Immune changes beyond Th2 pathways during rapid multifood immunotherapy enabled with omalizumab

Abstract

Background: Multifood oral immunotherapy (mOIT) with adjunctive anti-IgE (omalizumab, XOLAIR^® ) treatment affords safe, effective, and rapid desensitization to multiple foods, although the specific immune mechanisms mediating this desensitization remain to be fully elucidated.

Methods: Participants in our phase 2 mOIT trial (NCT02643862) received omalizumab from baseline to week 16 and mOIT from week 8 to week 36. We compared the immune profile of PBMCs and plasma taken at baseline, week 8, and week 36 using high-dimensional mass cytometry, component-resolved diagnostics, the indirect basophil activation test, and Luminex.

Results: We found (i) decreased frequency of IL-4⁺ peanut-reactive CD4⁺ T cells and a marked downregulation of GPR15 expression and CXCR3 frequency among γδ and CD8⁺ T-cell subsets at week 8 during the initial, omalizumab-alone induction phase; (ii) significant upregulation of the skin-homing receptor CCR4 in peanut-reactive CD4⁺ T and Th2 effector memory (EM) cells and of cutaneous lymphocyte-associated antigen (CLA) in peanut-reactive CD8⁺ T and CD8⁺ EM cells; (iii) downregulation of CD86 expression among antigen-presenting cell subsets; and (iv) reduction in pro-inflammatory cytokines, notably IL-17, at week 36 post-OIT. We also observed significant attenuation of the Th2 phenotype post-OIT, defined by downregulation of IL-4 peanut-reactive T cells and OX40 in Th2EM cells, increased allergen component-specific IgG4/IgE ratio, and decreased allergen-driven activation of indirectly sensitized basophils.

Conclusions: This exploratory study provides novel comprehensive insight into the immune underpinnings of desensitization through omalizumab-facilitated mOIT. Moreover, this study provides encouraging results to support the complex immune changes that can be induced by OIT.

Keywords: IgE; biologics; food allergy; immunotherapy and tolerance induction; immunotherapy clinical; lymphocytes.

Fig. 1.

(A) Timeline for administration of Omalizumab and OIT dosing to study participants. Time points of blood collection and performance of mechanistic assays are shown. (B) Overview of the assays that were performed for each participant, the foods in the OIT regimen that the active arm participants received and the foods in this panel to which the untreated controls were sensitized. One participant in the active arm failed the DBPCFC to egg at week 36 (highlighted in pink).

Fig. 2.. Induction phase with anti-IgE treatment alone leads to downregulation of GPR15 and peanut (allergen)-induced IL-4.

Mass cytometry dataset on ex vivo peanut-stimulated baseline, week 8 and week 36 PBMCs from 15 peanut-treated participants was screened to identify significant differences between baseline and week 8. For each significant feature, the corresponding frequency and median expression plots are shown. (A) The frequency of GPR15⁺ peanut-reactive CD8⁺ T cells, the median expression of GPR15 among CD8⁺ EM and Th2 EM cells, and the frequency of GPR15⁺ naïve γδ T cells was significantly downregulated at week 8 compared to baseline. (B) Reduced percentage of naïve γδ T cells among week 8 vs. baseline PBMCs migrating to GPR15L placed in the lower chamber of a transwell. % Specific migration = % migrated towards GPR15L minus mean basal % migration to control medium (pairwise P value by Wilcoxon signed-rank test; P < 0.05 considered significant. Note: For n=5, the lowest possible P value in a Wilcoxon signed-rank test is 0.0625.). (C) The frequency of IL4⁺ peanut-reactive CD4⁺ T cells decreased significantly at week 8 vs. baseline. (For A and C: Pairwise P value by χ² test < 0.05 considered significant. Pairwise P values were computed only when linear mixed effects models including the three time points had an FDR-adjusted P < 0.2 by χ² tests.)

Fig. 3.. Allergen ingestion during mOIT enhances skin-homing potential, and modulates activation marker expression of select T cell subsets.

Mass cytometry dataset on ex vivo peanut-stimulated baseline, week 8 and week 36 PBMCs from 15 peanut-treated participants was screened to identify significant differences in week 8 vs. week 36 comparison. For each significant feature, the corresponding median expression or frequency plot is shown. (A) Increase in median expression of CCR4 on peanut-reactive CD4⁺ T cells and Th2 EM cells at week 36 vs. week 8. (B) Increase in the frequency of CLA⁺ peanut-reactive CD8⁺ T cells and CD8⁺ EM cells at week 36 vs. week 8. (C) The frequency of OX40⁺ Th2 EM cells was significantly downregulated at week 36 vs. week 8, and the frequency of PD1⁺ CD8⁺ EM, and median expression of PD1 on CD8⁺ EM cells, was decreased at week 36 vs. week 8. (Pairwise P value by χ² test < 0.05 considered significant. Pairwise P values were computed only when linear mixed effects models including the three time points had an FDR-adjusted P < 0.2 by χ² tests.)

Fig. 4.. Role of allergen-specific OIT in immunomodulation of APC subsets.

Mass cytometry dataset on ex vivo peanut-stimulated baseline, week 8 and week 36 PBMCs from 15 peanut-treated participants was screened to identify significant differences in week 8 vs. week 36. For each significant feature, the corresponding median expression or frequency plot is shown. (A-C) Significant decrease in the frequency of cells expressing CD86, and/or the median expression of CD86, among the evaluated APC subsets at week 36 vs. week 8. (A) Monocyte subsets (viz. classical, intermediate and non-classical monocytes), (B) DC subsets (viz. mDC1 and mDC2) and (C) memory B cells. (Pairwise P value by χ² test < 0.05 considered significant. Pairwise P values were computed only when linear mixed effects models including the three time points had an FDR-adjusted P < 0.2 by χ² tests.)

Fig. 5.. Component-resolved diagnostics and indirect BAT.

(A) Cognate food-specific and allergen component-specific IgE and IgG₄ was measured in baseline and week 36 plasma from 30 active participants desensitized through rapid multifood OIT (A) and 12 untreated multifood allergic controls (B). Components in the storage protein family of peanut and treenut allergens showed a marked increase in IgG₄/IgE ratio. (P values calculated using Wilcoxon signed-rank test. FDR-adjusted P are shown as Q. Q < 0.1 were considered significant. [In (B), all Q = 1, thus not included in the figure] (C) Baseline and week 36 plasma from 11 active, peanut-allergic, peanut-treated participants desensitized by rapid multifood OIT, and 5 peanut-allergic untreated controls, was assessed for its potential to activate basophils (i.e., to increase frequency of CD63⁺ basophils) in whole blood from a healthy donor through the indirect BAT assay. (P values calculated using Wilcoxon signed-rank test. P < 0.05 considered significant.)

Fig. 5.. Component-resolved diagnostics and indirect BAT.

(A) Cognate food-specific and allergen component-specific IgE and IgG₄ was measured in baseline and week 36 plasma from 30 active participants desensitized through rapid multifood OIT (A) and 12 untreated multifood allergic controls (B). Components in the storage protein family of peanut and treenut allergens showed a marked increase in IgG₄/IgE ratio. (P values calculated using Wilcoxon signed-rank test. FDR-adjusted P are shown as Q. Q < 0.1 were considered significant. [In (B), all Q = 1, thus not included in the figure] (C) Baseline and week 36 plasma from 11 active, peanut-allergic, peanut-treated participants desensitized by rapid multifood OIT, and 5 peanut-allergic untreated controls, was assessed for its potential to activate basophils (i.e., to increase frequency of CD63⁺ basophils) in whole blood from a healthy donor through the indirect BAT assay. (P values calculated using Wilcoxon signed-rank test. P < 0.05 considered significant.)

Fig. 5.. Component-resolved diagnostics and indirect BAT.

(A) Cognate food-specific and allergen component-specific IgE and IgG₄ was measured in baseline and week 36 plasma from 30 active participants desensitized through rapid multifood OIT (A) and 12 untreated multifood allergic controls (B). Components in the storage protein family of peanut and treenut allergens showed a marked increase in IgG₄/IgE ratio. (P values calculated using Wilcoxon signed-rank test. FDR-adjusted P are shown as Q. Q < 0.1 were considered significant. [In (B), all Q = 1, thus not included in the figure] (C) Baseline and week 36 plasma from 11 active, peanut-allergic, peanut-treated participants desensitized by rapid multifood OIT, and 5 peanut-allergic untreated controls, was assessed for its potential to activate basophils (i.e., to increase frequency of CD63⁺ basophils) in whole blood from a healthy donor through the indirect BAT assay. (P values calculated using Wilcoxon signed-rank test. P < 0.05 considered significant.)

Fig. 6.. OIT-induced decrease in Th2 cytokines and inflammation-associated cytokines and chemokines.

Culture supernatants harvested from baseline, week 8 and week 36 PBMCs from 16 (4 each for #FA = 2, 3, 4 and 5 respectively) study participants were assessed for 41 select cytokine and chemokines using Luminex. (A) Heatmap depicting cytokines and chemokines with significant expression changes in unstimulated supernatants. (B) Heatmap depicting cytokines and chemokines with significant expression changes in PMA/Ionomycin-stimulated supernatants. (C) Expression pattern for IL-4, IL-17, and FLT3L in PMA/Ionomycin-stimulated supernatants at the three assessed time points is highlighted with individual violin plots. (Pairwise P value by Wilcoxon signed-rank test < 0.05 considered significant. Pairwise P values were computed only when Friedman rank sum test including the three time points had an FDR-adjusted P < 0.2.)