Heterologous ChAdOx1-nCoV19-BNT162b2 vaccination provides superior immunogenicity against COVID-19.

At the height of the COVID-19 vaccination campaign, supply and local availability issues for the approved vaccines—together with worries about rare side-effects (thrombotic thrombocytopenia)—necessitated the switch to heterologous vaccination schedules, commonly known as mixing vaccines. Several studies have now been completed addressing the efficacy and safety of this practice during the battle for immunisation against SARS-CoV-2 and its variants. In The Lancet Respiratory Medicine, David Hillus and colleagues report the interim results of a study from the Charité—Universitätsmedizin in Berlin, Germany, on the safety, reactogenicity, and immunogenicity of homologous and heterologous prime-boost immunisation with the non-replicating chimpanzee adenovirus vaccine ChAdOx1-nCoV19 (AstraZeneca, Cambridge, UK) and mRNA vaccine BNT162b2 (Pfizer-BioNtech, Mainz, Germany). For this prospective observational study, the authors enrolled health-care workers in Berlin who received either homologous ChAdOx1-nCoV19 or heterologous ChAdOx1-nCoV19–BNT162b2 vaccination with a 10–12-week vaccine interval or homologous BNT162b2 vaccination with a 3-week vaccine interval.

The study enrolled 380 participants, with 174 receiving homologous BNT162b2, 38 receiving homologous ChAdOx1-nCoV19, and 104 receiving ChAdOx1-nCoV19–BNT162b2 vaccination. Heterologous ChAdOx1-nCoV19–BNT162b2 vaccination was well tolerated, and its reactogenicity was similar to that of homologous vaccination. Antibodies against spike S1 subunit, receptor binding domain (RBD), and nucleocapsid antigens were quantitated by enzyme-linked antigen-printed microarrays. The nucleocapsid antibodies discriminated between individuals who were vaccinated and those who had viral infection, who were excluded. In subsequent analyses, Hillus and colleagues assessed neutralising antibodies from serum samples of participants using HRP-conjugated recombinant RBD (representing a surrogate SARS-CoV-2), which binds to the ACE2 receptor. This surrogate virus was pre-incubated with diluted serum and then incubated with ACE-2-coated plates (medac, Wedel, Germany). Alternatively, the authors did pseudovirus neutralisation tests (pNT) with HIV-1 pseudotypes containing the spike protein genes of alpha (B.1.1.7) and beta (B.1.351) SARS-CoV-2 variants. Reciprocals of values for diluted neutralising antibodies that inhibited infection of ACE2-293 cells by 50% were called the 50% inhibitory dose (ID50). Additionally, the authors characterised IgG avidity by incubating a 1:100 dilution of serum (in the presence or absence of urea) with subunit S1-coated ELISA plates for 10 min. Bound IgG in the presence or absence of urea was used to calculate the avidity index. Finally, spike protein-specific T-cell responses were measured with an ELISA designed to detect interferon-γ (IFN-γ) released from S1 peptide stimulated T-cells.

Most of the findings were straightforward. Spike-binding IgG responses were similar among participants vaccinated with ChAdOx1-nCoV19–BNT162b2, homologous ChAdOx1-nCoV19, or homologous BNT162b2. However, the median relative avidity index was higher after heterologous ChAdOx1-nCoV19–BNT162b2 prime-boost (93·6%, IQR 91·9–95·5) compared with both homologous ChAdOx1-nCoV19 (71·7%, 64·8–77·4, p=0·0026), and homologous BNT162b2 prime-boosts (73·9%, 63·0–81·6, p=0·014). Inhibition of surrogate virus neutralisation was found to be lower after homologous ChAdOx1-nCoV19 immunisation (median 92·4%, IQR 86·4–96·4) compared with homologous BNT162b2 (96·6%, 95·5–97·2, p=0·034) and ChAdOx1-nCoV19–BNT162b2 (97·1%, 96·9–97·3, p<0·0001). Antibody neutralisation assays directed against alpha and beta pseudoviruses showed that ChAdOx1-nCoV19–BNT162b2 prime-boost inhibited at an ID50 of 956·6 (95% CI 835·6–1095·0) for the alpha variant and 417·1 (349·3–498·2) for the beta variant. Serum samples from participants who received homologous BNT162b2 inhibited the pseudovirus variants, with ID50 of 369·2 (310·7–438·6) for alpha and 72·4 (60·5–86·5) for beta. Serum samples from participants who received homologous ChAdOx1-nCoV19 also inhibited the alpha (ID50 212·5, 131·2–344·4) and beta (48·5, 28·4–82·8) variants. Clearly, the alpha variant was more susceptible than the beta variant to neutralising antibodies. T-cell reactivity measured by IFN-γ release from S1-stimulated lymphocytes was greatest after ChAdOx1-nCoV19–BNT162b2 vaccination, followed by homologous BNT162b2 vaccination and homologous ChAdOx1-nCoV19 vaccination.

We noted that the difference in time interval between two doses of BNT162b2 vaccine (homologous BNT162b2, received 3 weeks apart) and between the ChAdOx1-nCoV19–BNT162b2 and homologous ChAdOx1-nCoV19 regimens (10–12 weeks apart) could reduce antibody production and T-cell reactivity in the homologous BNT162b2 vaccine group. The order of inoculation of the prime-boost in the heterologous schedule should also have been tested (ie, BNT162b2–ChAdOx1-nCoV19) because another study has shown that this can reduce immunogenicity. Another aspect of slight concern was the smaller number of participants (38 individuals) in the homologous ChAdOx1-nCoV19 group. It also remains to be shown whether the Moderna vaccine (elasomeran, Moderna, Cambridge, MA, USA) behaves in a similar manner as BNT162b2 in the context of heterologous vaccination. It would also be interesting to test the neutralising antibodies against the delta variant. Finally, the duration and degree of protection over time will also be of interest. Overall, Hillus and colleagues' study appears to be well conceived and carefully executed.

Small pilot studies describing heterologous vaccination against COVID-19 have previously been reported in the literature.3, 4, 5 An investigation in mice established groundwork for human cohort studies that used mixed ChAdOx1-nCoV19–mRNA vaccine regimens.2,6–10 Another study used flow cytometry to show that levels of stimulated CD4+ helper and CD8+ cytotoxic T lymphocytes producing key cytokines were elevated after ChAdOx1-nCoV19–BNT162b2 and homologous BNT162b2 vaccination compared with those after homologous ChAdOx1-nCoV19 vaccination. Two other studies have also shown increased levels of spike-specific CD4+ and CD8+ lymphocytes and neutralising activity against alpha and beta variants after ChAdOx1-nCoV19–BNT162b2 and homologous BNT162b2 vaccination.9, 10 Finally, a very recent study has suggested that the order of immunising agents (ChAdOx1-nCoV19–BNT162b2 vs BNT162b2–ChAdOx1-nCoV19) and the time interval between primary and secondary inoculation might have to be optimised to 10–12 weeks to deliver its full benefits. Clearly, heterologous vaccination in conjunction with mRNA vaccines has emerged as a key weapon in the arsenal against COVID-19.

I declare no competing interests.