Comparative immunogenicity and neutralizing antibody responses post heterologous vaccination with CoronaVac (Sinovac) and Vaxzevria (AstraZeneca) in HIV-infected patients with varying CD4+ T lymphocyte counts

ABSTRACT

The immune response to heterologous coronavirus disease (COVID-19) vaccination in people living with HIV (PLWH) is still unclear. Herein, our prospective cohort study aimed to compare the immune response of heterologous vaccination with CoronaVac (Sinovac) and Vaxzevria (AstraZeneca) between PLWH having CD4 counts ≤ 200 cells/µL (low CD4+) and > 200 cells/µL (high CD4+). Anti-receptor-binding domain (RBD) immunoglobulin G (IgG) levels and the percentage inhibition of neutralizing antibodies (nAbs) were analyzed at 2 and 12 weeks after immunization. Participants in the low and high CD4+ groups had mean CD4+ counts of 139 and 575 cell/µL, respectively. Two and 12 weeks after immunization, in the low CD4 group, the median anti-RBD-IgG levels were 159 IU/mL and 143 IU/mL, respectively, whereas the nAb level was 71% and decreased to 47.2%, respectively. Contrarily, the median anti-RBD-IgG levels in the high CD4+ group were 273 IU/mL and 294 IU/mL, respectively, whereas the nAb levels were 89.3% and relatively stable at 81.6%. However, although immune responses between the two study groups were not significantly different, a decline in nAb levels was observed at 12 weeks in the low CD4+ group. Therefore, a COVID-19 booster vaccine dose is suggested for immunoprotection.

KEYWORDS:

• COVID-19
• people living with HIV
• SARS-CoV-2 vaccine
• anti-receptor-binding domain immunoglobulin G
• neutralizing antibodies

Introduction

The coronavirus disease (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been associated with worse outcomes in patients with comorbidities. Immunosuppressive status is associated with a high risk of COVID-19 mortality. This mortality rate has not increased among people living with human immunodeficiency virus (PLWH) in the United States and Europe. In contrast, the mortality rate in UK and South Africa have doubled.^1,2 Low CD4+ T lymphocyte counts, non-suppressible human immunodeficiency virus (HIV) viral load, and co-infection with tuberculosis are risk factors for severe COVID-19 in PLWH.^3,4 Additionally, a higher burden of comorbidities associated with severe outcomes is observed in PLWH than in the general population. Cardiovascular disease, chronic lung disease and older age are more prevalent among PLWH.⁵ PLWH who had three or more comorbidities showed a significant risk of hospitalization and severe outcomes compared to PLWH with no comorbidities.⁶ Following the natural infection of SARS-CoV-2, the development of SARS-CoV-2 IgG concentration and pseudovirus neutralizing titers were both lower response among PLWH than among people without HIV. This mirrors the fact that the immune response in PLWH post-vaccination is lower than that in the general population.⁷ Vaccination against SARS-CoV-2 is considered a crucial approach for controlling the pandemic, particularly in individuals at a high risk of severe disease. Potential concerns are the immunogenicity and efficacy of SARS-CoV-2 vaccines in PLWH. Despite having a high CD4+ T lymphocyte count, humoral and cellular immunity are reduced in the magnitude of response and durability of protection. However, the efficacy of vaccines can be improved by increasing doses or booster schedules.⁸ Restoration of the immune system using highly active antiretroviral therapy also improves the immune response to vaccines, regardless of the CD4+ T lymphocyte count.^9,10 Several vaccines generate lower antibody levels than those generated by immunocompetent hosts. Vaccination against influenza virus, hepatitis A virus, hepatitis B virus, rabies virus, Streptococcus pneumoniae, and Neisseria meningitidis in PLWH having low CD4+ T lymphocyte counts exhibit poorer immunogenicity than in those with relatively higher CD4+ T lymphocyte count. The durable of neutralizing antibody is a point of concern. The antibody response for several vaccines wanes more rapidly among PLWH.^11–16 However, the efficacy of these vaccines in patients with advanced HIV is well-established.^17–19 Therefore, these vaccines are still recommended to protect PLWH, who are at a considerable risk of developing severe disease.

Several cohort studies on the immunogenicity of SARS-CoV-2 vaccines in PLWH have used mRNA and vector vaccines. In a study in Israel, 97–98% of PLWH vaccinated with BNT162b2 developed anti-receptor binding domain (RBD) immunoglobulin (IgG) and neutralizing antibodies, similar to those in the HIV-negative population. However, 95% of the participants had undetectable HIV viral load (VL), with a median CD4+ T lymphocyte count of 700 cells/µL.²⁰ In a prospective cohort study including PLWH with different CD4+ T lymphocyte counts and vaccinated with mRNA vaccines (57% received BNT162b2 and 43% received mRNA-1273), a significant reduction in anti-RBD IgG and neutralizing antibody levels were observed in those with current CD4+ T lymphocyte counts < 200 cells/µL (86.7% and 70%, respectively). However, the immune response in PLWH with current CD4+ T lymphocyte counts > 500 cells/µL was comparable to that in healthcare workers.²¹ Conversely, PLWH receiving mRNA vaccines have a 39.3% lower antibody concentration than that among HIV-negative controls, according to a nationwide prospective cohort in the Netherlands. However, the antibody response in PLWH with CD4+ T cell counts of 250–500 cells/µL and above is still significantly higher than those with counts < 250 cells/µL. Those receiving the vector vaccine have a 39.4% lower antibody concentration than those receiving the mRNA vaccine. Factors associated with no response to vaccination are as follows: receiving a vector vaccine, age >65 years, and having an HIV load > 50 copies/mL. Having an HIV status is associated with hyporesponsiveness to both mRNA vaccines and vector vaccines, compared to the control, although 86.4% of PLWH have current CD4+ T cells > 250 cells/µL.²² In contrast to the results for vector vaccines in South Africa, two doses of ChAdOx1 nCoV-19 vaccines showed similar antibody and neutralizing responses in PLWH and non-HIV participants. Robust immune responses in PLWH with previous SAR-CoV-2 infection were observed to exceed that in naïve participants.²³ Similar to that noted in the UK cohort, there was no difference in the antibody response against the SARS-CoV-2 spike protein between PLWH and those without HIV. The IFN-γ Elispot responses and T cell proliferation were similar in both groups.²⁴ Furthermore, three studies of immunogenicity of inactivated SARS-CoV-2 vaccine in PLWH were conducted in China. PLWH who have completed two doses of BBIBP-CorV (Sinopharm) or CoronaVac (Sinovac) vaccination display similar neutralizing antibody concentrations as HIV-negative participants. However, the antibody response had lower magnitude and was delayed among PLWH. Besides humoral responses, the specific T cell response was evaluated by determining cytokine expression. PLWH exhibit a decrease in Th2 and Th17 cytokine production. The neutralizing antibodies are positively correlated with CD4 T cell and B cell counts. These results indicate that the inactivated SARS-CoV-2 vaccines are immunogenic in PLWH, but the vaccines induce a lower immune response in those with lower CD4 T lymphocyte counts.^25–27

In Thailand, owing to inaccessibility to an mRNA COVID-19 vaccine in the early phase of the pandemic, two doses of CoronaVac (2–4 weeks apart) or Vaxzevria (10–12 weeks apart) were endorsed as a national guideline.²⁸ However, the procurement and delivery of the vaccines during pandemic have been uncertain. A first report of heterologous immunization showed that prime-boost regimen with ChAdOX1 nCoV-19 and BNT162B2 improves immunogenicity and efficacy compared with homologous ChAdOx1 nCoV-19 vaccination and BNT162B2 vaccination.²⁹ The immunogenicity of heterologous vaccines in Thailand showed that the mean antibody levels of the CoronaVac (Sinovac)-prime and Vaxzevria (AstraZeneca)-boost series (4 weeks apart) were comparable to those of patients who received two doses of Vaxzevria (10 weeks apart) (797 U/mL and 818 U/mL, respectively). The antibody responses observed among these two groups were significantly higher than that with two doses of CoronaVac (96.4 U/L).³⁰ Based on the limited vaccine supply vaccine and efficacy of heterologous vaccines, a CoronaVac (Sinovac)-prime and Vaxzevria (AstraZeneca)-boost schedule has been implemented across the country, according to the previous study. Thus, there are limited data on immunogenicity in PLWH for these heterologous primary schedules.

Consequently, this study aimed to compare the anti-RBD IgG and COVID-19 neutralizing antibody responses on day 14 and 84 after heterologous prime boost with CoronaVac-Vaxzevria between PLWH with a CD4+ counts ≤ 200 cells/µL and those with a CD4+ counts > 200 cells/µL.

Materials and methods

Study design and participants

This prospective cohort study was conducted in an outpatient clinic at Songklanagarind Hospital, a teaching-based tertiary care center in Thailand, from November 2021 to March 2022. The process of the enrollment was not blinded and randomized. The participants who received CoronaVac (Sinovac) 0.5 mL intramuscularly were subsequently administered Vaxzevria (AstraZeneca) 0.5 mL intramuscularly with a 4-week interval between administrations were recruited. Participants were included according to the following criteria:

• age over 18 years

• documented diagnosis of HIV infection

• receiving antiretroviral treatment (ART)

• undetectable HIV RNA VL (<20 copies/mL)

Patients were excluded from the study according to the following criteria:

• pregnant

• previously received any COVID-19 vaccine

• had a history of SARS-CoV-2 infection

• had signs or symptoms of active COVID-19

• had received any immunosuppressive agents

• had received or anticipated the requirement of any blood product or immunoglobulin preparation within 1 month

• had a history of hospitalization within 1 month

• had an active opportunistic infection

The participants were also withdrawn from the study if they did not receive CoronaVac followed by Vaxzevria and if they did not undergo a blood test for antibody level measurement within 14 days of the study protocol’s assigned date.

Baseline demographic data, current CD4+ T cell lymphocyte count, duration of undetectable HIV VL, current antiretroviral regimen, and presence of comorbidities were extracted from medical records after recruitment of participants.

Ethics approval

The study protocol was reviewed and approved by the Ethics Committee of Songklanagarind Hospital (REC-64-526-14-1). Written informed consent was obtained from each patient before enrollment.

Sample size calculation

We estimated that a sample size of 30 participants would have more than 80% power (with a one-sided α of 0.05) to show the superiority of antibody levels in PLWH with CD4+ counts > 200 cells/µL compared to that observed for a CD4+ count of ≤ 200 cells/µL using a margin of 50 IU/mL of anti-RBD IgG.

As in the following sample size calculation

$n_2=\frac{{\left({\mathit{z}}_{1-\mathit{\alpha }}+z_{1-\mathit{\beta }}\right)}^2{\sigma }^2\left(1+\frac{1}{k}\right)}{{\left(\mathit{ϵ}-\mathit{\delta }\right)}^2}$

$\mathit{ϵ}=u_2-u_1$

$\mathit{k}=\frac{n_1}{n_2}$

$n_1=\mathit{k}{n}_2$

Standard deviation (σ) = 200.00

Non-inferiority or superiority margin (δ) = 50.00

Ratio between two groups (k) = 1.00

Alpha (α) = 0.05, Z(0.950) = 1.644854

Beta (β) = 0.20, Z(0.800) = 0.841621

The hypothesis for the primary objective was that the anti-RBD IgG at day 14 after complete vaccination in the PLWH with CD4+ counts ≤ 200 cells/µL (low CD4+ group) was 30% lower than those with CD4+ counts > 200 cells/µL (high CD4+ group). Generally, several inactivated vaccines are suboptimal response in advanced HIV-infected patients. The seroconversion rate of influenza vaccine, hepatitis A vaccine and hepatitis B vaccine in PWHL with low CD4+ counts were lower 35%, 13% and 22% than those with high CD4+ counts, respectively.^11,13,14 Therefore, we estimated the antibody response was lower 30% in low CD4+ group than those in the high CD4+ group. The anti-RBD IgG of CoronaVac-Vaxzevria in a healthy Thai population is 797 IU/mL.³⁰ Hence, we assumed that the anti-RBD IgG level in PLWH with a CD4+ count > 200 cells/µL would not differ from that in the healthy population. The mean difference between the two groups (ε) was 239.00 IU/mL.

Using a sample size (n) of 14 in each group, we estimated a dropout rate of 10%. Accordingly, 30 participants were required to complete the primary objective analysis.

Procedure

Three blood tests were performed via direct venipuncture at the following time points: 1) day 0 (baseline prior to vaccination): 9 mL of blood was collected for measuring CD4+ T lymphocyte counts and anti-RBD IgG levels, 2) day 14 (following complete vaccination), and 3) day 28 (following complete vaccination): 6 mL of blood was collected on day 14 and 28 for measuring anti-RBD IgG levels and performing neutralization antibody assays.

Primary and secondary endpoints

The primary endpoint was the anti-RBD IgG and neutralizing antibody on day 14 after heterologous prime boost with CoronaVac-Vaxzevria between PLWH with a CD4+ counts ≤ 200 cells/µL and those with a CD4+ counts > 200 cells/µL. The secondary end point was the anti-RBD IgG and neutralizing antibody on day 84 after vaccination between two CD4+ groups.

Study tools and outcome measurement

The CD4+ T lymphocyte counts were measured via flow cytometry through BD Tritest™ CD3/CD4/CD45 (BD Biosciences, San Jose, USA) on the BD FACSCalibur™ flow cytometer, according to the manufacturer’s instructions. The results are reported in cells/µL.

Anti-RBD IgG levels were evaluated using Elecsys anti-SARS-CoV-2 S on a Cobas e 801 analyzer, according to the manufacturer’s instructions.³¹ The analysis involved a double-antigen sandwich enzyme-linked immunosorbent assay. The antigens in the reagent predominantly capture anti-SARS-CoV-2 IgG. The positive cutoff for test was 0.8.

A neutralization assay was performed using the surrogate virus neutralization test on the cPass™ kit (GenScript, USA) according to the manufacturer’s instructions.³² Values ≥ 30% inhibition cutoff were considered as the presence of SARS-CoV-2 neutralization antibodies. The seroconversion rate was defined as %inhibition ≥68%, based on the United States Food and Drug Administration (US-FDA) guidance for high titer COVID-19 convalescent plasma.³³

Statistical analyses

The demographic variables were also assessed. Descriptive statistics were used for all variables. Categorical variables were compared between groups using Fisher’s exact test or the chi-square test. Comparisons of continuous variables between groups were performed using the t-test or Wilcoxon rank-sum test. Statistical significance was set at p < .05. All analyses were performed using R Program 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria, 2018).

Results

Participants demographic and clinical characteristics

Thirty participants were included in the study. The baseline demographics and underlying diseases are shown in Table 1. The median CD4+ T lymphocyte count in the low and high CD4+ groups was 139 cells/µL (inter-quartile range (IQR) 129, 149.5) and 575 cells/µL (IQR 476.5, 681), respectively. Nucleoside reverse transcriptase inhibitors (NRTIs) in ART were similar in both groups. Tenofovir disoproxil fumarate (TDF) or tenofovir alafenamide (TAF) and 3TC (lamivudine) or FTC (emtricitabine) were the most commonly used as the NRTI backbone. The use of non-nucleoside reverse transcriptase inhibitors (NNRTIs) was significantly higher in the low CD4+ group (86.7%) than in the high CD4+ group (40%). The duration of ART exposure was longer in the high CD4+ group, but the difference was not statistically significant. The two groups differed significantly in the duration of undetectable VL. The low and high CD4+ groups showed median ART durations of 42 months (IQR 17, 197) and 139 months (IQR 110, 156), respectively, whereas the undetectable VL duration was 8.2 months (IQR 2.4, 78) and 158 months (IQR 119, 169), respectively. Majority of the participants had no underlying conditions; no significant difference was observed in comorbidities between the two groups. None of the participants demonstrated the presence of anti-RBD IgG at baseline prior to vaccination. Therefore, none of the participants had a history of SARS-CoV-2 infection before inclusion.

Table 1. Baseline demographic data and clinical characteristics of PLWH with CD4 ≤ 200 cells/µL and CD4 > 200 cells/µL.

	CD4 ≤ 200 cells/µL (n = 15)	CD4 > 200 cells/µL (n = 15)	P-value
Age, year (IQR)	47 (33, 55)	47 (43.5, 50.5)	.427
Male sex, n (%)	10 (66.7)	8 (53.3)	.709
CD4, cu.mm (IQR)	139 (129,149.5)	575 (476.5,681)	<.001
Current ART regimen, n (%)
NRTI backbone			.749
− 3TC/FTC+TDF/TAF	11 (73.3)	12 (80)
− 3TC+AZT	1 (6.7)	1 (6.7)
− TDF+AZT	1 (6.7)	1 (6.7)
− 3TC	2 (13.3)	0
Non-NRTIs			.004
− NNRTIs	6 (40)	13 (86.7)
− PIs	5 (33.3)	1 (6.7)
− INSTIs	4 (26.7)	0
Undetectable VL duration, months (IQR)	8.2 (2.4, 78)	139 (110, 156)	.004
Duration of ART, months (IQR)	42 (17, 197)	158 (119, 169)	.111
Underlying disease, n (%)			1
None	13 (86.7)	13 (86.7)
Mild hemophilia A	0	1 (6.7)
Metabolic disease	1 (6.7)	1 (6.7)
Panic disorder	1 (6.7)	0

IQR, inter-quartile range; VL, viral load; ART, antiretroviral treatment; NRTI, nucleoside reverse transcriptase inhibitor; 3TC, lamivudine; FTC, emtricitabine; TDF, tenofovir disoproxil fumarate; TAF, tenofovir alafenamide; AZT, zidovudine; NNRTIs, non-nucleoside reverse transcriptase inhibitors; PIs, protease inhibitors; INSTIs, integrase strand transfer inhibitors; PLWH, people living with HIV.

Anti-RBD IgG and neutralizing antibody response

The anti-RBD IgG and neutralizing antibody responses are shown in Table 2. The anti-RBD-IgG levels in the low CD4+ group were 159 IU/mL (IQR 35.8, 492) and 143 IU/mL (IQR 77.3, 344.5) on days 14 and 84, respectively. In the high CD4+ group, these values were 273 IU/mL (IQR 154, 791) and 294 IU/mL (IQR 140, 551.5) on days 14 and 84, respectively. However, the difference was not statistically significant (Figure 1).

Figure 1. Box plot of anti-RBD IgG responses on day 14 and 84 after vaccination.

Table 2. Anti-receptor binding domain (RBD) immunoglobulin G (IgG), neutralizing antibody responses, and seroconversion rate of neutralizing antibody.

	CD4 ≤ 200 cells/µL (n = 15)	CD4 > 200 cells/µL (n = 15)	P-value
Anti-RBD IgG (U/mL), median (IQR)
Day 14 after vaccination	159 (35.8, 492)	273 (154, 791)	.3
Day 84 after vaccination	143 (77.3, 344.5)	294 (140, 551.5)	.178
Neutralizing Ab (% inhibition), median (IQR)
Day 14 after vaccination	71 (49.4, 92.9)	89.3 (72.4, 93.8)	.089
Day 84 after vaccination	47.2 (42.3, 85)	81.6 (61.8, 86.9)	.112
Seroconversion of neutralizing Ab, n (%)
Day 14 after vaccination	8/15 (53.3)	12/15 (80)	.24
Day 84 after vaccination	7/15 (46.7)	10/15 (66.7)	.46

Neutralizing antibody responses against SARS-CoV-2 were elicited in 71% (IQR 49.4, 92.9) of the patients on day 14 and decreased to 47.2% (IQR 42.3, 85) on day 84 in the low CD4+ group. The neutralizing antibody levels were insignificantly higher in the high CD4+ group, at 89.3% (IQR 72.4, 93.8) on day 14, and slightly decreased to 81.6% (61.8, 86.9) on day 84 (Figure 2).

Figure 2. Box plot of neutralizing antibody responses on day 14 and day 84 after vaccination.

According to the US-FDA guidance on high titer COVID-19 convalescent plasma, inhibition using the cPass™ SARS-CoV-2 Neutralization Antibody Detection Kit should be ≥ 68%, which was determined as seroconversion. On day 14, 8/15 (53.3%) participants in the low CD4+ group achieved seroconversion compared to 12/15 (80%) participants in the high CD4+ group (p = .24). On day 84, 7/15 (46.7%) participants in the low CD4+ group achieved seroconversion compared to 10/15 (66.7%) participants in the high CD4+ group (p = .46). No significant differences were observed between the two groups.

Discussion

To our knowledge, this prospective observational cohort study is the first to report data on the immunogenicity of heterologous vaccination with CoronaVac, followed by Vaxzevria, in PLWH in the high and low CD4+ groups. Our results demonstrated that the antibody and neutralization responses in the high CD4+ group were higher than those in the low CD4+ group on day 14 and 84 after complete vaccination. However, the difference was not statistically significant.

We observed a weaker immune response in PLWH, even in the high CD4+ group, relative to that observed in a previous study conducted in a healthy Thai population after CoronaVac-Vaxzevria vaccination.³⁰ This finding is consistent with those of an immunogenic study using inactivated vaccines (CoronaVac and BBIBP-CorV) in PLWH.³⁴ Regardless of CD4+ T lymphocyte count and HIV VL, PLWH manifested lower antibody responses, neutralizing activities, and T cell – specific immune responses than those in HIV-negative participants. Notably, in a subgroup with CD4+ T lymphocyte counts > 500 cells/µL and undetectable HIV VL, the neutralizing activity was similar to that in the control group.³⁴ Vaccination in patients with advanced HIV infection results in poor immunogenicity, which is similar to our findings.^21,35 Moreover, each increase of 100 CD4+ T lymphocytes increases the levels of antibody responses and neutralizing activities.³⁶ All participants in our cohort were relatively young and most had no comorbidities to dysregulate the immune response to vaccination. Other than HIV, the comorbidities associated with chronic inflammation that contribute to the reduction of vaccine efficacy, such as cardiovascular disease, hypertension, and chronic pulmonary disease, were not presented in this study. We additionally selected PLWH with undetectable HIV VL because chronic HIV viremia has counteracted with the peak of antibody level, even in individuals with low viremia burden.³⁷ The low anti-RBD IgG level and surrogate virus neutralization test (sVNT) response have been described in most participants with HIV VL > 200 copies/mL.³⁵ Therefore, in our cohort, the low CD4+ T cell number is the single factor related to the reduction of the immunogenicity of heterologous vaccination. This is because the progressive memory B cell loss is significant in chronic HIV infection, and the low memory B cell number correlates with the low CD4+ T cell number. The progressive depletion of helper CD4+ T cells in advanced HIV infection impairs the differentiation of naïve B cells to memory B cells. Chronic immune exhaustion in even HIV patients with viral suppression impacts the vaccine responsiveness. The ART is unable to restore the memory B cell function in chronic HIV infection. Therefore, immunogenicity to the vaccine is suboptimal despite optimal ART treatment.³⁸ However, our results are inconsistent with those of some studies. This is because majority of the PLWH included in these studies showed high CD4+ T lymphocyte counts, and the sample sizes were not defined to detect the power of vaccine response in PLWH with different CD4+ T lymphocyte counts.^{20,23,24,39,40}

In addition to immunogenicity, waning vaccine protection is also a point of concern. Our study was designed as a longitudinal follow-up for durability of the protective immune response. The neutralizing activity dropped by approximately half after 84 days of complete vaccination in PLWH with low CD4+ T lymphocyte counts, despite similar anti-RBD-IgG levels 14 days after vaccination. In contrast, in PLWH with high CD4+ T cell counts, the neutralizing effect was largely retained. This finding has also been reported in a study by Benet et al., wherein PLWH having a CD4+ count < 200 cells/µL have shown reduced neutralizing activity and no vaccine-induced T cell response, despite having two doses of mRNA vaccine.⁴¹ The durability of the protective effect after vaccination is dynamic and differs among vaccine types. The initial response of sVNT with CoronaVac and BNT162B2 is similar. However, the decline of sVNT after the first 6 weeks in PLWH fully vaccinated with CoronaVac is remarkably more rapid than that in PLWH vaccinated with BNT162B2. sVNT continuously falls with time. The BNT162B2 vaccinees exhibit significantly higher seroconversion of 96% at 6 months after vaccination, whereas 57% of CoronaVac vaccinees maintain seroconversion. The magnitude of neutralizing activity in CoronaVac is significantly lower at 3 months. This study concludes that the protection duration is shorter for PLWH receiving inactivated vaccine than for those receiving mRNA vaccine.⁴² These results are consistent with our study in that the neutralization activity is markedly reduced after 3 months. Despite heterologous vaccination, the durability of vaccine protection is not improved among PLWH with low CD4+ T cell number. In line with observation studies in a meta-analysis, the efficacy of vaccines wanes over time but the effectiveness of vaccines against severe infection remained high over 6 months post-complete vaccination in the general population. Notably, the inactivated vaccine was not included in the study.⁴³

Our data highlight that PLWH, particularly those with baseline CD4+ T lymphocyte counts < 200 cells/µL, receiving heterologous vaccination with CoronaVac-Vaxzevria potentially exhibit lower immunogenicity than that in the healthy population. Although viral suppression was achieved in all PLWH, the immune response was lower in the low CD4+ T cell group than that in the high CD4+ T cell group. To date, the immunological correlation between vaccine protection against COVID-19 and the risk of infection and disease severity is unclear.⁴⁴ Different vaccines elicit different levels of neutralizing antibodies. A modeling study has shown that mRNA vaccines induce a higher neutralizing antibody level than those by vector and inactivated vaccines, and this is strongly associated with immune protection.⁴⁵ The administration of an mRNA vaccine in PLWH with CD4+ T cell counts > 500 cells/µL and an undetectable VL elicits an immunological response comparable to or slightly lower than that in a healthy population.⁴⁶ Moreover, mRNA-1273 demonstrates a better humoral response than that of BNT162b2.²¹ Overall, mRNA vaccines induce a better immune response in PLWH than that of less immunogenic vaccines, including heterologous inactivated vaccines and vector vaccines. A recent systematic review and meta-analysis revealed that SARS-CoV-2 vaccines have higher efficacy for preventing hospitalization, severe infection, and death than for preventing asymptomatic and mild symptomatic disease. However, the evidence does not clarify the difference in the efficacy of different types of vaccines.⁴⁷ The real-world effectiveness of vaccines in PLWH remains limited. A large cohort study across the United States showed that the risk of breakthrough infection is 28% higher in PLWH than in those without HIV. The highest effectiveness of risk reduction was observed for mRNA-1273 followed by those for BNT162 and Ad26.COV2.S.⁴⁸ In a large population study of PLWH in Canada, the effectiveness of vaccines against symptomatic infection, hospitalization and death in PLWH (94–97%) was comparable to that in the general population (82–98%) during a similar pandemic period.⁴⁹ In the SISONKE study, the effectiveness of Ad26.COV2.S vaccine against hospitalization and ICU admission was similar between PLWH and those without HIV among healthcare workers. However, the effectiveness in terms of mortality rate was reduced in PLWH (65%) compared to that in individuals without HIV (83%).⁵⁰ To date, no study of inactivated vaccine or heterologous inactivated vaccine with vector vaccine has been conducted for the assessment of vaccine effectiveness among PLWH.

The main limitation of our study is the lack of a non-HIV control group. However, data on immunogenicity in healthy adults can be compared using the same method of anti-RBD IgG measurement, which showed that the heterologous approach is less immunogenic among PLWH than among healthy volunteers, regardless of CD4+ T cell numbers. Moreover, our observational study had a short follow-up period, and we were unable to compare the decay in antibody levels and neutralizing antibodies until the protection waned. Although the appropriate timing for booster doses is unknown, our finding suggests a booster vaccine dose after 3 months owing to a marked drop in neutralizing antibody levels, particularly in PLWH whose CD4 T cell lymphocytes are ≤ 200 cells/µL and are vaccinated using the inactivated vaccine with vector vaccine schedule. Finally, the sample size was too small to identify the proposed difference of 30%. The difference in anti-RBD IgG levels at 2 weeks between the two groups was 41.7%. The enrollment of 30 participants achieved only 38.4% power of prediction.

Conclusions

CoronaVac-Vaxzevria can elicit immune responses against SARS-CoV-2 in PLWH. The resulting immunogenicity was strongly correlated with the CD4+ T lymphocyte counts prior to vaccination. However, the elicited immune response was lower than that in the healthy population, particularly in PLWH with a CD4±T lymphocyte < 200 cells/µL, although statistical significance was not observed. Thus, with our limited existing circumstances, additional strategies to prevent COVID-19 infection are still required for PLWH, rather than the heterologous CoronaVac-Vaxzevria. For patients receiving heterologous CoronaVac and Vaxzevria, a shortened booster vaccination interval is suggested to ensure immunoprotection, particularly in PLWH with a CD4±T lymphocyte < 200 cells/μL.

Author contributions

Conceptualization, S.Chi. and N.K.; formal analysis, S.Chi., S.Chu., and N.K.; investigation, S.Chi., P.S., S.Chu., S.K., B.C., T.H., P.K., and N.K.; data curation, S.Chi., S.Chu., and N.K.; writing – original draft preparation, S.Chi., P.S, S.K., and N.K; writing – review and editing, S.Chi. and N.K. All authors have read and agreed to the published version of the manuscript.

Acknowledgments

Special thanks to Marzukee Mayeng and Jirawan Jayuphan from the epidemiology unit for their assistance with statistical analyses.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support findings are available on request from corresponding author.

Additional information

Funding

This study was supported by a grant from Songklanagarind Hospital Foundation, the Faculty of Medicine, Prince of Songkla University.