Booster Vaccination Against SARS-CoV-2 Induces Potent Immune Responses in People With Human Immunodeficiency Virus

Abstract Background People with human immunodeficiency virus (HIV) on antiretroviral therapy (ART) with good CD4 T-cell counts make effective immune responses following vaccination against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). There are few data on longer term responses and the impact of a booster dose. Methods Adults with HIV were enrolled into a single arm open label study. Two doses of ChAdOx1 nCoV-19 were followed 12 months later by a third heterologous vaccine dose. Participants had undetectable viraemia on ART and CD4 counts >350 cells/µL. Immune responses to the ancestral strain and variants of concern were measured by anti-spike immunoglobulin G (IgG) enzyme-linked immunosorbent assay (ELISA), MesoScale Discovery (MSD) anti-spike platform, ACE-2 inhibition, activation induced marker (AIM) assay, and T-cell proliferation. Findings In total, 54 participants received 2 doses of ChAdOx1 nCoV-19. 43 received a third dose (42 with BNT162b2; 1 with mRNA-1273) 1 year after the first dose. After the third dose, total anti-SARS-CoV-2 spike IgG titers (MSD), ACE-2 inhibition, and IgG ELISA results were significantly higher compared to Day 182 titers (P < .0001 for all 3). SARS-CoV-2 specific CD4+ T-cell responses measured by AIM against SARS-CoV-2 S1 and S2 peptide pools were significantly increased after a third vaccine compared to 6 months after a first dose, with significant increases in proliferative CD4+ and CD8+ T-cell responses to SARS-CoV-2 S1 and S2 after boosting. Responses to Alpha, Beta, Gamma, and Delta variants were boosted, although to a lesser extent for Omicron. Conclusions In PWH receiving a third vaccine dose, there were significant increases in B- and T-cell immunity, including to known variants of concern (VOCs).

Third dose boosting with either homologous or heterologous combinations of COVID-19 vaccines results in vigorous immune responses [35,36]. A third dose of BNT162b2 protected against infection and severe COVID-19 disease in adults >60 years of age [37]. For PWH, the increased immune responses afforded by booster vaccination may therefore offer protection, help overcome antigenic variation seen in some SARS-CoV-2 strains [38], and reduce the incidence of COVID-19.
We performed qualitative and quantitative assessment of humoral and cellular immune responses to SARS-CoV-2 and circulating VOCs following a third booster dose vaccine in PWH.

Study Design and Cohort
The cohort has been described previously [3]. The study comprised people with HIV in an open-label non-randomized group within the larger multicentre phase 2/3 COV002 trial. Inclusion criteria were age 18-55 years, a diagnosis of HIV infection, virological suppression on ART at enrollment (plasma HIV viral load [VL] <50 copies per mL), and a CD4 count >350 cells/μL. Participants received 2 standard intramuscular doses of the ChAdOx1 nCoV-19 vaccine 4-6 weeks apart, and a third dose of any licensed COVID-19 vaccine after 1 year.
Participants with a history of laboratory-confirmed SARS-CoV-2 infection by anti-N protein immunoglobulin G (IgG) immunoassay (Abbott Architect, Abbott Park, Illinois, USA) at screening were excluded. Participants self-reported COVID-19 infection. Visits on day 0 (pre-ChAdOx1 nCoV-19 vaccine prime), 182 and "Post-Third Dose" were the main study timepoints for immunological analysis. As some participants did not attend their "Post-Third Dose" visit as they were lost to follow-up, there is a maximum of n = 43 at this timepoint. Where possible, we collected peripheral blood mononuclear cells (PBMCs) from participants before and after the third dose booster vaccine dose (n = 9).

SARS CoV-2 Spike IgG ELISA
Humoral responses at baseline and following vaccination were assessed using a standardized total IgG enzyme-linked immunosorbent assay (ELISA) against SARS CoV-2 spike as described previously [2]. Full details are in Supplementary Materials.

T-Cell Proliferation Assay
T-cell proliferation was measured use a CTV assay [3,9]. Full details are in Supplementary Table 2.

AIM Assay
The activation induced marker (AIM) assay was used to identify and characterise antigen-specific T cells [3,9]. Full details are in Supplementary Table 3.

ACE-2 Inhibition Assay
A multiplexed MSD immunoassay (MSD, Rockville, Maryland, USA) was used to measure the ability of human sera to inhibit ACE-2 binding to SARS-CoV-2 spike (B, B.

Statistical Analysis
We analyzed all outcomes in all participants who received specified doses of the vaccination schedule and with available samples, unless otherwise specified. We present medians and interquartile ranges (IQRs) for immunological endpoints. For comparison of 2 non-parametrically distributed unpaired variables, we used the Wilcoxon rank sum (Mann-Whitney U) test. Where multiple data points were compared, we used a Kruskal-Wallis test with Dunn's multiple comparison. For comparison of 2 non-parametrically distributed paired data sets, we used the Wilcoxon matched pairs signed rank test. All analyses were carried out using Prism 9 (GraphPad Software).

Study Approval
Study approval in the United Kingdom was by the Medicines and Healthcare products Regulatory Agency (reference 21584/0424/001-0001) and the South Central Berkshire Research Ethics Committee (reference 20/SC/0145). COV002 is registered with ClinicalTrials.gov, NCT04400838.

Participants
Participants with HIV (n = 54; all male) were recruited as a substudy group in the COV002 clinical trial (NCT04400838) in November 2020. Participants were administered 2 doses of ChAdOx1 nCoV-19 vaccine at day 0 and after 4-6 weeks. They were offered a third dose with a heterologous vaccine around 365 days after their first ChAdOx1 nCoV-19 dose. All participants had undetectable VL (<50 HIV RNA copies/mL) and a median CD4 count of 694 cells/µL (IQR: 573.5-859.5) at the time of recruitment. Ethnicity was mostly White (81.5%). Other reported ethnicities were Asian (3.7%), mixed (7.4%), and other (7.4%). Participants returned for study visits on day 14, 28, 42, 56, 182, and "Post-Third Dose," The "Post-Third Dose" visit was recorded as the first study visit following the third dose of vaccine (mean number of days post third dose = 33, range: 5-115, IQR: . Participants received mostly BNT162b2 vaccine for their day 365 boost (42/ 43; 1/43 received mRNA-1273; Moderna) (Supplementary Table 1). For this study, baseline (Day 0), 6 months (Day 182), and "Post-Third Dose" samples were considered. The introduction of the booster vaccine as National Health Service (NHS) policy by the UK government meant some third doses were given out of sync with the study protocol, and so blood draws before the third dose were not available for all participants. However, for some (n = 9), samples were available either side of the third dose, as pre-and post-third dose visits ( Figure 1A and Table 1). All participants self-reported an absence of SARS-CoV-2 infection at every study visit based on interviews with the study team, and SARS-CoV-2 nucleocapsid responses measured for 6 months after recruitment.

Increased Magnitude of T-Cell Responses After Third COVID-19 Vaccine Dose in PWH
T-cell immune responses were first measured using an ex vivo AIM assay to measure effector-type responses and then a CTV proliferation assay on 7-day expanded cells to quantify recall response. (Flow cytometric gating strategy for AIM and proliferation assays are shown in Supplementary Figures 2A and F,  respectively). Staphylococcal enterotoxin B (SEB) and cytomegalovirus (CMV) responses were used as mitogenic and antigenic control responses in the AIM assay (Supplementary Figure 2B-E), whereas phytohemagglutinin (PHA) and Flu, EBV, CMV, Tetanus (FECT) optimal peptides were used as controls in the proliferation assays (Supplementary Figure 2G-J).
These observed T-cell responses from the AIM assay were also seen when measuring T-cell proliferation, although with a greater magnitude. Proliferative CD4+ and CD8+ T-cell responses to SARS-CoV-2 S1 and S2 following the third dose were significantly greater than responses at baseline (day 0) and day 182 after first dose ( Figure 2E-H ). Analysis of the magnitude of the CD4+ and CD8+ proliferative response following vaccination showed that T-cell responses were primed after initial vaccine, peaking between days 28 and 42, had waned by day 182 [13], and then increased again following the third dose (Supplementary Figure 3A-H). These assays indicate potent boosting of T-cell responses by vaccination and efficient recall upon antigen re-exposure.

Phenotypic Analysis of SARS-CoV-2 Specific Cells Following Booster Vaccination
As we had observed an increase in the magnitude of SARS-CoV-2 T cells following third dose vaccination, we assessed if there were changes in the distribution of the phenotype of the CD4+ T helper cell subsets following the booster vaccine. We first compared the magnitude of all antigen-specific cells within CD4 and CD8+ T-cell compartments using the AIM assay. We observed that despite the recent boost of SARS-CoV-2 spike-specific T cells, CMVpp65-specific T-cell response remained at a higher frequency compared to SARS-CoV-2 spike-specific responses ( Figure 3A and 3B). We then used chemokine receptors CXCR3 and CCR6 to evaluate the distribution of CD4+ T-cell subsets within the antigen-specific AIM+ CD4+ T cells 6 months after priming vaccination and after the third dose. We found no change in the frequency of SARS-CoV-2 spike-specific CD4+ T cells that exhibited a Th1 (CXCR3+ CCR6−), Th17 (CXCR3− CCR6+) or circulating Tfh (CXCR5+) phenotype following a third dose ( Figure 3C, E, and F). We noted an increase in the frequency of Th2 (CXCR3− CCR6−) cells within the CD4+ antigen-specific compartment; however, this was found with all antigens (including CMV) and, in the absence of functional data, larger studies would be needed to determine if this was reproducible ( Figure 3D).

Potent VOC Immune Responses Are Induced Following Booster Vaccines
Finally, we evaluated the magnitude of humoral and T-cell responses to circulating VOCs (including the recently categorized Omicron BA1 variant) after a third dose. Compared to total anti-SARS-CoV-2 spike IgG titers in the ancestral strain, total anti-spike antibody responses to all VOCs were significantly reduced ( Figure 4A). This was also found with the SARS-CoV-2 ACE-2 binding assay, which indicated a decreased potency of neutralising antibodies in the "Post-Third Dose" sample to bind to spike protein from VOCs ( Figure 4B). For VOCs-Alpha, Beta, and Gamma-for which we had historical day 0 and day 182 data, we assessed the kinetics of the antibody response after the third dose. We noted a striking increase in ACE inhibition (Supplementary Figure 1D-F) and antibody titers (Supplementary Figure 4A-C). after the third dose compared to samples tested at baseline and 6 months after the first of the 2 ChAdOx1 nCoV-19 doses. We also investigated T-cell responses to VOCs in comparison to the ancestral SARS-CoV-2 Victoria strain. Similar to our previous report [13], the magnitude of the proliferative CD4+ and CD8+ T-cell response was comparable between the ancestral strain and the Beta, Gamma, and Delta variants -with the exception of the CD8+ T-cell response to SARS-CoV-2 S2 peptide pool. Interestingly, we found the proliferative T-cell response to the Omicron variant targeting both spike S1 and S2 peptide pools was significantly reduced in the CD4 and CD8+ T-cell compartments ( Figure 4C-F). Where sample availability allowed, we compared the kinetics of the T-cell response 6 months after first vaccination and after a third dose, and found an increase in T-cell responses to all variants tested after a third dose, with the sole exception of the CD8 T-cell response to the SARS-CoV-2 Beta variant S1 subunit (Supplementary Figure 4D-O). As Omicron-directed antibody and T-cell responses were significantly lower than the responses to the ancestral SARS-CoV-2 strain, we looked in more detail in participants sampled before and shortly after their third dose of COVID-19 vaccine (n = 9). We found moderate but statistically significant increases in both humoral (Supplementary Figure 5A and B) and CD4+ and CD8 + T cell (Supplementary Figure 5C and D) responses to the Omicron variant after the third dose. Taken together our data show that booster vaccination in PWH significantly boosts antibody and T-cell responses to Alpha, Beta, Gamma, and Delta VOCs, and to a lesser extent to Omicron.

DISCUSSION
We show evidence that a third dose of the licensed COVID-19 vaccines significantly boosted antibody and T-cell responses in PWH (VL undetectable and CD4 count >350 cells/µL). The robust responses generated in our cohort of PWH following heterologous third dose regimen are consistent with reports in people without HIV [39][40][41] and are reassuring, especially as the ChAdOx1 nCov-19 vaccine is well designed for distribution in low-to middle-income countries including those with a significant prevalence of PWH [42].
Equally crucial in the strategic management of the COVID-19 pandemic is that boosted SARS-CoV-2 immune responses can target circulating VOCs, especially as immune escape has been reported [16][17][18]43]. We found humoral responses to VOCs to be boosted although to a lesser degree than responses targeting the ancestral strain. There was no difference between the magnitude of T-cell responses to the VOCs except for the Omicron variant, which was boosted but to lower levels than other VOCs. The relatively high number of mutations on key sites of antibody target including K417N and N501Y in the Omicron spike protein may account for this [13,44]. Interestingly, our data may suggest that antibody immune evasion is more prevalent than T cell escape in immune response to VOCs-whether T cells may therefore play a role in protection from VOC-mediated COVID-19 needs further investigation [45]. Real world data would also be needed to determine if boosted VOC responses confer protection from severe COVID-19 disease in PWH. Finally, the quality of the induced immune response may be impacted by the vaccine platform. For example, there is evidence that the ChAdOx1 nCoV-19 vaccine results in a more dominant Th1-driven response [46]  and mRNA vaccines may induce stronger antibody responses [47], possibly by soliciting Tfh cell help [48][49][50].
Our study has some limitations. We do not have access to a control group of HIV seronegative volunteers tested with the same assays in the same conditions post-boost, and so cannot comment on how the magnitude of immune response in our cohort of PWH would compare to HIV negative controls. We assessed breakthrough infection with SARS-CoV-2 by direct questioning of participants at every study visit. This was supported by nucleocapsid responses, but only for the first six months of the study. Our cohort of PWH represent the scenario of ART suppressed volunteers with an undetectable VL and high CD4 count. This is not the case for many PWH. As such, the data from our cohort should be extrapolated cautiously to other populations with HIV, especially as our cohort was also biaised to male participants in the United Kingdom. Due to the roll-out of the UK vaccination program during the study, we were only able to obtain pre-third dose samples from nine participants. It is therefore difficult to state exactly what the immediate increase in immune response was, although it is clear that the overall response was significantly augmented. Finally, as most participants received the BNT162b2 vaccine as the third dose after the two ChAdOx1 nCoV-19 doses, we did not have the scope to perform a comparative analysis of immune responses following a different third dose vaccine, which may be especially relevant in countries without access to RNA vaccines. In summary, we show a robust booster effect on antibody and T-cell responses to SARS-CoV-2 in PWH after a third dose in a heterologous vaccination schedule.

Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.