BNT162b2 vaccine effectiveness against SARS-CoV-2 omicron BA.4 and BA.5.

SARS-CoV-2 omicron (B.1.1.529) subvariants BA.4 and BA.5 were first detected in South Africa in December, 2021. Their spike (S) proteins are identical (hereafter referred to collectively as BA.4/5) and include L452R and F486V mutations in the receptor binding domain, which might lead to increased immune evasion or the ability to infect host cells, or both.1Tegally H Moir M Everatt J et al.Emergence of SARS-CoV-2 omicron lineages BA.4 and BA.5 in South Africa.Nat Med. 2022; 28: 1785-1790Crossref PubMed Scopus (72) Google Scholar Evidence also suggests that COVID-19 vaccine responses are less effective at neutralising BA.4/5 than BA.1 or BA.2 subvariants of omicron.2Qu P Faraone J Evans JP et al.Neutralization of the SARS-CoV-2 omicron BA.4/5 and BA.2.12.1 subvariants.N Engl J Med. 2022; 386: 2526-2528Crossref PubMed Scopus (40) Google Scholar Subsequently, BA.4/5 have become the predominant subvariants in the USA and globally.3Our World in DataSARS-CoV-2 sequences by Variant, Mar 1, 2021. Oxford Martin School University of Oxford, 2022https://ourworldindata.org/grapher/covid-variants-bar?time=earliest&country=CAN~BWA~ESP~ZAF~AUS~GBR~USA~DEU~ITA~BEL~FRADate accessed: September 10, 2022Google Scholar To our knowledge, no studies evaluating the effectiveness of COVID-19 vaccines against BA.4/5 have been published to date. Using the same test-negative design approach as in our previous analyses,4Tartof SY Slezak JM Puzniak L et al.BNT162b2 effectiveness and durability against BA.1 and BA.2 hospital and emergency department admissions in a large US health system: a test-negative design.SSRN. 2022; (published online June 30.) (preprint).https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4150500Google Scholar we determined the effectiveness of BNT162b2 (Pfizer–BioNTech) against BA.4/5 among members of the health insurance provider Kaiser Permanente, based in Southern California, CA, USA aged 18 years or older, who were diagnosed with an acute respiratory infection and tested for SARS-CoV-2 by PCR at one of four health-care settings (in ascending order of acuity of care: outpatient visits [including virtual appointments], urgent care centres, emergency departments, and the hospital) between May 9 and Aug 26, 2022 (appendix p 2). Variant sublineage was defined using a combination of whole genome sequencing, S-gene target failure as measured by TaqPath COVID-19 Combo Kit (ThermoFisher, Waltham, MA, USA), and calendar time (additional methods are in the appendix [p 2]). We assessed effectiveness of BNT162b2 against omicron subvariants BA.4 and BA.5, by highest level of care and number and timing of receipt of BNT162b2 doses. Kaiser Permanente's institutional review board granted a waiver for informed consent. 24 356 health-care encounters comprised the primary analyses of two-dose and three-dose effectiveness. Of which, 5182 (38%) of 13 718 outpatient, 1556 (20%) of 7977 urgent care, 575 (31%) of 1867 emergency department, and 123 (16%) of 794 hospital encounters had a positive SARS-CoV-2 test. Overall, 5793 (24%) of 24 356 patients who had a health-care encounter during this time were unvaccinated and 5997 (25%) had received two doses and 12 566 (52%) had received three doses of BNT162b2 vaccine (appendix p 3). Median age was 44 years (IQR 32–58), 15 755 (65%) patients were women, 8599 (35%) were men, fewer than six (<1%) had missing sex data, 12 536 (51%) were Hispanic, 5726 (24%) were White, 2519 (10%) were Asian, 2068 (8%) were Black, and 1507 (6%) were other races or ethnicities. Participant characteristics by vaccination status and outcome are shown in the appendix (pp 4–6). In adjusted analyses, point estimates for vaccine effectiveness against BA.4/5 after two doses were 50% or lower (albeit with wide 95% CIs for most estimates), regardless of outcome and time since last dose (table). At less than 6 months after a third dose of vaccine, vaccine effectiveness was 73% (95% CI 25–91) against BA.4/5-related hospitalisation; however, point estimates were less than 50% against milder outcomes (table). Vaccine effectiveness point estimates were greater than 50% against BA.4/5-related outpatient, urgent care, and emergency department encounters only in the first 3 months after a third dose of vaccine (table). At 6 months or longer after a third dose of vaccine, less protection was seen against BA.4/5 than at earlier timepoints, even for hospitalisation, although 95% CIs were wide (table; appendix p 7).TableAdjusted effectiveness*Adjusted for week of COVID-19 health-care encounter, age, sex, race or ethnicity, previous SARS-CoV-2 infection, BMI, Charlson score, and history of previous influenza and pneumococcal vaccination, and nirmatrelvir plus ritonavir receipt. of BNT162b2 vaccine against omicron (B.1.1.529) subvariants BA.4 and BA.5, by highest level of care and number and timing of receipt of BNT162b2 dosesHospitalEmergency departmentUrgent careOutpatientTwo doses of BNT162b2<6 months since second doseNC30 (−86 to 74)50 (10 to 72)30 (4 to 49)≥6 months since second dose−4 (−118 to 50)44 (20 to 61)7 (−11 to 22)19 (9 to 29)Overall−4 (−116 to 50)44 (19 to 61)11 (−7 to 25)21 (11 to 30)Three doses of BNT162b2<3 months since third doseNC71 (18 to 90)59 (35 to 74)55 (41 to 65)3–5 months since third dose72 (13 to 91)36 (−3 to 60)28 (10 to 42)23 (11 to 33)<6 months since third dose73 (25 to 91)43 (10 to 63)34 (18 to 46)29 (19 to 37)≥6 months since third dose38 (−31 to 71)37 (8 to 57)11 (−7 to 26)6 (−7 to 17)Overall50 (−1 to 76)39 (14 to 57)20 (5 to 33)17 (7 to 26)Four doses of BNT162b2†Analysis done among individuals aged ≥50 years (for whom a fourth dose was recommended at the time of the study).<3 months since fourth dose66 (20 to 85)65 (35 to 82)35 (10 to 54)28 (10 to 43)≥3 months since fourth dose33 (−112 to 79)78 (50 to 91)20 (−23 to 48)11 (−18 to 34)Overall60 (11 to 82)69 (44 to 83)32 (7 to 50)25 (7 to 39)Data are vaccine effectiveness, with 95% CIs in parentheses. NC=not calculated (ie, fewer than five total cases).* Adjusted for week of COVID-19 health-care encounter, age, sex, race or ethnicity, previous SARS-CoV-2 infection, BMI, Charlson score, and history of previous influenza and pneumococcal vaccination, and nirmatrelvir plus ritonavir receipt.† Analysis done among individuals aged ≥50 years (for whom a fourth dose was recommended at the time of the study). Open table in a new tab Data are vaccine effectiveness, with 95% CIs in parentheses. NC=not calculated (ie, fewer than five total cases). Among eligible participants aged 50 years and older, 3029 (24%) of 12 630 had received a fourth dose of vaccine. A fourth dose improved protection beyond that seen 6 months or longer after a third dose, back to levels that were comparable to those at less than 6 months after a third dose among all adults (table). Excluding immunocompromised individuals yielded similar results (appendix p 8). Our study has several limitations, including the potential for unmeasured confounding between vaccinated and unvaccinated individuals. Additionally, although we controlled for week of SARS-CoV-2 infection, it is possible that natural immunity, which we could not sufficiently measure, affected our estimates. Specifically, if many unvaccinated individuals gained undocumented natural immunity during omicron BA.1 or BA.2 waves, this could bias our vaccine effectiveness estimates downward. Our results suggest that two doses of BNT162b2 offered little protection against all BA.4/5 outcomes measured, including hospital admission. A booster (third or fourth dose) did provide protection against BA.4/5, but this protection probably wanes after 3 months against milder outcomes like outpatient, urgent care, or emergency department encounters and after roughly 6 months against BA.4/5-related hospitalisation. Approximately half of individuals who are eligible for a booster vaccination in the USA have not yet received a booster dose,5Centers for Disease Control and PreventionCOVID data tracker: COVID-19 vaccinations in the United States. Data as of: August 24, 2022 6:00am ET.https://covid.cdc.gov/covid-data-tracker/#vaccinations_vacc-people-additional-dose-totalpopDate: Aug 25, 2022Date accessed: August 31, 2022Google Scholar and of those who have, many did so at least 6 months ago. Moreover, only a third of US individuals aged 50 years and older who previously received a booster have received a second booster.5Centers for Disease Control and PreventionCOVID data tracker: COVID-19 vaccinations in the United States. Data as of: August 24, 2022 6:00am ET.https://covid.cdc.gov/covid-data-tracker/#vaccinations_vacc-people-additional-dose-totalpopDate: Aug 25, 2022Date accessed: August 31, 2022Google Scholar Thus, much of the US population—and other populations globally—probably have low levels of vaccine-derived immunity, underscoring the importance of booster programmes. The degree to which protection will be extended by BA.4/5-adapted vaccines in the real-world setting, however, is still unknown and requires future assessments in the months ahead. LJ, LP, JMZ, and JMM are employees of and hold stock or stock options in Pfizer. SYT, TBF, JMS, VH, HT, OAO, SS, and BKA received research support from Pfizer during the conduct of this study that was paid directly to Kaiser Permanente Southern California. BKA has received research support for work unrelated to this study provided by Moderna, Dynavax, GSK, Novovax, and Genentech. JMS has received research support from ALK, Dynavax, and Novavax for work unrelated to this study. HT has received research support from GSK, Moderna, ALK, and Wellcome Trust for work unrelated to this study. TBF has previously owned stock in Pfizer. SYT received research support from Genentech and GSK for work unrelated to this study. Download .pdf (.36 MB) Help with pdf files Supplementary appendix