Background
Prospective trials have not previously compared the efficacy of boosting with the ancestral/monovalent mRNA vaccine (WA-1) versus a bivalent ancestral plus BA.4/BA.5-targeted vaccine in African countries, with high HIV and SARS-CoV-2 seroprevalence.
Methods
The second phase of CoVPN 3008 (Ubuntu trial) was a double-blind randomized trial in 7 countries, among people with HIV (PWH) or another comorbidity linked to severe Covid-19. It activated during the BA.4/5 omicron wave, progressing through to XBB.1. At month 6 (M6), previously-vaccinated volunteers received 100 mcg of either a monovalent or bivalent boost (1:1), then monthly PCR testing until M18. We compared the risk of symptomatic Covid-19 between arms using cumulative incidences and Cox regression overall and by HIV status, SARS-CoV-2 serology (SARS+/SARS-), CD4 count and HIV viremia. We assessed neutralizing anti-Spike antibody ID50 titers against BA.4/5 and XBB.1.5 (dominant strain during the trial) before boost and 1-month after, in 100 PWH per arm.
Results
From 10/2022-03/2023, 3988 participants (3086 PWH; 3491 SARS+) received the monovalent (n=1976) or bivalent (n=2012) boost and were followed until 05/2024. At M6, characteristics were balanced by arm with 68% born female and median age 37 years overall; among PWH, median CD4 count was 653 cells/mm3, 18% had HIV viremia and 95.8% were on antiretroviral therapy. At M12 (6 months after boosting), we observed 113 Covid-19 cases (all Omicron) including 1 non-hospitalized severe case. Incidence was 5.58 per 100 person-years, with no difference in efficacy overall (HR=1.00, 95% CI [0.69,1.45], Fig 1); by HIV status (PWH HR=0.99 [0.64,1.51]; PWoH HR=1.03 [0.50,2.14]); by prior SARS-CoV-2 infection (SARS+ HR=0.98 [0.66,1.46]; SARS- HR=1.17 [0.42,3.47]); or by CD4 count or HIV viremia (not shown). HRs were similar at M18. Geometric mean titers to BA.4/5 increased from 835 at M6 to 2544 at M7 in the monovalent arm vs. 773 to 3639 in the bivalent; and to XBB.1.5 from 135 to 559 in the monovalent arm vs. 156 to 1191 in the bivalent (Fig 1). The <2 fold higher titers among bivalent recipients were not associated with discernable clinical benefit. Severe disease was rare based on daily O2 readings, temperatures and symptom checks following each positive PCR.
Conclusions
Among PWH, despite modestly higher neutralizing titers to circulating strains of Omicron after bivalent boosting, the bivalent boost provided no added protection against symptomatic Covid-19 over the monovalent boost.