalent.
All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted February 25, 2022. All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.