T cell activation markers CD38 and HLA-DR indicative of non-seroconversion in anti-CD20-treated patients with multiple sclerosis following SARS-CoV-2 mRNA vaccination

Abstract

Background: Messenger RNA (mRNA) vaccines provide robust protection against SARS-CoV-2 in healthy individuals. However, immunity after vaccination of patients with multiple sclerosis (MS) treated with ocrelizumab (OCR), a B cell-depleting anti-CD20 monoclonal antibody, is not yet fully understood.

Methods: In this study, deep immune profiling techniques were employed to investigate the immune response induced by SARS-CoV-2 mRNA vaccines in untreated patients with MS (n=21), OCR-treated patients with MS (n=57) and healthy individuals (n=30).

Results: Among OCR-treated patients with MS, 63% did not produce detectable levels of antibodies (non-seroconverted), and those who did have lower spike receptor-binding domain-specific IgG responses compared with healthy individuals and untreated patients with MS. Before vaccination, no discernible immunological differences were observed between non-seroconverted and seroconverted OCR-treated patients with MS. However, non-seroconverted patients received overall more OCR infusions, had shorter intervals since their last OCR infusion and displayed higher OCR serum concentrations at the time of their initial vaccination. Following two vaccinations, non-seroconverted patients displayed smaller B cell compartments but instead exhibited more robust activation of general CD4⁺ and CD8⁺ T cell compartments, as indicated by upregulation of CD38 and HLA-DR surface expression, when compared with seroconverted patients.

Conclusion: These findings highlight the importance of optimising treatment regimens when scheduling SARS-CoV-2 vaccination for OCR-treated patients with MS to maximise their humoral and cellular immune responses. This study provides valuable insights for optimising vaccination strategies in OCR-treated patients with MS, including the identification of CD38 and HLA-DR as potential markers to explore vaccine efficacy in non-seroconverting OCR-treated patients with MS.

Keywords: COVID-19; MULTIPLE SCLEROSIS.

Figure 1

Circulating immune profiles of OCR-treated patients with MS. (A) SARS-CoV-2-unexposed patients with MS without treatment (n=21), patients with MS treated with OCR (n=57) and healthy controls (HC; n=30) were followed over time after Moderna SARS-CoV-2 mRNA vaccination. Blood or serum samples were collected at specified time points pre-vaccination and post-vaccination. (B) IgG titres to SARS-CoV-2 RBD before vaccination, 28 days after the first, before the second, and 7 and 28 days after the second vaccination. (C) The area under the curve was obtained to quantify the overall anti-RBD IgG antibody responses. (D,E) Age (D) and sex (E) distribution in HC, MS control, and anti-RBD IgG⁺ or anti-RBD IgG⁻ OCR-treated patients with MS. (F–H) The time since the last OCR treatment (F), the concentration of OCR at the day of the first vaccination (G). (H) Correlation plot between the OCR concentration and the time since last OCR treatment at baseline. (I) The number of OCR infusions before the first vaccination (J) FIt-SNE two-dimensional map and cluster identification from FlowSOM analysis of the high-dimensional flow cytometry immune panel, separated across groups and major adaptive immune populations displayed in each projection. (K) Surface expression intensity of indicated markers projected on the FIt-SNE map. (L) The number of circulating neutrophils, monocytes, NK cells, B cells, CD4⁺ T cells, CD8⁺ T cells and γδ T cells per μL of blood on day 0 before the first vaccination. (M) Volcano plot showing the abundance of circulating immune cells fold change of anti-RBD IgG⁺ versus anti-RBD IgG⁻ OCR-treated patients with MS (x-axis) and their Wilcoxon signed-rank test p values (y-axis) at day 0. Statistical significance between groups was determined with Wilcoxon signed-rank test (F, G, I and L) with Bonferroni-Holm multiple comparison correction (C, D and K). Associations in (H) were calculated using Spearman rank correlation. The p values are depicted as *<0.05, **<0.01 and ****<0.0001. AU, arbitrary units; MS, multiple sclerosis; NK, natural killer; ns, not significant; OCR, ocrelizumab; RBD, receptor-binding domain.

Figure 2

Distinct B cell responses in seroconverted OCR-treated patients with MS after vaccination. (A) The number of circulating B cells per μL of blood on day 49 (7 days after the second vaccination) in healthy control (HC, n=30), untreated MS controls (MS, n=21), seroconverted OCR-treated patients with MS (anti-RBD IgG+, n=21) and non-seroconverted OCR-treated patients with MS (anti-RBD IgG⁻, n=36). (B) FIt-SNE two-dimensional projection and cluster identification of circulating B cell populations from FlowSOM analysis of high-dimensional flow cytometry immune panel. The FIt-SNE projection is separated across groups and time points (day 0 and day 49). (C) The surface expression intensity of indicated markers is projected on the FIt-SNE map. (D) The number of circulating naïve B cells, transitional B cells, plasmablast, plasma cells, memory B cells, double negative (DN)-like B cells and activated B cells per μL of blood 7 days after the second vaccination (day 49). (E) Volcano plot showing the abundance of circulating immune cells fold change of patients with anti-RBD IgG⁺ versus anti-RBD IgG⁻ MS (x-axis) and their Wilcoxon signed-rank test p values (y-axis) at day 49. (F) Heatmap representing changes in B cell population counts per μL blood when comparing baseline (day 0) with day 7 post-second vaccination (day 49) within the different groups, blue indicating a reduction of that specific population between the time points, red an increase and white populations were not affected. (G) Correlations between the number of B cells 7 days after the second vaccination (day 49) and levels of anti-RBD IgG at 7 (day 49; top) or 28 days (day 70; bottom) after the second vaccination. Statistical significance was determined using a Wilcoxon signed-rank test (F,G) with Bonferroni-Holm multiple comparison correction (A and E). Associations in (G) were calculated using Spearman rank correlation. The p values are depicted as *<0.05, **<0.01, ***<0.001 and ****<0.0001. AU, arbitrary units; MS, multiple sclerosis; ns, not significant; OCR, ocrelizumab; RBD, receptor-binding domain.

Figure 3

Non-seroconverted OCR-treated patients with MS experience enhanced T cell activation after the second SARS-CoV-2 vaccination. (A) Number of spike-specific IFN-γ–producing T cell spots (SFU) before vaccination and 7 days post-second vaccination are plotted side by side in healthy controls (HC, n=8), untreated MS controls (MS, n=10), seroconverted OCR-treated patients with MS (anti-RBD IgG⁺, n=7) and non-seroconverted OCR-treated patients with MS (anti-RBD IgG⁻, n=17). Results are shown as the average number of SFUs of S1 and S2 together per 2×10⁵ cells after subtracting the SFU of unstimulated wells. (B) Heatmap representing changes in circulating T cell population counts per μL blood when comparing baseline (day 0) with day 7 post-second vaccination (day 49) within the different groups, blue indicating a reduction of that specific population between the time points, red an increase and white populations were not affected (HC, n=30; MS, n=21; anti-RBD IgG⁺, n=21; and anti-RBD IgG⁻, n=36). (C) Representative flow analysis plots showing the expression of HLA-DR and CD38 in CD8 central memory (CM) T cells pre-vaccination and 7 days post-second vaccination. (D) Volcano plot showing the circulating T cell populations expressing both HLA-DR and CD38 fold change of day 49 vs day 0 (x-axis) and their Wilcoxon signed-rank test p values (y-axis). (E–I) Frequency of CD38⁺HLA-DR⁺ CD8⁺ T CM (E), CD8⁺ T effector memory (EM; F), CD8⁺ T EF CD45RA+ (EMRA; G), CD4⁺ T CM (H) and CD4⁺ T EM (I) before (D0) and 7 days after second vaccination (D49). (J) Correlations of the number of spike-specific IFN-γ-producing T cell spots and the frequency of CD38⁺HLA-DR⁺ circulating CD8⁺ and CD4⁺ T cell populations at day 0 and day 49. (K) Correlation plots between the number of spike-specific IFN-γ-producing T cell spots and the frequency of CD38⁺HLA-DR⁺ circulating CD8 CM separated by group. Statistical significance was determined using a Wilcoxon signed-rank test (A (left)–C, E–I) with Bonferroni-Holm multiple comparison correction (A: right). Associations in J and K were calculated using Spearman rank correlation. The p values are depicted as *<0.05, **<0.01, ***<0.001 and ****<0.0001. IFN-γ, interferon-gamma; MS, multiple sclerosis; ns, not significant; OCR, ocrelizumab; RBD, receptor-binding domain; S1, spike 1; S2, spike 2; SCM, stem cell memory; SFU, spot-forming unit; Tfh, T follicular helper; Th, T helper.