mRNA vaccines against SARS-CoV-2 induce divergent antigen-specific T-cell responses in patients with lung cancer

Abstract

Background

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron variant is highly transmissible and evades pre-established immunity. Messenger RNA (mRNA) vaccination against ancestral strain spike protein can induce intact T-cell immunity against the Omicron variant, but efficacy of booster vaccination in patients with late-stage lung cancer on immune-modulating agents including anti-programmed cell death protein 1(PD-1)/programmed death-ligand 1 (PD-L1) has not yet been elucidated.

Methods

We assessed T-cell responses using a modified activation-induced marker assay, coupled with high-dimension flow cytometry analyses. Peripheral blood mononuclear cells (PBMCs) were stimulated with various viral peptides and antigen-specific T-cell responses were evaluated using flow cytometry.

Results

Booster vaccines induced CD8⁺ T-cell response against the ancestral SARS-CoV-2 strain and Omicron variant in both non-cancer subjects and patients with lung cancer, but only a marginal induction was detected for CD4⁺ T cells. Importantly, antigen-specific T cells from patients with lung cancer showed distinct subpopulation dynamics with varying degrees of differentiation compared with non-cancer subjects, with evidence of dysfunction. Notably, female-biased T-cell responses were observed.

Conclusion

We conclude that patients with lung cancer on immunotherapy show a substantial qualitative deviation from non-cancer subjects in their T-cell response to mRNA vaccines, highlighting the need for heightened protective measures for patients with cancer to minimize the risk of breakthrough infection with the Omicron and other future variants.

Correspondence to Dr Zihai Li; [email protected]

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Messenger RNA vaccines targeting SARS-CoV-2 can induce long-lasting cellular responses, particularly with booster shots. Recent studies indicated that patients with cancer might display weaker T-cell responses, possibly owing to their compromised immune systems or cancer therapies. However, the qualitative characteristics of antigen-specific T cells in patients with cancer remain unexplored.

WHAT THIS STUDY ADDS

• Booster vaccines induced CD8⁺ and CD4⁺ T-cell responses in patients with lung cancer, although induction levels were not as robust as those observed in healthy donor samples. On characterizing these antigen-specific T cells, we identified qualitative distinctions in antigen-specific T cells from patients with lung cancer when compared with those in healthy donor samples. Additionally, our study highlights the influence of factors such as sex and responsiveness to immunotherapy on the quality of antigen-specific T cells.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Emerging evidence indicates that the cellular response to COVID-19 plays a crucial role in disease management, and this is often evaluated through the measurement of antigen-specific T-cell frequencies. This manuscript demonstrates that, despite patients with lung cancer exhibiting a persistent cellular response in terms of quantity, its effectiveness may be compromised due to elevated expressions of inhibitory molecules. Consequently, heightened protective measures are essential for patients with cancer, necessitating a deeper understanding of antigen-specific T cells.

Background

The novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and its associated disease syndrome (COVID-19) resulted in a global pandemic of unprecedented modern-day proportions. As of July 2023, over 767 million people have been diagnosed with COVID-19, resulting in more than 6.9 million deaths worldwide.¹ In an effort to curb the deadly impact of this pandemic, multiple strategies have been used to treat and prevent COVID-19 and its most severe symptoms.^2–6 Messenger RNA (mRNA) vaccines were rapidly developed and deployed in the general population, and showed promising efficacy against infection caused by the ancestral strain and earlier variants, especially when combined with booster shots, leading to a remarkable reduction in overall infection rates.^7–12 In November 2021, a highly-infectious variant Omicron emerged,¹³ possessing mutations that facilitated its escape from humoral immunity established by vaccination and prior infection by other variants.^14–17 Certain individuals display weakened T cell-mediated immunity against Omicron following vaccination; for instance, one study showed patients with multiple sclerosis exhibit reduced T-cell response against Omicron compared with the ancestral strain.^18–20 However, in other studies, a comparable response rate was detected for both the ancestral strain and Omicron variant, potentially due to differing timelines for collection of peripheral blood mononuclear cells (PBMCs) and the assay parameters.^21–23

Patients with cancer are at a higher risk of developing severe disease and mortality following SARS-CoV-2 infection, which can be explained by multiple factors including advanced age; comorbidities; and active treatments that impact the immune system, such as corticosteroids, anti-B-cell therapy, and immune checkpoint blockade.^24–28 Considering the increased risk of severe COVID-19 in patients with cancer, research has focused on the humoral and cellular responses to SARS-CoV-2 variants (ancestral, beta, and delta strains) in the context of vaccination status or prior infection.²⁹ These studies found a modest reduction in both immune compartments for patients with cancer against the ancestral strain and earlier variants.^{25 30–33} Similarly, for the Omicron variant, recent publications indicated that the viral neutralizing capacity established by vaccination is impaired in individuals who received the two-dose vaccination course (both healthy individuals and patients with cancer), but this phenotype was partially rescued by a third booster dose.^34–36 Patients with multiple myeloma also showed decreased B and T-cell immune responses against Omicron variant.³⁷ As the balance between humoral and cellular responses to infection is critical for disease control, understanding the cellular response to infection in patients with cancer is of paramount importance. Vaccination against ancestral strain elicits considerable CD8⁺ and CD4⁺ T-cell response, but the effect of the booster vaccine in patients with cancer has not been thoroughly addressed.^18–20 It is also unclear whether CD8⁺ and CD4⁺ T-cell responses against the Omicron variant are qualitatively impaired in patients with lung cancer after booster mRNA vaccines.

Methods

Study design and enrollment

All study participants were enrolled under an approved Institutional Review Board protocol (2021C004) at The Ohio State University Wexner Medical Center. Eligible participants were at least 18 years old, previously received two doses of mRNA COVID-19 vaccine, and able to complete online surveys related to the study. Participants included cancer and non-cancer cohorts, with data presented herein focusing on patients with lung cancer. PBMCs were isolated after mRNA vaccine course (designated “pre-booster”) and booster (designated “post-booster”) in non-cancer (NC) and lung patient cohorts. This yielded mixed paired and unpaired sets of pre-booster and post-booster samples for each cohort. Demographic and clinical characteristics (including age, sex, race, cancer staging, and therapy status) were obtained from internal electronic medical records (EMR) and patient-provided information via survey. Relevant cancer therapies included those administered within 9 months of primary vaccination and subcategorized into chemotherapy (primarily platinum agents), targeted therapy, immune checkpoint inhibitor (ICI; eg, anti-programmed cell death protein 1 (PD-1)/programmed death-ligand 1 (PD-L1)), and combination treatment. Additional COVID-19-specific details (including vaccine and booster type, sample time points, history of diagnosis or treatment of COVID-19) were also abstracted from EMR (online supplemental table 1). Two independent (preliminary and validation) experiments were combined to determine the effect of the booster vaccine in activation-induced marker (AIM)⁺ T-cell frequencies. Validation cohort was used for high-dimensional flow cytometry analysis and paired analysis comparing frequencies of AIM⁺ cells stimulated with ancestral and Omicron peptides. 3 (non-cancer) and 15 (lung-cancer) samples were matched for pre-booster and post-booster vaccination samples when determining AIM⁺ T-cell frequencies.

Cryopreservation of PBMCs

Blood samples were collected in tubes containing acid citrate dextrose (BD). Samples were diluted with an equal volume of Dulbecco's Phosphate-Buffered Saline(DPBS) (Gibco) and poured into SepMate tubes (STEMCELL) preloaded with FICOLL (BD) solution. SepMate tubes were spun, and the layer enriched with PBMCs was isolated. Isolated PBMCs were washed with DPBS twice. Cells were counted and resuspended in freezing media. Each vial was placed in a freezing container and transferred later to a liquid nitrogen tank per the manufacturer’s protocol.

Activation-induced marker assay

Activation-induced marker (AIM) assay was performed according to published protocols with an adapted flow cytometry panel (online supplemental table 2) to capture the characteristics of antigen-specific T cells.³⁸ Briefly, cryopreserved PBMCs were plated on U-bottom 96-well plates at 1 million cells per well. To obtain enough cells for downstream analysis in ancestral strain spike peptide and Omicron variant spike peptide stimulated conditions, two wells per condition were plated for downstream analysis. For the no-peptide and CEF (cytomegalovirus, Epstein-Barr virus, and influenza virus) groups, only a single well was plated under the same conditions. Cells were incubated overnight at 37°C, 5% CO2, and subsequently treated with a CD40 blocking antibody (Miltenyi, 0.5 µg/mL final concentration) for 15 min and stimulated with peptide pools (ancestral strain spike and Omicron variant spike, GenScript; CEF peptide pool, STEMCELL) at 2 µg/mL in the presence of a costimulatory reagent (CD28/CD49d, BD) for 24 hours. At 20 hours post-stimulation, 37°C staining cocktails (online supplemental table 2) were added along with GolgiStop (BD) and incubated for 4 hours. Cells were collected, pooled as described, and washed with DPBS. Viability dye staining (eBioscience) was applied for 15 min at room temperature. Cells were washed and incubated at room temperature with staining antibody mixture (online supplemental table 2) for 45 min. Fc receptor blocking reagent (BioLegend) was added to both viability and room temperature staining antibody cocktails. Cells were washed and incubated with fix/permeabilization solution (eBioscience) for 30 min, followed by permeabilization solution and intracellular staining antibodies (online supplemental table 2) overnight at 4°C. Finally, cells were washed with flow cytometry buffer for downstream analysis.

Flow cytometry and data analysis

Flow cytometry data was acquired on a 5-laser Cytek Aurora instrument. UltraComp eBeads (eBioscience) were used to set single-color controls to perform unmixing prior to sample requisition. To check AIM⁺ cell frequencies from acquired data, FlowJo software (BD) was used for quantification and visualization. High-dimensional flow cytometry data analysis was performed using the web-based software OMIQ. Initially, AIM⁺ cells were gated from each sample and concatenated per group for further analysis. Uniform Manifold Approximation and Projection (UMAP) visualization was applied, and the FlowSOM clustering technique was used to distinguish cells with unique features.^{39 40} To determine the optimal number of clusters, elbow metaclustering was performed to decide the proper number of clusters, and consensus metaclustering was used for data analysis. Consensus metaclustering was performed with 10 xdim and 10 ydim with Euclidean distance metric. Clustered heatmap was generated using normalized expression values with indicated markers. Normalization was done considering the entire heatmap. For high-dimensional flow cytometry analysis, a few samples were excluded due to insufficient AIM⁺ cells.

Statistics

For each stimulation type (no-peptide, CEF, ancestral, and Omicron), we evaluated changes in cell proportions (CD8⁺ or CD4⁺ T cells) on a log scale between pre-booster and post-booster vaccination for lung cancer versus NC cohorts. For statistical evaluation, we used a linear mixed-effects model, including time point (pre vs post), sex, and age as fixed effects and a subject-specific random effect (the lme4 R package). To address statistical significance in the proportion of clusters comparing NC and lung cancer samples, non-paired Student’s t-test was applied. A comparison of AIM⁺ T-cell frequencies against ancestral and Omicron variants with pooled samples was performed using a paired t-test.

Results

We performed in-depth studies of T-cell responses against the spike protein antigen in patients with cancer after SARS-CoV-2 mRNA vaccinations. A total of 63 samples collected under the COVID-19 Vaccine Study of Infections and Immune Response trial were assessed, and demographic and clinical characteristics were summarized (online supplemental table S1). Patients with lung cancer were primarily of advanced clinical staging (stages 3–4), with 93.3% of pre-booster and 100% of post-booster cohorts at stage 3 or greater. All patients were receiving active therapy primarily with ICIs. Only two subjects in each cohort had a history of SARS-CoV-2 infection, including one who received treatment for acute infection (via monoclonal antibody therapy). We performed modified AIM assays using cryopreserved PBMC samples, coupled with high-dimensional flow cytometry analysis of both CD8⁺ and CD4⁺ T cells. Two independent experiments (preliminary and validation) were combined to determine and quantify the AIM⁺ T-cell frequencies, and the validation cohort result was used for high-dimensional flow cytometry analysis.

To capture antigen-specific CD8⁺ T-cell frequencies, PBMCs were stimulated with costimulatory signals (anti-CD28/CD49d) in the presence of CEF peptide pool (containing 32 peptides from the cytomegalovirus, the Epstein-Barr virus, and the influenza virus), overlapping 15-mer peptide pools from ancestral strain or Omicron spike proteins, or no peptide control. High-dimensional flow cytometry assessed two parameters. We evaluated the frequency of activated T cells, to reflect the antigen-experienced and antigen-specific T cells. Additionally, we evaluated subpopulation dynamics of T cells based on differentiation markers, including immune checkpoint molecules (eg, PD-1, cytotoxic T-lymphocyte associated protein 4(CTLA-4), CD39 and lymphocyte-activation gene 3 (LAG-3)), transcription factors (eg, T cell factor 1 (TCF1)), and effector molecules (eg, interferon (IFN)-γ and Granzyme B (GZMB); online supplemental table S2). As previously established,^{9 22} we identified AIM⁺ cells by CD69 and 4-1BB double positivity.

After 24-hour stimulation, the cells treated with CEF peptide pool, ancestral strain spike peptide, and Omicron variant spike peptide showed a higher frequency of AIM⁺ cells compared with no-peptide control (figure 1A). We quantified AIM⁺ cells based on stimulating peptide, cancer status, and vaccination condition (figure 1B). Using a linear mixed model with log-transformed response values, we evaluated statistical differences, considering matching structure of samples and multiple other variables (sex, age, and cancer stage). By using log-transformed values, analyses were performed in a more conservative manner, minimizing the effect of potential outliers. As expected, the frequency of CEF-specific T cells did not change with SARS-CoV-2 mRNA booster vaccines. By comparison, and in line with previous reports, the third-dose booster vaccine increased CD8⁺ T-cell response in non-cancer subjects against both the ancestral strain and Omicron variant.^{11 41} By the third booster dose, patients with lung cancer also showed increased frequency of AIM⁺ CD8⁺ T cells against both spike peptides, but the induction rate against the Omicron variant (2.2-fold induction) was not as high as that detected against the ancestral strain (3.2-fold induction). All groups showed considerable antigen-specific CD8⁺ T-cell response when compared with the no-peptide stimulation group, implying that antigen-specific CD8⁺ T-cell response induced by vaccination was intact in this cohort. There was a significantly reduced CD8⁺ T-cell response against Omicron versus the ancestral strain in both cancer and non-cancer cohorts, similar to previous reports (online supplemental figure 1A).^{18 19} Next, we compared AIM⁺ CD8⁺ T cells in male versus female samples, as sex plays a role in vaccine efficacy.⁴² We detected higher frequency of AIM⁺ CD8⁺ T cell in female samples when compared with male samples (online supplemental figure 1B).⁴³

Figure 1. CD8 + T-cell responses in non-cancer donors and patients with lung cancer following viral peptide stimulation. (A) Representative flow cytometry gating strategy for AIM + CD8 + T cells. CEF peptide pool, ancestral spike protein, and Omicron spike protein were used to stimulate pre-booster and post-booster peripheral blood mononuclear cell samples; no-peptide group served as a negative control. (B) Percentage of AIM + CD8 + T cells were plotted to compare frequencies between the groups. NC; (non-cancer), pre; (pre-booster vaccination), post; (post-booster vaccination). Gray lines connecting pre-vaccine and post-vaccine dots indicate matched samples. Results from two independent sets of experiments were combined to determine AIM + CD8 + T-cell frequencies. Data is presented as mean+-SEM. Statistical significance evaluated using linear mixed models. **p<0.01, ***p<0.001. AIM, activation-induced marker; CEF, cytomegalovirus, Epstein-Barr virus, and influenza virus.

Vaccination-induced SARS-CoV-2 antigen-specific T cells can exhibit heterogenous population dynamics, reflecting diverse differentiation states.³⁸ Also, non-naïve T cells from COVID-19-infected patients demonstrated various activation and differentiation states, requiring a more in-depth understanding of antigen-specific T cells.⁴⁴ To address this, we used high-dimensional flow cytometry data from the validation cohort to determine the differentiation status of antigen-specific CD8⁺ T cells. Using OMIQ, AIM⁺ cells (CD69⁺ and 4-1BB⁺) were gated and concatenated for further analysis. We used UMAP dimension reduction for visualization and applied FlowSOM for clustering analysis on gated cells. We identified 18 clusters with distinct marker expression patterns, which we projected into UMAP space (figure 2A). Marker expression patterns were assessed with a clustered heatmap to evaluate characteristics (figure 2B). Clusters 15 and 16 exhibited higher PD-1, CTLA-4, and LAG-3 expression without CD39 expression, whereas clusters 4 and 5 demonstrated intermediate CD39 expression but low PD-1. Clusters 2 and 3 showed higher OX40 and CD40L expression with low-to-intermediate levels of inducible T cell costimulator (ICOS). Cluster 12 expressed the highest levels of PD-1 and LAG-3 with a higher CD39 expression level. Cluster 1 showed the least activation/dysfunction marker expression within AIM⁺ CD8⁺ T cells. To better understand the population dynamics between the groups, we used a contour plot and visualized in UMAP space (figure 2C). Booster vaccines did not cause noticeable changes in population dynamics in AIM⁺ CD8⁺ T cells. However, when we compared patient with lung cancer and non-cancer donor samples, we noticed enriched accumulation of a PD-1⁺ CTLA-4⁺ LAG-3⁺ population in patient with lung cancer samples, represented by clusters 16 and 12 (figure 2C, left top arrow; quantified in figure 2D and E). Cluster 1 with the less-activated phenotype (figure 2C, lower right arrow) did not significantly change, but patients with lung cancer tended to develop fewer cells in this cluster (figure 2F). AIM⁺ CD8⁺ T cells stimulated with CEF peptide pool showed no significant differences in population dynamics between non-cancer and lung cancer groups (online supplemental figure 2). To evaluate the impact of the sex and immunotherapy responsiveness in AIM⁺ CD8⁺ T-cell dynamics, we performed subgroup analysis using patient with lung cancer samples. A similar analytical approach was applied, and we visualized 14 distinct clusters in UMAP space (online supplemental figure 3A). Key marker expressions were overlaid into UMAP space to visualize marker expression patterns (online supplemental figure 3B). Interestingly, male patients had higher frequencies of clusters 3 and 5, which showed higher expressions of inhibitory molecules (PD-1, LAG-3, CD39, and CTLA-4). Also, cluster 13 (CD45RO⁺ GZMB⁻) was increased in male samples (online supplemental figure 3C,D). To further explore CD8⁺ T-cell dynamics, we compared AIM⁺ CD8⁺ T cells between immunotherapy responders and non-responders. Cluster 4 (PD-1^– CD45RA⁺ GZMB⁺) which showed phenotype similar to terminally differentiated effector memory cells (Temra) was increased, and cluster 10 (PD-1^– CD45RA⁺ GZMB^int) and cluster 11 (CD45RO⁺ CD62L⁺) which represented less activated cells were decreased in immunotherapy responder samples (online supplemental figure 3E,F). Our data suggests that AIM⁺ CD8⁺ T cells comprise a heterogenous population in various states of differentiation, and patients with lung cancer are more likely to generate antigen-specific T-cell responses that express immune checkpoint molecules.

Figure 2. Population dynamics of activation-induced marker + CD8 + T cells from non-cancer donor and patient with lung cancer samples. (A) UMAP dimension reduction and FlowSOM clustering were performed to visualize and define clusters in UMAP space. 18 clusters with distinct marker expression patterns were identified; circles indicate populations of interest. (B) Heatmap indicating relative marker expression by cluster. (C) Contour plot was used to visualize population dynamics across cohorts. Upper row, samples stimulated with ancestral strain spike protein; lower row, samples stimulated with Omicron variant spike protein. Arrows indicate notable changes between patient without cancer and patient with lung cancer samples. The numbers of cells concatenated per condition indicated at upper right corner. (D-F) Pre-vaccination and post-vaccination samples were pooled per cancer status; proportions of clusters 16, 12, and 1 were plotted. Data are presented as the mean+-SEM. Statistical significances (D-F) were evaluated using an independent-sample t-test. *p<0.05. CTLA-4, cytotoxic T-lymphocyte associated protein 4; GZMB, granzyme B; IFN, interferon; LAG-3, lymphocyte-activation gene 3; PD-1, programmed cell death protein 1; UMAP, Uniform Manifold Approximation and Projection.

We next examined the vaccine-induced AIM⁺ CD4⁺ T-cell frequencies based on the co-expression of CD200 and CD40L (figure 3A). The CEF peptides represent major histocompatibility complex (MHC) class I epitopes and indeed showed minimal stimulation of CD4⁺ T cells. We observed a significantly increased frequency of AIM⁺ cells by third-dose booster vaccine in all groups tested (figure 3B). However, the induction level of AIM⁺ CD4⁺ T cells by third-dose booster vaccine was lower when compared with CD8⁺ T-cell response. We detected 1.4-fold and 1.3-fold induction in non-cancer individuals against ancestral and Omicron strain, respectively, and observed 1.3-fold induction in patients with lung cancer for both strains. These findings were consistent with reports that the CD4⁺ T-cell response is less durable than CD8⁺ T-cell response following mRNA vaccine, and that CD8⁺ T cells—but not CD4⁺ T cells—are preserved even with B-cell depletion therapy.^{20 45} However, the 15-mer peptide pool we used may not be optimal for CD4⁺ T-cell stimulation, so further validation is required. Next, we pooled pre-booster and post-booster samples and compared AIM⁺ CD4⁺ T-cell frequencies in non-cancer donor and patient with lung cancer samples against the ancestral strain and the Omicron variant. Similar to our results evaluating the CD8⁺ T-cell response, non-cancer samples showed reduced AIM⁺ CD4⁺ T-cell frequencies against Omicron when compared with the ancestral strain, but patient with lung cancer samples did not follow this trend (online supplemental figure 4A). Also, female patient with cancer samples displayed greater AIM⁺ CD4⁺ T-cell frequencies than males (online supplemental figure 4B).

Figure 3. CD4 + T-cell responses in non-cancer donors and patients with cancer following peptide stimulation. (A) Representative flow cytometry gating for AIM + CD4 + T cells. CEF peptide pool, ancestral spike protein, and Omicron spike protein were used to stimulate pre-booster and post-booster peripheral blood mononuclear cell samples; no-peptide group served as a negative control. (B) Percentages of AIM + cells were plotted to compare frequencies between the groups. Gray lines connecting pre-vaccine and post-vaccine dots indicate matched samples. Results from two independent sets of experiments were combined to determine AIM + CD4 + T-cell frequencies. Data is presented as the mean+-SEM. Statistical significances evaluated using linear mixed model. *p<0.05, **p<0.01. AIM, activation-induced marker; CEF, cytomegalovirus, Epstein-Barr virus, and influenza virus; NC, non-cancer;

To examine the impact of cancer status, booster, and sex on CD4⁺ T-cell differentiation program, we gated CD200 and CD40L double-positive AIM⁺ CD4⁺ T cells and performed OMIQ analysis, similar to the CD8⁺ T cells. We identified 15 CD4⁺ T-cell subclusters with distinct features, which were visualized in UMAP space (figure 4A). Marker expression was characterized per cluster and visualized using a clustered heatmap (figure 4B). Cluster 7 represented effector cells, which have the highest IFN-γ expression with intermediate-to-high PD-1 and LAG-3 expression; cluster 3 exhibited similar expression patterns but did not express IFN-γ. Cluster 4 expressed intermediate levels of CD45RA and CD62L. Cluster 14 had characteristics of regulatory T cells, including the highest expression of FOXP3, CTLA-4, and CD39. To evaluate population dynamics, we visualized the data using a contour plot (figure 4C). We found that cluster 3 increased while cluster 4 decreased in patient with lung cancer samples, compared with non-cancer samples (figure 4D and E). Cluster 9, which did not show distinct characteristics, was not affected by cancer status (figure 4F). We also evaluated the impact of sex and immunotherapy responsiveness in CD4⁺ AIM⁺ T-cell dynamics using patient with lung cancer samples. Similar to CD8⁺ T-cell analysis, we visualized 14 distinct clusters in UMAP space with marker expression patterns (online supplemental figure 5A,B). Clusters 5 and 8 which express higher levels of inhibitory molecules (PD-1, LAG-3, and CTLA-4), were increased in male samples (online supplemental figure 5C,D). When we compared immunotherapy responders and non-responders, responders showed decreased frequencies of cluster 6 (PD-1⁺ CD45RO⁺ CD27^–) and cluster 11 (PD-1^– CD45RA⁺), while cluster 13 (PD-1⁺ CD45RO⁺ CXCR3⁺ CD27⁺) was increased (online supplemental figure 5E,F). Thus, we conclude that the AIM⁺ T cells can harbor various differentiation patterns that are impacted by multiple factors including cancer status, sex, and immunotherapy responsiveness.

Figure 4. Population dynamics of AIM + CD4 + T cells from non-cancer donor and patient with lung cancer samples. (A) UMAP dimension reduction and FlowSOM clustering were performed to visualize and define clusters in UMAP space. 15 clusters with distinct marker expression patterns were identified; circles indicate populations of interest. (B) Heatmap indicating relative marker expression by cluster. (C) Contour plot was generated to visualize population dynamics across cohorts. Upper row, samples stimulated with ancestral strain spike protein; lower row, samples stimulated with Omicron variant spike protein. Arrows indicate notable changes between non-cancer and lung cancer samples. The numbers of cells concatenated per condition indicated at bottom right corner. (D-F) Pre-vaccination and post-vaccination samples were pooled per cancer status and proportions of clusters 3, 4, and 9 were plotted. Data are presented as the mean+-SEM. Statistical significances (D-F) were evaluated using an independent-sample t-test. **p<0.01, ***p<0.001. CTLA-4, cytotoxic T-lymphocyte associated protein 4; GZMB, granzyme B; IFN, interferon; LAG-3, lymphocyte-activation gene 3; PD-1, programmed cell death protein 1; UMAP, Uniform Manifold Approximation and Projection.

Discussion

In this study, we evaluated antigen-specific T-cell response against ancestral and Omicron strains of SARS-CoV-2 following mRNA-vaccine in patient without and with lung cancer samples. When evaluating the CD8⁺ T-cell response, we found that booster shots increased AIM⁺ frequencies in both patient cohorts against the ancestral strain and Omicron variant, but Omicron-specific T cells were induced to a lower level in patients with lung cancer, as previously reported.^18–20 When evaluating the CD4⁺ T-cell response, the booster vaccine increased the AIM⁺ frequency in all groups tested, but the levels induced were marginal when compared with CD8⁺ T-cell response. Although it is possible that the 15-mer peptide pool was suboptimal for CD4⁺ T-cell activation, 15-mer peptide has been widely used for studying CD4⁺ T-cell response, and we also observed consistent CD4⁺ T-cell activation using this peptide pool. We consider our assay reliable with appropriate positive and negative controls and showed considerable CD4⁺ and CD8⁺ T-cell response against the spike protein antigen in vaccinated individuals when compared with no-peptide controls. The dampened T-cell response to Omicron in patients with cancer could suggest that vaccination established less protection for patients with lung cancer, especially given that humoral response is also reduced in this population.²⁹

Growing evidence shows that sex acts as a critical factor in shaping immune responses against viral infection and cancer,^46–49 which warrants more systemic studies.^{50 51} Previous reports show vaccination provides greater protection against SARS-CoV-2 infection in women than men.⁵² Our data supports this, as we observed a more efficient antigen-specific T-cell responses from female patients than males. We also characterized the phenotype of male and female AIM⁺ T cells and observed a qualitative deviation. Further studies are required to address the mechanism underlying our findings, including the emerging roles of T cell-intrinsic androgen receptors in driving T-cell dysfunction or exhaustion.^{48 49 53}

mRNA booster vaccines are recommended for individuals with impaired immune systems to augment humoral and cellular immunity against SARS-CoV-2 mutant strains, such as Omicron.⁵⁴ However, our findings indicate that current mRNA vaccine booster shots may offer an inadequate level of protection against the Omicron strain due to suboptimal T-cell responses in patients with cancer, specifically in lung malignancies. This is supported by the fewer number of antigen-specific T cells observed in patients with lung cancer against the Omicron variant spike peptides, as well as the appearance of dysfunctional T-cell phenotypes from high-dimensional flow cytometry analysis. Our finding may explain the disproportionately high number of hospitalizations and breakthrough Omicron cases reported in fully vaccinated/boosted patients with cancer, especially in those with advanced cancer receiving active therapy for their underlying malignancies.^{55 56}

One key message from our study is that the total or relative numbers of antigen-specific T-cell responses alone does not give a complete picture of immune response following vaccine or infection. High-dimensional flow analysis revealed a high degree of AIM⁺ T-cell heterogeneity in vaccinated individuals. While the underlying reason for this heterogeneity remains unclear, it could be due to the polyclonal nature of these cells, as well as their differentiation states. Clonal variation may be due to differences in the affinity of T-cell receptors for MHC-peptide complexes, as well as in the avidity of antigen-specific T-cell interactions with the cognate antigen-presenting cells. When we assessed the differentiation states of both CD8⁺ and CD4⁺ T cells, we saw a significant deviation of patients with cancer from non-cancer subjects. AIM⁺ CD8⁺ T cells from patients with lung cancer showed more activated/dysfunctional signatures with higher PD-1 expression, despite the fact that patients with lung cancer enrolled in the trial were receiving immunotherapy with anti-PD-1/PD-L1 agents. A similar trend was observed in AIM⁺ CD4⁺ T cells. The patient with lung cancer cohort contained more PD-1-positive cells without IFN-γ expression. Although patients with lung cancer showed higher expression of PD-1 and other inhibitory molecule expressions in certain clusters, these markers can also represent activated phenotypes. Due to the low number of AIM⁺ cells, we were not able to perform functional assay to determine which population can more effectively kill the virus-infected cells; however, this is an important point to address in the future. These population dynamics could also explain the discrepancies in T-cell response against Omicron reported to date by various laboratories, as each researcher assesses T-cell response using different markers and methodologies. Finally, an obvious limitation of our study is that we do not have a separate cohort of patients with cancer who did not receive immunotherapy. Thus, the possibility of immunotherapy drugs affecting T-cell differentiation heterogeneity in patients with cancer cannot be ruled out and requires further evaluation.

In summary, patients with advanced-stage lung cancer undergoing active therapy had preserved T-cell response against the ancestral strain but compromised response to the Omicron variant following SARS-CoV-2 mRNA vaccination using current technology. Moreover, these patients with cancer have distinct differentiation statuses from non-cancer subjects in both CD8⁺ and CD4⁺ T-cell compartments post-booster shot. Further investigations are needed to understand the underlying mechanism of our findings and the implications of our work to ensure protection of patients with cancer (and other vulnerable populations) from breakthrough infection by SARS-CoV-2 Omicron variant and other future variants.

We thank the Wherry laboratory, including Mathew Divij, for providing the detailed protocol for AIM assay and discussion. Additional gratitude is extended to those who participated in sample collection and preparation, including Jamie Hamon, Donna Bucci, Mohamed Yusuf, Robert Davenport, and Taylor Chatlos from the Pelotonia Institute for Immuno-Oncology and the Immune Monitoring and Discovery Platform. We thank Angela Dahlberg for editing and proofreading this paper.

Data availability statement

The data sets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

All study participants were enrolled under an approved Institutional Review Board protocol (2021C004) at The Ohio State University Wexner Medical Center. Participants gave informed consent to participate in the study before taking part.

Footnote

Contributors Conceptualization: N-JS, KBC, MPR, PGS, ZL. Methodology: N-JS, KBC, KR, YW, AL, KPW, HJ, DC. Investigation: N-JS, MPR, ZL. Visualization: N-JS, KBC, CB. Funding acquisition: ZL, PGS. Project administration: CB, KPW, PGS, SR, ZL. Supervision: PGS, ZL. Writing—original draft: N-JS, KBC, CB, ZL. Writing—review and editing: N-JS, KBC, CB, ZL, SJ, QM, DHB, EMO. Guarantor: ZL.

Funding This work is supported by the OSU Cancer Center Support Grant (P30-CA016058) and administrative supplements to study the impact of SARS-CoV-2 vaccine boosts in immunocompromised populations (Center for Serological Testing to Improve Outcomes from Pandemic COVID-19) (U54-CA260582).

Competing interests No, there are no competing interests.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ World Health Organization (WHO). WHO Coronavirus (COVID-19) Dashboard: WHO, Available: https://covid19.who.int/ [Accessed 29 Mar 2023 ].

² Hu B, Guo H, Zhou P, et al. Characteristics of SARS-Cov-2 and COVID-19. Nat Rev Microbiol 2021; 19: 141–54. doi:10.1038/s41579-020-00459-7

³ Sanders JM, Monogue ML, Jodlowski TZ, et al. Pharmacologic treatments for Coronavirus disease 2019 (COVID-19): A review. JAMA 2020; 323: 1824–36. doi:10.1001/jama.2020.6019

⁴ Owen DR, Allerton CMN, Anderson AS, et al. An oral SARS-Cov-2 M(Pro) inhibitor clinical candidate for the treatment of COVID-19. Science 2021; 374: 1586–93. doi:10.1126/science.abl4784

⁵ Jayk Bernal A, Gomes da Silva MM, Musungaie DB, et al. Molnupiravir for oral treatment of COVID-19 in Nonhospitalized patients. N Engl J Med 2022; 386: 509–20. doi:10.1056/NEJMoa2116044

⁶ Song N-J, Allen C, Vilgelm AE, et al. Treatment with soluble Cd24 attenuates COVID-19-associated systemic Immunopathology. J Hematol Oncol 2022; 15: 5. doi:10.1186/s13045-021-01222-y

⁷ Falsey AR, Frenck RW, Walsh EE, et al. SARS-Cov-2 neutralization with BNT162b2 vaccine dose 3. N Engl J Med 2021; 385: 1627–9. doi:10.1056/NEJMc2113468

⁸ Gruell H, Vanshylla K, Tober-Lau P, et al. mRNA booster immunization elicits potent neutralizing serum activity against the SARS-Cov-2 Omicron variant. Nat Med 2022; 28: 477–80. doi:10.1038/s41591-021-01676-0

⁹ Goel RR, Painter MM, Apostolidis SA, et al. mRNA vaccines induce durable immune memory to SARS-Cov-2 and variants of concern. Science 2021; 374: abm0829. doi:10.1126/science.abm0829

¹⁰ Bar-On YM, Goldberg Y, Mandel M, et al. Protection of BNT162b2 vaccine booster against COVID-19 in Israel. N Engl J Med 2021; 385: 1393–400. doi:10.1056/NEJMoa2114255

¹¹ Liu Y, Zeng Q, Deng C, et al. Robust induction of B cell and T cell responses by a third dose of Inactivated SARS-Cov-2 vaccine. Cell Discov 2022; 8: 10. doi:10.1038/s41421-022-00373-7

¹² Liu J, Yu J, McMahan K, et al. Cd8 T cells contribute to vaccine protection against SARS-Cov-2 in macaques. Sci Immunol 2022; 7: eabq7647. doi:10.1126/sciimmunol.abq7647

¹³ Karim SSA, Karim QA. Omicron SARS-Cov-2 variant: a new chapter in the COVID-19 pandemic. The Lancet 2021; 398: 2126–8. doi:10.1016/S0140-6736(21)02758-6

¹⁴ Hu J, Peng P, Cao X, et al. Increased immune escape of the new SARS-Cov-2 variant of concern Omicron. Cell Mol Immunol 2022; 19: 293–5. doi:10.1038/s41423-021-00836-z

¹⁵ Kuhlmann C, Mayer CK, Claassen M, et al. Breakthrough infections with SARS-Cov-2 Omicron despite mRNA vaccine booster dose. Lancet 2022; 399: 625–6. doi:10.1016/S0140-6736(22)00090-3

¹⁶ McCallum M, Czudnochowski N, Rosen LE, et al. Structural basis of SARS-Cov-2 Omicron immune evasion and receptor engagement. Science 2022; 375: 864–8. doi:10.1126/science.abn8652

¹⁷ Cao Y, Wang J, Jian F, et al. Omicron escapes the majority of existing SARS-Cov-2 neutralizing antibodies. Nature 2022; 602: 657–63. doi:10.1038/s41586-021-04385-3

¹⁸ Naranbhai V, Nathan A, Kaseke C, et al. T cell reactivity to the SARS-Cov-2 Omicron variant is preserved in most but not all individuals. Cell 2022; 185: S0092-8674(22)00332-4. doi:10.1016/j.cell.2022.03.022

¹⁹ Keeton R, Tincho MB, Ngomti A, et al. Author correction: T cell responses to SARS-Cov-2 spike cross-recognize Omicron. Nature 2022; 604: E25. doi:10.1038/s41586-022-04708-y

²⁰ Madelon N, Heikkilä N, Sabater Royo I, et al. Omicron-specific cytotoxic T-cell responses after a third dose of mRNA COVID-19 vaccine among patients with multiple sclerosis treated with Ocrelizumab. JAMA Neurol 2022; 79: 399–404. doi:10.1001/jamaneurol.2022.0245

²¹ GeurtsvanKessel CH, Geers D, Schmitz KS, et al. Divergent SARS-Cov-2 Omicron-reactive T and B cell responses in COVID-19 vaccine recipients. Sci Immunol 2022; 7: eabo2202. doi:10.1126/sciimmunol.abo2202

²² Tarke A, Coelho CH, Zhang Z, et al. SARS-Cov-2 vaccination induces immunological T cell memory able to cross-recognize variants from alpha to Omicron. Cell 2022; 185: 847–59. doi:10.1016/j.cell.2022.01.015

²³ Liu J, Chandrashekar A, Sellers D, et al. Vaccines elicit highly conserved cellular immunity to SARS-Cov-2 Omicron. Nature 2022; 603: 493–6. doi:10.1038/s41586-022-04465-y

²⁴ Russell B, Moss CL, Shah V, et al. Risk of COVID-19 death in cancer patients: an analysis from guy’s cancer centre and king’s college hospital in London. Br J Cancer 2021; 125: 939–47. doi:10.1038/s41416-021-01500-z

²⁵ Fendler A, Au L, Shepherd STC, et al. Functional antibody and T cell immunity following SARS-Cov-2 infection, including by variants of concern, in patients with cancer: the CAPTURE study. Nat Cancer 2021; 2: 1321–37. doi:10.1038/s43018-021-00275-9

²⁶ Johannesen TB, Smeland S, Aaserud S, et al. COVID-19 in cancer patients, risk factors for disease and adverse outcome, a population-based study from Norway. Front Oncol 2021; 11: 652535. doi:10.3389/fonc.2021.652535

²⁷ Yang L, Chai P, Yu J, et al. Effects of cancer on patients with COVID-19: a systematic review and meta-analysis of 63,019 participants. Cancer Biol Med 2021; 18: 298–307. doi:10.20892/j.issn.2095-3941.2020.0559

²⁸ Ruiz JI, Lopez-Olivo MA, Geng Y, et al. COVID-19 vaccination in patients with cancer receiving immune Checkpoint inhibitors: a systematic review and meta-analysis. J Immunother Cancer 2023; 11: e006246. doi:10.1136/jitc-2022-006246

²⁹ Fendler A, de Vries EGE, GeurtsvanKessel CH, et al. COVID-19 vaccines in patients with cancer: Immunogenicity, efficacy and safety. Nat Rev Clin Oncol 2022; 19: 385–401. doi:10.1038/s41571-022-00610-8

³⁰ Fendler A, Shepherd STC, Au L, et al. Adaptive immunity and neutralizing antibodies against SARS-Cov-2 variants of concern following vaccination in patients with cancer: the CAPTURE study. Nat Cancer 2021; 2: 1321–37. doi:10.1038/s43018-021-00274-w

³¹ Zeng C, Evans JP, Reisinger S, et al. Impaired neutralizing antibody response to COVID-19 mRNA vaccines in cancer patients. Cell Biosci 2021; 11: 197. doi:10.1186/s13578-021-00713-2

³² Terpos E, Gavriatopoulou M, Fotiou D, et al. Poor neutralizing antibody responses in 132 patients with CLL, NHL and HL after vaccination against SARS-Cov-2: A prospective study. Cancers (Basel) 2021; 13: 4480. doi:10.3390/cancers13174480

³³ Ehmsen S, Asmussen A, Jeppesen SS, et al. Antibody and T cell immune responses following mRNA COVID-19 vaccination in patients with cancer. Cancer Cell 2021; 39: 1034–6. doi:10.1016/j.ccell.2021.07.016

³⁴ Rössler A, Riepler L, Bante D, et al. SARS-Cov-2 Omicron variant neutralization in serum from vaccinated and Convalescent persons. N Engl J Med 2022; 386: 698–700. doi:10.1056/NEJMc2119236

³⁵ Schmidt F, Muecksch F, Weisblum Y, et al. Plasma neutralization of the SARS-Cov-2 Omicron variant. N Engl J Med 2022; 386: 599–601. doi:10.1056/NEJMc2119641

³⁶ Zeng C, Evans JP, Chakravarthy K, et al. COVID-19 mRNA booster vaccines elicit strong protection against SARS-Cov-2 Omicron variant in patients with cancer. Cancer Cell 2022; 40: 117–9. doi:10.1016/j.ccell.2021.12.014

³⁷ Azeem MI, Nooka AK, Shanmugasundaram U, et al. Impaired SARS-Cov-2 variant neutralization and Cd8+ T-cell responses following 3 doses of mRNA vaccines in myeloma: correlation with breakthrough infections. Blood Cancer Discov 2023; 4: 106–17. doi:10.1158/2643-3230.BCD-22-0173

³⁸ Apostolidis SA, Kakara M, Painter MM, et al. Cellular and humoral immune responses following SARS-Cov-2 mRNA vaccination in patients with multiple sclerosis on anti-Cd20 therapy. Nat Med 2021; 27: 1990–2001. doi:10.1038/s41591-021-01507-2

³⁹ McInnes L, Healy J, Melville J. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction,. 2018 Available: https://ui.adsabs.harvard.edu/abs/2018arXiv180203426M

⁴⁰ Van Gassen S, Callebaut B, Van Helden MJ, et al. Flowsom: using self-organizing maps for visualization and interpretation of Cytometry data. Cytometry A 2015; 87: 636–45. doi:10.1002/cyto.a.22625

⁴¹ Nguyen AT, Szeto C, Chatzileontiadou DSM, et al. n.d. COVID-19 vaccine booster induces a strong Cd8+ T cell response against Omicron variant epitopes in HLA-A*02:01+ Individuals.. bioRxiv; 2022: 2022. doi:10.1101/2022.01.12.473243

⁴² Fink AL, Klein SL. Sex and gender impact immune responses to vaccines among the elderly. Physiology (Bethesda) 2015; 30: 408–16. doi:10.1152/physiol.00035.2015

⁴³ Karlsson AC, Humbert M, Buggert M. The known unknowns of T cell immunity to COVID-19. Sci Immunol 2020; 5: eabe8063. doi:10.1126/sciimmunol.abe8063

⁴⁴ Mathew D, Giles JR, Baxter AE, et al. Deep immune profiling of COVID-19 patients reveals distinct Immunotypes with therapeutic implications. Science 2020; 369: 6508. doi:10.1126/science.abc8511

⁴⁵ Cox RJ, Brokstad KA. Not just antibodies: B cells and T cells mediate immunity to COVID-19. Nat Rev Immunol 2020; 20: 581–2. doi:10.1038/s41577-020-00436-4

⁴⁶ Takahashi T, Iwasaki A. Sex differences in immune responses. Science 2021; 371: 347–8. doi:10.1126/science.abe7199

⁴⁷ Klein SL, Flanagan KL. Sex differences in immune responses. Nat Rev Immunol 2016; 16: 626–38. doi:10.1038/nri.2016.90

⁴⁸ Kwon H, Schafer JM, Song N-J, et al. Androgen conspires with the Cd8(+) T cell exhaustion program and contributes to sex bias in cancer. Sci Immunol 2022; 7: eabq2630. doi:10.1126/sciimmunol.abq2630

⁴⁹ Guan X, Polesso F, Wang C, et al. Androgen receptor activity in T cells limits Checkpoint blockade efficacy. Nature 2022; 606: 791–6. doi:10.1038/s41586-022-04522-6

⁵⁰ Bischof E, Wolfe J, Klein SL. Clinical trials for COVID-19 should include sex as a variable. J Clin Invest 2020; 130: 3350–2. doi:10.1172/JCI139306

⁵¹ Brady E, Nielsen MW, Andersen JP, et al. Lack of consideration of sex and gender in COVID-19 clinical studies. Nat Commun 2021; 12: 4015. doi:10.1038/s41467-021-24265-8

⁵² Bignucolo A, Scarabel L, Mezzalira S, et al. Sex disparities in efficacy in COVID-19 vaccines: A systematic review and meta-analysis. Vaccines (Basel) 2021; 9: 825. doi:10.3390/vaccines9080825

⁵³ Furman D, Hejblum BP, Simon N, et al. Systems analysis of sex differences reveals an immunosuppressive role for testosterone in the response to influenza vaccination. Proc Natl Acad Sci U S A 2014; 111: 869–74. doi:10.1073/pnas.1321060111

⁵⁴ Heyne S, Esser P, Werner A, et al. Attitudes toward a COVID-19 vaccine and vaccination status in cancer patients: a cross-sectional survey. J Cancer Res Clin Oncol 2022; 148: 1363–74. doi:10.1007/s00432-022-03961-y

⁵⁵ Wang W, Kaelber DC, Xu R, et al. Breakthrough SARS-Cov-2 infections, hospitalizations, and mortality in vaccinated patients with cancer in the US between December 2020 and November 2021. JAMA Oncol 2022; 8: 1027–34. doi:10.1001/jamaoncol.2022.1096

⁵⁶ Mair MJ, Mitterer M, Gattinger P, et al. Enhanced SARS-Cov-2 breakthrough infections in patients with hematologic and solid cancers due to Omicron. Cancer Cell 2022; 40: 444–6. doi:10.1016/j.ccell.2022.04.003