Some HIV Drugs Can Cause Overestimation of Antibody Activity in Lentivirus-Based COVID-19 Neutralization Assays
Drug interference behind high neutralization readings
Some antiretroviral drugs taken by people with HIV may interfere with COVID-19 antibody neutralization assays, making antibody activity appear higher than it is, according to a study. The interference was observed in lentivirus-based pseudovirus assays. When assessed using authentic SARS-CoV-2 and vesicular stomatitis virus (VSV)-based assays, neutralizing antibody levels in people with HIV were similar to or somewhat lower than those in controls. The findings do not show that the drugs reduced the actual protective effectiveness of vaccines.
Published in Scientific Reports, the study examined assay interference alongside differences in HIV-related immune characteristics while analyzing immune responses after mRNA vaccination. The researchers investigated not only CD4 T-cell depletion but also how T-cell populations responding to HIV antigens were associated with vaccine immune responses.
Samples analyzed up to 383 days after two doses
The researchers analyzed plasma and peripheral blood mononuclear cells from a prospectively recruited cohort of 22 people with HIV and 30 healthy controls, all of whom had received two doses of a Pfizer or Moderna mRNA vaccine. Participants with a history or serological evidence of COVID-19 were excluded. Samples were grouped by time after completion of the two-dose series: 21–150 days, 151–230 days, and 231–383 days. Some participants provided repeat samples over more than a year.
All participants with HIV were receiving antiretroviral therapy. Their CD4 T-cell counts ranged from 13 to 1,465 cells per µL of blood, reflecting a wide range of immune status. The researchers measured antibody types and functions, as well as T-cell responses to stimulation with SARS-CoV-2 spike and HIV-1 Gag antigens. This was not a clinical trial that directly assessed protection against infection or severe disease.
Inhibitory activity persisted after IgG removal
In the lentivirus-based pseudovirus neutralization assay, measured neutralizing antibody titers in participants with HIV were higher than those in controls. However, inhibitory activity remained in samples from participants taking integrase strand transfer inhibitors (INSTIs) even after IgG antibodies were removed from plasma. No evidence of the same residual activity was found in the comparison group taking non-nucleoside reverse transcriptase inhibitors (NNRTIs).
Among SIV-infected nonhuman primates that had not been vaccinated against COVID-19, samples from animals receiving antiretroviral therapy also showed greater inhibitory activity in the lentivirus-based assay. By contrast, when human samples were evaluated using an authentic SARS-CoV-2 microneutralization assay and a VSV-based pseudovirus assay, the higher neutralization readings in the HIV group were not reproduced. This shows how inhibition of the assay system by drugs in samples can be misread as antibody-mediated neutralization.
Immune responses were not uniformly weaker
At some post-vaccination time points, the HIV group had lower levels of spike-specific total IgG, IgG1, and IgG3, as well as lower antibody-dependent cellular phagocytosis (ADCP). ADCP is a function that helps cells engulf targets recognized by antibodies. In contrast, spike-specific IgA2 levels were higher. For antibodies against the spike receptor-binding domain (RBD), increases in IgG1 and IgA1 and a decrease in IgG3 were observed. The direction of change differed according to antibody target and subclass.
Spike-reactive CD4 T-cell responses were similar in the two groups early on, but were less well maintained over time in the HIV group. Spike-reactive CD8 T-cell responses were similar between groups throughout the observation period. Network analysis identified the proportion of CD4 T cells as a key variable linking antibody and cellular immune measures, while Gag-specific T cells were associated with antibody measures. Additional analyses also found associations between some Gag-specific T-cell populations and spike antibody subclasses, but these findings do not establish a direct causal mechanism.
Implications for assay choice and vaccine design
The researchers proposed considering assays that avoid drug interference or additional sample processing, such as drug removal or antibody isolation. They also pointed to the need to consider immune characteristics associated with chronic HIV infection in vaccine design research.
However, the sample was small, the median age was 59 years in the HIV group compared with 37 years in the control group, and the groups also differed in sex and racial composition. Coinfections with other chronic viruses were not assessed, and the blood-focused analysis limits the ability to establish mechanisms in tissues. Therefore, these findings alone cannot determine additional vaccination schedules or changes to therapy.
The study design and findings can be reviewed in the provided full-text article and bibliographic and abstract record.
People living with HIV infection (PLWH) often have attenuated responses to infections and vaccination. This study aimed to better understand how HIV-associated inflammation and chronic T-cell activation influenced the immune responses to mRNA vaccination or neutralization assay analyses. PLWH on ART or healthy donor controls were analyzed using systems serology and viral T-cell phenotyping to Spike or HIV-1 Gag peptide stimulation after primary mRNA COVID-19 vaccination. Neutralization assays using a lentiviral pseudovirus construct were compromised by the presence of integrase strand transfer inhibitor (INSTI) drugs in plasma from HIV + subjects taking certain ART. This combination of lentiviral pseudovirus reporter assays and INSTIs led to false positive neutralization results. Spike-specific IgG1, IgG3, IgA1, IgA2, and antibody-dependent cellular phagocytosis (ADCP) were altered post-vaccination in PLWH compared to controls. Network and multivariate analyses revealed post-vaccination outcomes were strongly correlated to CD4 immunodeficiency and Gag-specific T-cells, including effector CD8 T-cells and Th1 CD4 T-cells. Given the growing use of pseudovirus neutralization assays for serological evaluation and mRNA technology in novel vaccines that could be recommended for PLWH Pseudovirus neutralization assays need to be carefully selected to prevent ART drugs in patient samples from impacting results. Spike-specific antibody and CD4 T-cell phenotypes are influenced by both CD4 immunodeficiency and Gag-specific T-cell effector populations. This work has clinical relevance beyond COVID, with future considerations of pseudovirus assay evaluations and mRNA vaccine design for chronically infected hosts.
Comparing vaccine immunogenicity requires distinguishing inhibition of the assay system by drugs in samples from antibody-mediated neutralizing activity. This study shows that high neutralization readings observed in lentivirus-based assays may not be reproduced with other assay methods. It provides a basis for checking medications and assay methods together in studies of people with HIV and for cross-validating results with a different type of neutralization assay. It also provides a reason to examine not only CD4 counts but also antibody subclasses and functions and their associations with HIV-specific T cells. Whether these measures predict actual protection against infection or long-term protection remains to be determined in follow-up studies.