Abstract
Expansion of the SARS-CoV-2 BA.4/BA.5 Omicron lineages in populations with prevalent immunity from prior infection and vaccination has raised concerns about the association of these lineages with immune escape. Here we show that COVID-19 vaccination and documented prior infection are associated with reduced protection against infection with BA.4/BA.5. Compared to time-matched BA.2 cases, BA.4/BA.5 cases had 9% (95% confidence interval: 2-17%) and 27% (15-41%) higher adjusted odds of having received 3 and 4 COVID-19 vaccine doses, respectively, and 55% (39-71%) higher adjusted odds of documented infection ≥90 days previously. However, BA.4/BA.5 infection was not associated with differential risk of emergency department presentation, hospital admission, or intensive care unit admission following an initial outpatient diagnosis. This finding held after correcting for potential exposure misclassification resulting from unascertained prior infections. Despite increased risk of BA.4/BA.5 breakthrough infection observed among previously vaccinated or infected individuals, the reduced severity associated with prior (BA.1 and BA.2) Omicron lineages has persisted with BA.4/BA.5.
The SARS-CoV-2 Omicron (B.1.1.529) variant emerged in late 2021 and rapidly achieved global dissemination, accounting for a majority of incident SARS-CoV-2 infections within the United States by late December, 2021.1,2 As of February, 2022, 58% of US adults and 75% of US children aged ≤17 years were estimated to have acquired SARS-CoV-2 infection, with nearly half of these infections occurring during the initial expansion of the BA.1 subvariant lineage.3 COVID-19 vaccination and naturally-acquired immunity from infection with pre-Omicron variants have generally been found to confer robust protection against clinically severe disease involving the Omicron variant, but both confer weaker protection against Omicron variant infection compared with prior variants.4–7 Although widespread transmission of the Omicron variant was thus associated with surges in COVID-19 hospitalizations and deaths, the proportion of Omicron cases resulting in severe illness has been lower than that experienced with prior variants and during outbreaks in immunologically naïve populations.4,8
Following the initial peak in BA.1 infections within the US from December, 2021 to February, 2022, multiple Omicron lineages have driven subsequent surges in cases. Although not associated with enhanced severity or risk of breakthrough infection after vaccination or natural infection,4,9 the BA.2 lineage (and associated BA.2.12.1 sublineage) surpassed BA.1 in incident cases within the US beginning in March, 2022. Subsequently, the BA.4 and BA.5 lineages have become dominant globally.10 These novel lineages share an identical spike (S) protein, which represents the primary target of host immune responses, with changes at multiple amino acids (relative to BA.2) expected to alter binding to both human angiotensin-converting enzyme-2 and antibodies derived from prior vaccination or infection.11 However, clinical implications of the emergence of BA.4/BA.5 remain uncertain, as the burden of hospitalized and fatal COVID-19 cases observed during BA.4/BA.5 waves has varied widely across settings.12
Establishing the relative severity of BA.4/BA.5 lineage infections and their capacity to evade vaccine- or infection-derived immunity is of key importance to informing public health responses. We therefore compared clinical outcomes and characteristics of contemporaneous cases with BA.2 and BA.4/BA.5 lineage Omicron variant infections within the Kaiser Permanente Southern California (KPSC) healthcare system from 29 April to 2 July, 2022, replicating the approach taken in prior analyses which established the relative severity of infections with BA.1, BA.2, and Delta variant lineages.4 As a comprehensive, integrated care organization, KPSC delivers healthcare across telehealth, outpatient, emergency department, and inpatient settings for over 4.7 million members. Electronic health records (EHRs) across all clinical settings, together with laboratory, pharmacy, and immunization data, provide a comprehensive view into care delivered by KPSC. These observations are augmented by insurance claims for out-of-network diagnoses, prescriptions, and procedures.
In total, 65,694 SARS-CoV-2 cases (out of 81,880 diagnosed at KPSC during the study period) were identified in outpatient settings with samples processed using the ThermoFisher TaqPath COVID-19 Combo Kit, and thus eligible for inclusion in the study (see Methods). Within this sample, 16,753 (25.5%) and 48,941 (74.5%) cases were infected with the BA.4/BA.5 and BA.2 lineages, respectively. Age distributions were similar among cases infected with either lineage, with 13.8% of all cases aged 0-17 years, 33.9% aged 18-39 years, 40.8% aged 40-64 years, and 11.1% aged ≥65 years (Table S1). Other attributes including race/ethnicity, sex, body mass index, Charlson comorbidity index, neighborhood deprivation index, prior-year healthcare utilization, and receipt of vaccines targeting respiratory pathogens other than SARS-CoV-2 likewise did not differ markedly between cases infected with the BA.4/BA.5 and BA.2 lineages.
Among BA.4/BA.5 cases, 16.9% had not received any COVID-19 vaccine doses while 2.5%, 24.2%, 49.0%, and 7.5% had received 1, 2, 3, and 4 doses, respectively, before their diagnosis (Table 1). Among BA.2 cases, 17.5%, 2.5%,
%, 48.9%, and 5.9% had received 0, 1, 2, 3, and 4 doses, respectively. In logistic regression analyses adjusting for all measured covariates among cases (Table S1), including calendar time of diagnosis, adjusted odds of having received 3 and 4 COVID-19 vaccine doses were 1.09 (95% confidence interval: 1.02-1.17) and 1.27 (1.15-1.41) fold higher among BA.4/BA.5 cases than BA.2 cases. As compared to 4.5% of BA.4/BA.5 cases, 2.9% of BA.2 cases had documentation of a prior SARS-CoV-2 infection ≥90 days before their positive test. Adjusted odds of documented prior infection were 1.55 (1.39-1.71) fold higher among BA.4/BA.5 cases than BA.2 cases.
Among 4,349 outpatient-diagnosed BA.4/BA.5 cases who were followed for ≥30 days from their initial outpatient test, crude 30-day incidence of ED presentation, symptomatic ED presentation, inpatient admission, symptomatic inpatient admission, and ICU admission was 26.0, 25.1, 3.0, 2.3, and 0.2 per 1,000 cases, respectively; no deaths or instances of mechanical ventilation occurred (Figure 1). Among 32,592 outpatient-diagnosed BA.2 cases followed ≥30 days, crude 30-day incidence of ED presentation, symptomatic ED presentation, inpatient admission, symptomatic inpatient admission, ICU admission, mechanical ventilation, and death was 26.0, 24.4, 3.0, 2.6, 0.3, 0.1, and 0.2 per 1,000 cases. Adjusted hazards ratios comparing BA.4/BA.5 cases to BA.2 cases, estimated via Cox proportional models defining strata on cases’ week of diagnosis, were 1.07 (0.92-1.25) for any ED presentation, 1.07 (0.92-1.26) for symptomatic ED presentation, 1.06 (0.66-1.70) for any inpatient admission, 1.04 (0.62-1.72) for symptomatic hospital admission, and 0.93 (0.17-5.28) for ICU admission. No BA.4/BA.5 cases died or received mechanical ventilation, precluding estimation of adjusted hazard ratios for these outcomes. This observation did not depart from expectations within the available sample size and follow-up period under the null hypothesis of equal risk of mechanical ventilation or death among cases infected with the BA.4/BA.5 and BA.2 lineages (binomial p=0.54 for 0 cases with mechanical ventilation and p=0.19 for 0 death, each parameterized using unadjusted outcome probabilities observed among BA.2 cases). Coefficient estimates from Cox proportional hazards models did not reveal weaker associations of prior vaccination or infection with risk of severe clinical outcomes among BA.4/BA.5 cases relative to BA.2 cases (Table S2).
Plots illustrate cumulative 30-day risk of severe clinical outcomes among cases first ascertained in outpatient settings, stratified by SGTF status for infecting subvariant (BA.4/BA.5 [SGTF]: red; BA.2 [No SGTF]: blue), for endpoints of any emergency department (ED) presentation (a); symptomatic ED presentation (b); hospital admission (c); and symptomatic hospital admission (d). Shaded areas denote 95% confidence intervals around median estimates (center lines). Data presented in the table (e) include the hazard ratio (HR) and adjusted hazard ratio (aHR) of each outcome, comparing cases with BA.4/BA.5 infection to those with BA.2 infection, estimated via Cox proportional hazards models including strata for cases’ week of diagnosis. Events and crude rates per 100,000 person-days at risk are calculated among cases with ≥30 days of follow up from the date of testing.
Our findings of non-differential severity held in analyses subset to cases with documented history of prior infection, although the sample size within this stratum was sufficient only for assessment of ED endpoints (Table S3). Further risk-of-bias analyses allowing for differential misclassification of prior infection status among BA.4/BA.5 or BA.2 cases identified that findings would not differ even under extreme scenarios with >20-fold higher-than-reported prevalence of prior infection among cases with BA.4/BA.5 infection (Figure S1).
We illustrate estimates of the corrected adjusted hazard ratio (aHR) under scenarios of differing prevalence of potentially under-reported prior infection among cases with BA.2 and BA.4/BA.5 infection. Analyses define α (for α ∈ 1, 2, …, 5) as the ratio of true prior infections to observed prior infections among cases with BA.2 infection (panels), and further allow for (∈ 1, 2, 3, 4 fold higher prevalence of unobserved prior infection among cases with BA.4/BA.5 infection as compared to cases with BA.2 infection (defined along the x-axis). Vertical lines denote 95% confidence limits around point estimates (circles), with x-axis markers indicating the modeled ratio (equal to α × ρ) of true prior infections to observed prior infections among cases with BA.4/BA.5 infection. Horizontal lines indicate primary estimates (uncorrected for potentially unobserved prior infections, as plotted in Figure 1) accompanied by grey shaded areas delineating 95% confidence limits. Panels are presented for endpoints of symptomatic emergency department presentation (A-E), symptomatic hospital admission (F-J), and intensive care unit admission (K-O).
Our analysis has revealed several important epidemiologic characteristics of SARS-CoV-2 BA.4/BA.5 Omicron lineage infections. First, outpatient-diagnosed BA.4/BA.5 cases had 55% higher adjusted odds of a prior documented infection than contemporaneous BA.2 cases tested in the same settings, as well as modestly higher adjusted odds of having received ≥3 COVID-19 vaccine doses. These findings corroborate earlier suggestions of immune escape in BA.4/BA.5 infections, which to date have been based largely on data from genomic10 and neutralization11,13,14 analyses rather than direct clinical evidence. Importantly, prior COVID-19 vaccination was not associated with differential protection against severe clinical endpoints among BA.4/BA.5 and BA.2 cases in our study population. As our study is limited to infected cases, is important to note our findings do not measure the effectiveness of prior infection or vaccination against infection with the BA.4/BA.5 or BA.2 lineages. At least one previous study has demonstrated that prior infection, especially with BA.1 or BA.2 Omicron lineages, remains modestly protective against BA.4/BA.5 infection.15
Consistent with a smaller recent study in South Africa,9 we did not identify evidence of differential severity of infections with the BA.4/BA.5 and BA.2 lineages in either unadjusted analyses comparing all incident cases or adjusted analyses controlling for measured characteristics of cases with each lineage. Previous analyses within our study population4 and South Africa16 have likewise established equivalent severity of BA.1 and BA.2 infections. Thus, our findings likely imply that the reduced severity associated with BA.1 Omicron variant infections relative to Delta variant infections has persisted with BA.4/BA.5. While this result is encouraging, it is important to note that disease burden is influenced by further variant-specific properties including the intrinsic capacity to transmit and to infect individuals with immunity from prior vaccination or infection.17 These fitness advantages are relevant to consider in the context of BA.4/BA.5, which has outcompeted BA.2 in the context of substantial population immunity.10
Our analysis has limitations. Prior infections are likely undercounted among both BA.4/BA.5 and BA.2 cases. This misclassification may cause our analyses to underestimate the true difference in prevalence of prior infection among cases acquiring each lineage. However, our findings of equivalent risk of severe clinical outcomes with each lineage are unlikely to be driven by this factor. Sensitivity analyses identified that ascertainment of <1 in 20 prior infections among BA.4/BA.5 cases would be required to explain our null findings if BA.4/BA.5 truly caused more severe illness. The proportion of infections ascertained in our study population is not precisely known; however, this represents a considerable departure from prior estimates of the reporting fraction in California.18 Whereas bias would require not only the prevalence of prior infection, but the proportion of prior infections ascertained, to differ among cases with BA.4/BA.5 and BA.2 infection, we are unaware of a priori reasons to suspect the latter would occur. It is important to note that our analyses do not distinguish causes of ED presentations and hospital admission, although analyses subset to symptomatic endpoints19 and our restriction to cases tested in outpatient settings4 may help to exclude incidentally-identified infections. Last, our analyses do not distinguish cases infected with BA.4 and BA.5, or BA.2.12.1 and other BA.2 lineages, which may be associated with distinct epidemiologic and clinical characteristics.
While it is reassuring that we do not identify an association of BA.4/BA.5 infection with increased disease severity, and that we do not identify evidence of differential vaccine protection against severe disease with either lineage, our findings support the role of immune escape in BA.4/BA.5 expansion. Updating COVID-19 vaccine compositions to match circulating SARS-CoV-2 lineages may play an important role in mitigating both disease burden and societal disruption associated with ongoing widespread SARS-CoV-2 transmission.
Data Availability
Individual level clinical data are not publicly available for sharing.
Author contributions statement
JAL, VH, and SYT contributed to the study concept and design. JAL, VH, and SYT led acquisition and statistical analysis of data. JAL and SYT led interpretation of data. JAL drafted the manuscript, and VH and SYT critically revised the manuscript for important intellectual content. SYT obtained funding and provided supervision.
Competing interests
JAL has received research grants and consulting honoraria unrelated to this study from Pfizer. SYT has received research grants unrelated to this study from Pfizer. VH discloses no competing interests.
Methods
Setting, procedures, and study population
Care delivery within KPSC has been described previously.4 Briefly, approximately 19% of the population of southern California receives care from KPSC through employer-provided, prepaid, or federally sponsored insurance plans (∼4.7 million members as of 2022). In-network care delivery data encompassing diagnoses (and accompanying clinical notes), immunizations, laboratory tests administered and test results, and prescriptions are captured in real time via patient EHRs, while out-of-network care is captured through insurance claim reimbursements. Delivery of COVID-19 vaccine doses by other providers was identified via linkage to California Immunization Registry data. Online portals provide an automated platform for individuals to upload or notify providers of positive at-home test results or test results received from other providers.
Molecular diagnostic testing for COVID-19 was made available to KPSC patients across both outpatient and inpatient settings throughout the study period. Most tests conducted in outpatient settings were processed in regional laboratories using the ThermoFisher TaqPath COVID-19 Combo Kit assay, which includes probes targeting the S, nucleocapsid (N), and Orf1a/b genes. As BA.4/BA.5 lineages harbor the Δ69-70 amino acid deletion in the S protein, S-gene target failure (SGTF) has been proposed elsewhere as a proxy for distinguishing BA.4/BA.5 from BA.2 lineages.9,10 Within a sample of cases KPSC randomly selected for sequencing throughout 2022, 99.6% (243/244) of BA.4 cases and 99.0% (406/410) of BA.5 cases exhibited SGTF. In contrast, SGTF did not occur in 97.8% (2,000/2,046) of BA.2 cases, validating the use of SGTF to distinguish contemporaneous infections with the BA.4/BA.5 and BA.2 lineages.
Consistent with prior analyses,4 we restricted our analytic sample to cases with a positive SARS-CoV-2 test result from testing undertaken in outpatient settings using the ThermoFisher TaqPath COVID-19 Combo Kit. In addition to enabling longitudinal follow-up for severe endpoint ascertainment, restricting analyses to cases tested as outpatients was expected to provide two design advantages helping to mitigate bias. First, excluding individuals first ascertained in hospital settings helped to reduce bias driven by differential healthcare-seeking behavior among cases tested as outpatients versus those who deferred testing to more severe stages of illness. Second, this approach enabled us to minimize the inclusion of cases hospitalized cases who were identified incidentally via SARS-CoV-2 infection screening at admission. Although ED presentations, hospital admissions, and other study endpoints are generally rare events, individuals’ risk of each endpoint is greatly increased by SARS-CoV-2 infection. Thus, a majority of observed severe endpoints observed within the outpatient-diagnosed sample were expected to be attributable to COVID-19.
To ensure our analyses captured new-onset infections, we excluded cases with a positive SARS-CoV-2 test result within the prior 90 days. To ensure adequate follow-up time among cases, we further excluded cases who disenrolled prior to the study end date or 60 days after their test date, whichever was earlier. The study protocol was approved by the KPSC Institutional Review Board.
Outcomes
Outcomes of interest to our analyses included: (1) any ED presentation; (2) symptomatic ED presentation, defined as those occurring ≤14 days after the first day that patients reported experiencing any new-onset solicited symptoms (as described below);19 (3) any inpatient admission; (4) symptomatic inpatient admission, applying the same definition as above; (5) ICU admission; (5) mechanical ventilation; and (6) death. We limited follow-up time for ED presentation, symptomatic ED presentation, hospital admission, and symptomatic hospital admission to 30 days following the initial positive outpatient test; we included follow-up time through 60 days from the initial positive outpatient test for endpoints of ICU admission, mechanical ventilation, and death. We censored observations at study end date for cases who had not experienced each outcome, provided a minimum of 2 weeks of follow-up were available. As cases diagnosed in outpatient settings were enrolled in a home-based monitoring program with standardized criteria for ED referral and inpatient admission,20 we expected severity of illness associated with each endpoint to be internally comparable within the study cohort.
Data on symptoms were collected for all individuals receiving SARS-CoV-2 testing at KPSC, and analyzed via natural language processing of both structured and unstructured patient HER data, as described elsewhere.19 Solicited symptoms potentially associated with COVID-19 included fever, cough, headache, fatigue, dyspnea, chills, sore throat, myalgia, anosmia, diarrhea, nausea/vomiting, and abdominal pain.
Case characteristics
We recorded the following characteristics for each case: age (defined in 10-year age bands), sex, race/ethnicity (white, black, Hispanic of any race, Asian, Pacific Islander, and other/mixed/unknown race), neighborhood deprivation index, measured at the Census block level; smoking status (current, former, or never smoker); body mass index (BMI; underweight, normal weight, overweight, obese, and morbidly obese); Charlson comorbidity index (0, 1-2, 3-5, and ≥6); prior-year emergency department visits and inpatient admissions (each defined as 0, 1, 2, or ≥3 events); prior-year outpatient visits (0-4, 5-9, 10-14, 15-19, 20-29, or ≥30 events); documented prior SARS-CoV-2 infection; and history of COVID-19 vaccination (receipt of 0, 1, 2, 3, or 4 doses, and time from receipt of each dose to each case’s testing date), and receipt of Paxlovid ≤14 days after the initial outpatient diagnosis date.
Multiple imputation of missing data
Variables with missing data included cases’ age (0.4% of observations), neighborhood deprivation index (0.7% of observations), BMI (21.4% of observations), and smoking status (17.1% of observations). We populated 5 complete pseudo-datasets sampling from the distribution of missing values (according to the joint distribution of all variables) via multiple imputation and repeated all statistical analyses across each pseudo-datsaet, pooling resulting estimates according to Rubin’s rules.21
Logistic regression analysis
We compared the distributions of prior vaccination status and prior infection status among BA.4/BA.5 cases versus BA.2 cases via logistic regression. Models controlled for all variables listed above, with the exception of Paxlovid receipt (which occurred after diagnosis), to define aORs in relation to infecting lineage. Models included distinct intercepts for each calendar week to control for potential changes in testing and healthcare-seeking practices over the period of BA.4/BA.5 emergence.
Survival analysis
We fit Cox proportional regression models including data from all outpatient-diagnosed cases with ≥14 days of follow-up from their initial positive test, censoring at either the study end date, 30 days (for ED and hospital admission endpoints), or 60 days (for ICU admission, mechanical ventilation, and mortality endpoints). Models defined covariates for each case characteristic listed above, and included time-varying covariates for Paxlovid exposure (before versus during/after treatment) according to the Andersen-Gill framework.22 Models defined strata according to cases’ calendar week of testing to control for potential changes in testing and healthcare-seeking practices over the period of BA.4/BA.5 emergence.
Sensitivity analyses
Because protection from prior infection could contribute to lower risk of clinical progression among cases with BA.4/BA.5 infection,23 we undertook several sensitivity analyses aiming to determine whether our results were robust to bias driven by potentially differential prevalence of unrecorded prior infections among cases infected with each lineage. First, we repeated survival analyses of the ED presentation and symptomatic ED presentation endpoints within the subset of cases known to have experienced a prior infection, as differential prevalence of prior infection could not lead to differences in disease progression within this stratum. However, sample sizes were inadequate to allow similar analyses for hospital admission, ICU admission, mechanical ventilation, and death endpoints within this subset. We therefore conducted risk-of-bias analyses allowing for non-differential or differential undercounting of prior infections among cases with BA.4/BA.5 and BA.2 infection, similar to prior work in the study population4 and described in detail below.
Within each imputed pseudo-dataset, we fit logistic regression models to define cases’ propensity for prior SARS-CoV-2 infection as a function of all measured characteristics as well as the occurrence of endpoints of ED presentation, hospital admission, ICU admission, mechanical ventilation, and death. Among cases with BA.2 infection, we multiplied these individual propensities by a factor α ∈ (1, 2, 3, 4, 5) to assign individuals probabilities of prior infection under scenarios where the ratio of recorded prior infections to true prior infections was 1:1, 1:2, …, 1:5. To allow for potentially differential ascertainment of prior infections among cases with BA.4/BA.5 infection, we multiplied propensities within this subset by a factor of α × ρ, with ρ ∈ (1, 2, 3, 4), such that the likelihood a prior infection was unobserved was up to 4 times higher among BA.4/BA.5 cases than among BA.2 cases. Analyses thus allowed up to 20 times higher-than-observed prevalence of prior infection among BA.4/BA.5 cases. We repeated survival analyses for endpoints of symptomatic ED presentation, symptomatic hospital admission, and symptomatic ICU admission in pseudo-datasets sampling individual prior infection histories, among individuals without a documented prior infection, according to these modeled individual propensities. We plot resulting estimates of the “corrected” aHR of each outcome, comparing BA.4/BA.5 cases to BA.2 cases, in Figure S1.
Acknowledgments
This work was funded by the US Centers for Disease Control and Prevention (CDC). The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC. JAL was supported by grant R01-AI14812701A1 from the National Institute for Allergy and Infectious Diseases (US National Institutes of Health) which had no role in design or conduct of the study, or decision to submit for publication.
References
