Early decrease in basophil sensitivity to Ara h 2 precedes sustained unresponsiveness after peanut oral immunotherapy

Abstract

Background:

Only some peanut-allergic subjects undergoing oral immunotherapy (OIT) achieve sustained clinical response. Basophil activation could provide a functional surrogate of efficacy.

Objective:

We hypothesized that changes in basophil sensitivity and AUC to the immunodominant allergen Ara h 2 correlate with clinical responses to OIT.

Methods:

Peanut-allergic children, aged 7–13, enrolled in a single-center, open-label peanut OIT trial. Specific immunoglobulins were measured throughout OIT. Peripheral blood from multiple time points was stimulated in vitro with peanut allergens for flow cytometry assessment of the percent CD63^hi activated basophils.

Results:

Twenty-two of 30 subjects were successfully treated by OIT; post-avoidance, 9 achieved sustained unresponsiveness (SU) and 13 had transient desensitization (TD). Basophil sensitivity, measured by ED50, to Ara h 2 stimulation decreased from baseline in SU (post-OIT p= 0.0041, post-avoidance p=0.0011). At 3 months of OIT, basophil sensitivity in SU decreased from baseline compared to TD (median 18-fold vs. 3-fold, p=0.01) with an ROC of 0.84 and optimal fold-change of 4.9. Basophil AUC, measured by area-under-the-curve, to Ara h 2 was suppressed post-OIT equally in SU and TD (p=0.4). After avoidance, basophil AUC rebounded in TD but not SU (p<0.001). Passively sensitized basophils suppressed with post-avoidance SU plasma had lower AUC than TD plasma (6.4% vs. 38.9%, p=0.03).

Conclusions:

Early decreases in basophil sensitivity to Ara h 2 correlate with SU. Basophil AUC rebounds post-avoidance in TD. Therefore, different aspects of basophil activation may be useful for monitoring of OIT efficacy.

Clinical Implication:

Basophil sensitivity and AUC correlate with OIT efficacy.

Capsule summary:

Early decrease in basophil sensitivity to Ara h 2 correlates with sustained unresponsiveness after peanut oral immunotherapy.

Graphical Abstract

Introduction

The rising incidence and persistence of IgE-mediated peanut allergy^{1, 2} highlight the growing need for effective therapies. Peanut oral immunotherapy (OIT) induces desensitization, raising the threshold of exposure required to cause a reaction in most peanut allergic individuals who successfully complete it; however, only about 30% achieve more durable clinical efficacy.^{3, 4} Multiple immunological changes occur during OIT,⁵ including the suppression of basophil AUC⁶, the depletion of peripheral allergen-specific Th2 effector cells,⁷ the expansion of antigen-specific B cells,⁸ and the induction of allergen-specific antibodies.^{8, 9} However, the extent to which any of these changes are mechanistically related to clinical outcomes, and durable efficacy in particular, remains unclear.

Basophil activation to allergen stimulation has emerged as an informative surrogate marker of clinical allergic status.^{5, 10} For example, among untreated allergic patients subjected to a graded oral peanut challenge, the clinical severity of reaction correlates closely to the in vitro magnitude of basophil degranulation, reflected by the maximal or peak response. The threshold dose provoking a clinical reaction also correlates to in vitro basophil sensitivity, reflected by the ED50 of the dose-response.¹¹ These two features of basophil responses, basophil AUC and sensitivity, are somewhat independently influenced by distinct features of the antigen-antibody-Fc receptor complexes that form, including the IgE/non-IgE clonality, affinity,¹² glycosylation and intrinsic basophil reactivity.^13–17 Measures of basophil activation have been shown to have additional diagnostic utility for IgE mediated food allergies,¹⁸ though not for evaluating for sustained unresponsiveness.

In the context of oral immunotherapy, several clinical trials have demonstrated suppression of in vitro basophil AUC in subjects with decreased clinical reactivity to peanut.^{6, 19, 20} Previous work on basophil suppression with post-OIT serum has shown that this suppression can be mediated in vitro by antigen-specific IgG₄ through inhibitory FcγRIIb engagement.¹⁴ However, sustained efficacy of OIT, as measured by sustained clinical unresponsiveness to peanut after strict allergen avoidance, has been repeatedly shown to not correlate with the concentration of allergen-specific IgG₄ induction in previous trials.^{3, 4} On the other hand, in other forms of immunotherapy, functional antibody testing as opposed to specific antibody levels correlated better with clinical outcomes.¹⁷

We therefore hypothesized that OIT-induced suppression of basophil sensitivity to the immunodominant antigen Ara h 2 is a biomarker of sustained unresponsiveness. After longitudinally studying basophil activation both during OIT and after avoidance, we conclude that basophil sensitivity and AUC can not only better inform clinical outcomes than bulk measurements of allergen-specific antibodies but may also provide a surrogate functional measure of the induction of protective antigen-specific antibodies.

METHODS

Peanut Oral Immunotherapy Trials

Peripheral blood was obtained after IRB-approved consent from peanut allergic participants of an open-label randomized trial of peanut OIT (). Study inclusion criteria included participants aged 7–21 years old with a diagnosis of peanut allergy based on a clinical history of reaction to peanut within one hour of ingestion and either an elevated skin prick test (>8 mm wheal) or elevated peanut-specific IgE (CAP FEIA >10 kU/L).

A total of 30 patients were enrolled, including 4 in an observational control arm, who were monitored for 1 year, after which they also received active OIT with premeasured peanut flour (Medium Roast, Golden Peanut Co., Alpharetta, GA). The protocol, as previously described,²¹ began with a 1 day modified rush, 44 week long build up, and 12 week maintenance phase, at the end of which desensitization was evaluated with an oral food challenge. After 4 weeks of peanut avoidance, subjects underwent a double blind oral food challenge to peanut. Peripheral blood sampling was performed at baseline, every 4 weeks during the buildup phase and every 6 weeks during maintenance, and then at both food challenges.

Prior to the start of active OIT, 2 subjects voluntarily withdrew and 1 subject was a screen fail due to failure to tolerate a minimal dose on the modified rush. During the study, 3 subjects were withdrawn by the investigator due to persistent reactivity to peanut, 3 subjects withdrew during the study (1 due to difficulty adhering to timeline, 1 due to dose reaction, 1 due to concern about underlying peanut allergy), and 1 due to failure to tolerate the starting dose during the modified rush. Twenty-three patients reached the clinical endpoint, of which 1 subject failed to be effectively desensitized. Of the remaining 22 subjects, 9 had sustained unresponsiveness after 4 weeks of peanut avoidance and 13 regained responsiveness to peanut and were labeled as having transient desensitization (TD) (Figure 1).

Fig 1: Peanut OIT CONSORT diagram.

Flowchart of subjects enrolled in the peanut OIT clinical trial, in both the control-crossover and active OIT arms. Summary of clinical outcomes is located below the diagram, with the total number of patients who achieved sustained unresponsiveness (blue, n=9) and those with transient desensitization (red, n=13) defined by peanut tolerance on double-blind placebo-controlled challenge after 1 month of peanut avoidance.

Specific immunoglobulin measurement

Measurement of peanut and antigen specific immunoglobulin levels in the serum or plasma from peripheral blood of subjects undergoing OIT was determined using the ImmunoCAP 1000 instrument (Phadia AB) according to the manufacturer’s instructions.

Basophil activation testing

Using the FlowCAST assay (BÜHLMANN Laboratories, Switzerland), basophils in whole blood were stimulated with allergens (whole peanut extract, Ara h 1, Ara h 2, and Ara h 6) and stained with fluorescent CCR3 and CD63 antibodies according to the manufacturer’s protocol. Flow cytometry data was acquired on a LSRII or Fortessa (Becton, Dickinson and Company, NJ USA) in FCS file format. Using Bioconductor tools in R with basophil activation testing data from patients undergoing peanut oral immunotherapy (OIT), we developed a data-driven, automated gating strategy in which we used clustering approaches to define activated basophils as SSC^loCCR3⁺CD63^hi cells. In each individual experiment, the CD63 cutoff to determine activated CD63^hi basophils was determined as the 97.5 percentile of CD63 expression of unstimulated basophils. Gate visualization on plots of all stimulation conditions provided additional quality control (R 3.2.1, Biconductor 1.22.3).^22–24 Statistical characterization of basophils was performed and aggregated for additional analysis.

Basophil activation in passively sensitized basophils

PBMCs from healthy donors were isolated with a Ficoll density gradient and stripped of endogenous IgE using lactic acid stripping protocol12 followed by a one-hour incubation in undiluted plasma, from 1–2 months after the start of OIT. After washing the PBMCs, stimulation with Ara h 2 at various concentrations was performed in medium or in medium mixed with separate aliquots of autologous plasma at a 1:20 ratio from an early time point (1–3 months), at the end of immunotherapy, or after avoidance.

Statistical analysis

Longitudinal data was analyzed using generalized non-linear, mixed modeling in R. Dose response curves were fitted using the R package drc²⁵ assuming a four-parameter log-logistic dose-response model and the effective dose ED50 was derived. The default, non-robust least square unconstraint estimation fitting procedure (“mean”) was not always the optimal solution, so we also fitted the same dose-response using robust methods²⁵: median estimation (“median”), least median of squares (“lms”). Because the percent CD63^hi basophils were bounded (they cannot be less than 0 or larger than 100) we also extended the mean fitting procedure by a “constrain” estimation²⁵ restricting the lower and upper bounds (so-called box constraints) to those allowed by the data. For each of these four estimating procedures we selected the dose-response curve with the smallest AIC (Akaike’s Information Criterion). Non-determinable ED50 values were imputed using multivariate imputation by chained equations²⁶ with the application of the random forest algorithm. Area under the curve was calculated using trapezoidal approximation. Relative importance was calculated using the relimpo package.²⁷ Classifier performance was conducted using the RORC package.²⁸ Passive basophil experiments were analyzed using ANOVA and paired t-tests. Data analysis was conducted in R (R 3.2.1).²⁴

RESULTS

Peanut OIT clinical outcomes

Of the 30 subjects recruited for the clinical trial, 27 subjects received treatment with peanut (Table 1) (Fig 1). During immunotherapy, all subjects experienced at least one adverse event related to dosing. Of the adverse events, 96% were mild, with only 4% moderate and no severe adverse events. A total of 17 events (0.5% of all reported adverse events) required epinephrine administration among a total of 10 of the subjects. Adverse events requiring epinephrine treatment occurred most often during the build-up phase (11 events, 65%), followed by food challenges (5 events, 30%) and maintenance (1 event, 5%). During the study, 3 subjects withdrew due to persistent adverse events, and 1 subject withdrew due to failure to tolerate the minimal dose of peanut at the beginning of dosing.

Table 1:

Clinical demographics by outcome after peanut OIT.

	SU (n=9)	TD (n=13)	*p-values
Age (mean +/− SD)	9.33 (2.40)	9.62 (2.22)	0.78
Sex (% male)	7 (77.8)	9 (69.2)	1.0
Ethnicity			1.0
Non-Hispanic (%)	9 (100.0)	12 (92.3)
Race			0.46
White	0	2 (15.4)
Asian	1 (111)	1 (7.7)
> 1 race (%)	8 (88.9)	10 (76.9)
History of allergic rhinitis (%)	7 (77.8)	5 (38.5)	0.17
History of atopic dermatitis (%)	3 (33.3)	9 (69.2)	0.22
History of anaphylaxis (%)	1 (111)	6 (46.2)	0.20
History of other food allergies (%)	5 (55.6)	11 (84.6)	0.31
Age of diagnosis	1.44 (0.46)	2.23 (0.82)	0.02
Age of first reaction	2.11 (1.60)	3.36 (2.68)	0.23

^*P-values were calculated by Wilcox rank-sum test.

Of the 23 subjects that completed active immunotherapy and underwent an oral challenge to assess desensitization, 1 subject failed to tolerate 5 grams of peanut. Of the remaining 22 subjects who reached the primary endpoint, 9 achieved sustained unresponsiveness (SU) and tolerated 5 grams of peanut on challenge after 4 weeks of peanut avoidance, while 13 regained responsiveness to peanut post-avoidance and were designated as achieving transient desensitization (TD).

Subjects with TD required more time to build up to their maintenance dose (434 ± 39 vs. 394 ± 41 days, p=0.034) due to reactions during that build-up phase requiring dose modifications. Subjects with SU were diagnosed with peanut allergy earlier in life than TD subjects (1.44 ± 0.46 vs. 2.23 ± 0.82 years, p=0.02); no other differences in baseline age, gender, or ethnicity were found between these subjects (Table 1).

Kinetics of specific immunoglobulins

Consistent with previously published studies,^{4, 21, 29} antigen-specific immunoglobulins changed over the course of OIT (Fig 2), and pre-treatment serum peanut and Ara h 2 specific IgE levels tend to be higher in TD subjects compared to SU (Fig E1, A). In this clinical trial, we focused on Ara h 2, an immunodominant peanut antigen.³⁰ Circulating Ara h 2 specific IgE increased transiently and early in all subjects, increasing from baseline for the first 20 weeks (p<0.001) in TD subjects. Total circulating IgE levels also increased from 6–20 weeks (p<0.05) in TD subjects, but a similar trend did not reach significance in SU subjects (Fig E1). A significant decrease in Ara h 2 specific IgE from pre-OIT levels occurred after OIT in TD subjects (70.3 ± 1.6 to 42.7 ± 1.5 kU/L, p=0.004), and while there was a similar trend in SU subjects, it did not reach significance.

Fig 2: Serum immunoglobulins during peanut OIT.

A. Clinical schema of the active arm of peanut OIT, including pre-OIT, build-up with increasing peanut doses, 3-month maintenance, post immunotherapy oral food challenge (OFC), post-avoidance double-blind food challenge (DBFC) after 1 month of peanut avoidance, and follow-up period (FU01, FU02) with ad hoc peanut consumption. Black arrows mark peripheral blood and serum sampling. B. Serum Ara h 2 IgE, IgG, and IgG₄ in SU (blue) and TD (dashed, red) over time using a logistical regression model with random intercepts. C. Serum IgE:IgG ratio, and IgE:IgG₄ ratio in SU (blue) and TD (dashed, red) over time using a logistical regression model with random intercepts. Lines indicate means and shaded regions indicate standard deviation. X-axis lettered labels refer to visits, PreOIT (P), Build-up (B), Maintainance (M), Post-OIT OFC (O), Post-avoidance DBFC (D), Follow-up (F).

In this study, peanut-specific IgE levels at baseline correlated with clinical outcomes after OIT. Before the start of OIT, subjects who developed only TD had significantly higher levels of Ara h 2-specific IgE than those with SU (70.3 ± 1.6 vs. 8.6 ± 1.7 kU/L, p=0.002, Fig 2, B). However, the ratio of Ara h 2-specific IgE to total IgE was not significantly different (0.2 ±0.1 vs. 0.3 ±0.2, p=0.5).

Similar to previous studies, circulating Ara h 2-specific IgG and IgG₄ increased during OIT. Notably, the increase in Ara h 2-specific IgE and IgG levels occurred early in OIT, and Ara h 2-specific IgG levels peaked at the end of build-up (Fig. 2B). In this study, after peanut OIT, TD subjects had higher Ara h 2 IgG levels than SU subjects (51.2 ± 1.3 vs. 17.5 ± 1.4 ng/mL, p=0.01), with persistence through post-avoidance (43.8 ± 1.3 vs. 15.1 ± 1.4 ng/mL, p=0.01, Fig 2, B). There was a trend towards an increased Ara h 2 specific IgG₄ level in TD compared to SU subjects after OIT (24.7 ± 1.3 vs. 11.6 ± 1.4 ng/mL, p=0.07) that disappeared post-avoidance (18.1 ± 1.3 vs. 8.9 ± 1.4 ng/mL, p=0.11, Fig 2, B).

The ratios of Ara h 2 IgE to Ara h 2 IgG and Ara h 2 IgE to Ara h 2 IgG₄ were significantly increased in those with TD compared to those with SU at baseline (IgE/IgG: 27.4 ± 3.1 vs. 3.0 ± 3.4, p=0.003; IgE/IgG₄: 135.8 ± 3.6 vs. 21.4 ± 4.0, p=0.006) but were more similar after OIT (IgE/IgG 5.3 ± 2.9 vs. 3.4 ± 3.1, p=0.1; IgE/IgG₄, 8.2 ± 3.1 vs. 4.0 ± 3.4, p=0.09) or 1 month of avoidance (IgE/IgG 5.8 ± 2.9 vs. 3.6 ± 2.9, p=0.1; IgE/IgG₄ 9.4 ± 3.1 vs. 4.3 ± 3.4, p=0.06, Fig 2, C).

Basophil suppression during OIT

To study a functional correlate of the humoral response to immunotherapy, we monitored peanut allergen-induced basophil activation ex vivo in peripheral blood from subjects during the clinical trial (Fig 3, A). Basophil AUC, measured as the area-under-the-curve, can correlate with the degree of clinical symptoms on challenge³¹ (Fig 3, B). Stimulation with whole peanut protein was compared to Ara h 2 (Fig 3, C and D).

Fig 3: Ex vivo direct basophil activation during peanut OIT.

A. Peripheral basophil activation testing to Ara h 2 to identify degranulation as percentage of CD63^hi basophils by flow cytometry. B. Basophil sensitivity (ED50) and AUC (AUC, filled green) in representative SU (left, blue) and TD (right, red) subjects. C, D. Longitudinal basophil sensitivity and AUC using a generalized additive mixed model (with smoothing by Laplace approximate) to Ara h 2 (C) and peanut (D) in SU (blue) or TD (dashed, red). The y-axis is logarithmically scaled for basophil AUC, the peanut avoidance period is shaded in grey, and the black arrow marks 3 months of OIT. Labeled visits on the x-axis include pre-OIT (Bl), build-up visits (B), maintenance (M), post-OIT oral food challenge (O), post-avoidance double blinded food challenge (D), and the follow-up visits (F). Lines indicate means and shaded regions indicate standard deviation. X-axis lettered labels refer to visits, PreOIT (P), Build-up (B), Maintainance (M), Post-OIT OFC (O), Post-avoidance DBFC (D), Follow-up (F). E. The proportion of the variance in Ara h 2 ED50 change from baseline to post-avoidance explained by quantitative measurements of serum Ara h 2 specific immunoglobulins.

Despite the significant differences in Ara h 2 IgE levels, basophil AUC prior to OIT was similar in SU and TD subjects when basophils were stimulated with either Ara h 2 (68.2 ± 1.5 vs. 85.3 ± 1.4, p=0.7) (Fig 3, C) or with peanut (85.3 ± 1.5 vs. 112.2 ± 1.4, p=0.6)(Fig 3, D). Post-OIT, basophil AUC was significantly suppressed from baseline in SU subjects (Ara h 2 68.2 ± 1.5 vs. 8.1 ± 1.5, p<0.001; peanut 85.3 ± 1.5 vs. 12.6 ± 1.5, p<0.001), as well as in TD subjects (85.3 ± 1.4 vs. 17.6 ±1.4, p<0.001, for Ara h 2 and 112.2 ± 1.4 vs. 19.0 ± 1.4, p<0.001, for peanut). Basophil AUC to both Ara h 2 and peanut was similar in SU and TD subjects at the end of OIT. However, after 1 month of avoidance, basophil AUC in TD subjects rebounded, as did clinical reactivity, as opposed to SU subjects in whom both basophil AUC and clinical reactivity remained suppressed (Ara h 2, 12.0 ± 1.5 vs. 93.4 ± 1.4, p=0.0001; peanut 14.6 ± 1.5 vs. 101.8 ± 1.4, p=0.0005). These patterns were also found with basophil activation to Ara h 6 (Fig E2).

Basophil sensitivity, measured as the ED50 of the dose-dependent basophil activation curve, has been shown to change with other forms of immunotherapy.³² Previous work suggests that basophil sensitivity can change due to the valency of allergen and surface IgE interactions,¹² which we expect to change with induction of an effective inhibitory antibody response. We therefore sought to test whether basophil sensitivity was more suppressed in SU compared to TD subjects.

At baseline, basophil sensitivity was similar in SU and TD subjects to both Ara h 2 (SU 1.9 ± 1.5 vs. TD 3.8 ± 1.4 ng, p=0.2) and whole peanut (SU 3.5 ± 1.6 vs. TD 2.4 ± 1.4, p=0.5). (Fig 3, C and D). Post-OIT, basophil sensitivity to Ara h 2 significantly decreased in SU but not TD subjects (SU 22.4 ± 1.5 vs. TD 5.0 ± 1.4 ng, p=0.005), which was persistent post-avoidance (SU 15.8 ± 1.5 vs. TD 2.8 ± 1.4 ng, p= 0.001). In comparison, the differences in basophil sensitivity to peanut post-OIT and post-avoidance were not significant (post-OIT SU 29.1± 1.5 vs. TD 12.1 ± 1.4 ng, p= 0.1, post-avoidance SU 23.1 ± 1.5 vs. TD 9.0 ± 1.4 ng, p=0.08). Basophil sensitivity and AUC to Ara h 6 was similar to Ara h 2 (Fig E2).

To evaluate the influence of immunoglobulin levels on the OIT-induced change in basophil sensitivity to Ara h 2, we compared the change from pre-OIT to post-OIT in the bulk Ara h 2-specific serum IgE and IgG levels to the change in the ED50 to Ara h 2. Using a multiple linear regression model to analyze the relative importance of the antigen-specific immunoglobulin levels, we found that only about 20.8% of the change of basophil sensitivity to Ara h 2 is accounted for by the changes in bulk measurements of Ara h 2 specific serum immunoglobulins, with the largest contributions by the difference in Ara h 2 specific IgE and total IgE levels 27 (Fig 3, E). Our data suggests that additional qualitative factors of induced immunoglobulins, such as clonality or affinity, may be more relevant to the decrease in basophil sensitivity and clinical efficacy of OIT.

Early changes in basophil sensitivity to Ara h 2 as a biomarker of SU

We have previously described early immunological changes in the humoral response to Ara h 2.⁸ To test our hypothesis that the decrease in basophil sensitivity to Ara h 2 may be an effective biomarker of future SU, we used SU and TD in a binary classifier system. The fold change in early basophil sensitivity, as measured by the change in ED50 to Ara h 2 from baseline to 3 months on OIT, was significantly higher in subjects with SU than TD (median 18.9 vs. 3.1 fold-change, p = 0.01) (Fig 4, A). The receiver operator curve (ROC) for the early change of basophil AUC to Ara h 2 was 0.84 with an optimal fold change cutoff of 4.9 compared to the ROC of the change in peanut-stimulated basophil AUC (0.66), which has an optimal cutoff of 8.5 (Fig 4, B–E).

Fig 4: Early decrease in basophil sensitivity to Ara h 2 associated with sustained unresponsiveness after peanut OIT.

A. The fold change in basophil sensitivity to Ara h 2 from baseline to 3 months (Figure 3) was compared in subjects SU (blue) and TD (red). B. The receiver operator curve (ROC) of the early change of basophil sensitivity to Ara h 2 with an optimal cutoff (C) compared to the ROC of the early change of basophil sensitivity to peanut (D) with an optimal cutoff (E).

Indirect basophil suppression to Ara h 2 by serum

We hypothesized that basophil suppression is persistent post-avoidance due to serum immunoglobulins in SU subjects. To test this hypothesis, we used an indirect basophil activation test with peripheral blood mononuclear cells from healthy donors. After removal of endogenous cell-bound IgE and passive sensitization using plasma from individual SU or TD subjects at baseline, we stimulated the cells with Ara h 2 alone or in the presence of plasma from individual SU or TD subjects during or after OIT. We were able to compare suppression using the same individual’s plasma across 3 times points: early OIT (1–3 months after starting OIT), post-OIT, or post-avoidance (Fig E3).

Similar to the kinetics of basophil activation in ex vivo basophils, we demonstrated suppression of Ara h 2 stimulated basophils derived from healthy donors and sensitized with plasma from early OIT-treated subjects. Relative basophil AUC, normalized to stimulation in medium, decreased post-OIT in the entire group of subjects (62.5 ± 39.8% vs. 32 ± 36%, p=0.0005) and remained similar through post-avoidance (32 ± 33.5%, p=0.98). Moreover, basophil AUC was significantly lower when using plasma from SU subjects than from TD subjects post-OIT (median 6.4 ± 27.1%, vs. 38.9 ± 39.3%, p=0.02) and post-avoidance (median 12.3 ± 26.5% vs. 30.2 ± 37.2%, p=0.03, Fig 5). Sensitivity to Ara h 2 tended to be lower in SU compared to TD post-avoidance plasma treated basophils (ED50 mean 2.1 ± 2.2 vs. 1.8 ± 1.7, p=0.06, Fig E4).

Fig 5: Basophil AUC suppression is increased post-OIT and post-avoidance in SU.

The percent AUC of basophils stimulated with Ara h 2 and plasma from subjects with SU (blue) or TD (red) were normalized to Ara h 2 stimulation alone. The suppressive plasma was collected from 3 time points: early OIT (1–3 months), post-OIT, or post-avoidance period. These indirect basophil activation assays were performed after stripping off endogenous IgE and passively sensitizing healthy donor peripheral mononuclear blood cells (n=5).

DISCUSSION

In this study, we demonstrate that basophil sensitivity to Ara h 2 is a useful biomarker of sustained clinical efficacy after peanut oral immunotherapy. In those with sustained unresponsiveness post-avoidance, basophil sensitivity was significantly suppressed, as evidenced by an increase in the ED50 of the Ara h 2-stimulated basophil activation. On the other hand, basophils from subjects who developed only transient desensitization not only failed to change their basophil sensitivity, either early or later in OIT, but also had a significant rebound in their basophil AUC, as evidenced by AUC. Hence, monitoring basophil activation to Ara h 2 has two potential uses during OIT, not only as an early biomarker to identify those likely to develop sustained unresponsiveness, but also as a marker post-OIT to evaluate when clinical reactivity to peanut may be likely to return.

Our findings underscore other recent studies on several points. First, the overall kinetics of basophil activation are similar to those in other studies, with suppression of AUC while the dose of peanut administered increases during immunotherapy.³³ Second, the mirroring of clinical reactivity and basophil AUC post-avoidance is parallel to a previous study of peanut allergic patients showing that severity of clinical reactions correlates to basophil AUC.¹¹ Basophil sensitivity, on the other hand, correlated to the clinical threshold on oral challenge, which may be similar to our finding that basophil sensitivity decreases in patients who have long-term tolerance of higher amounts of peanut post-OIT and avoidance. To facilitate reproducible basophil activation data analysis for future applications, we pioneered automated flow cytometry analysis on the commercially available basophil activation kit used in this study and refined a robust statistical analysis for measures of basophil activation.³⁴ Third, allergen-specific IgG or IgG₄ are consistently induced by OIT, but without any significant correlation to clinical outcomes. Serum from those with clinical protection post-avoidance demonstrated greater suppression of indirect basophil activation testing suggesting that neutralizing antibody activity, rather than concentration, does correlate with clinical reactivity, specifically with sustained clinical protection. Lastly, while not predictive, pre-treatment specific IgE levels tend to be higher among individuals who do not attain persistent clinical tolerance – at least in the relatively short time frames that have been evaluated in trials of this nature. While four weeks of avoidance does not indicate long term tolerance, it does stratify the the patients in a manner that correlates with differences in OIT-induced peanut-specific immune response.

As allergic effector cells, basophils can serve as an integrative measure of clinical efficacy in allergen immunotherapy. The valency of the interaction between antigen and basophil surface IgE has been shown to directly influence basophil sensitivity.^{12, 35} Correlation of basophil suppression with clinical tolerance post-immunotherapy highlights its utility as a monitoring tool for the efficacy of peanut immunotherapy³⁶ without subjecting patients to repeat oral challenges. Moreover, the indirect basophil test captures the functional ability of specific antibodies to suppress basophil activation. In this study, qualitative rather than quantitative Ara h 2 specific immunoglobulin changes correlated with clinical tolerance post-immunotherapy, as subjects with SU had lower levels of induced Ara h 2-specific IgG with stronger suppressive ability. At baseline, those with TD have significantly higher levels of IgE, but by the end of active immunotherapy, the Ara h 2 IgE to IgG ratio is similar, again underscoring that qualitative rather than quantitative differences in SU sera account for their stronger suppressive capabilities. Basophil sensitivity is modulated by the effective valency of antigen-immunoglobulin interactions, which in turn are affected by specific antibody repertoire. Post-immunotherapy suppression of basophil activation by sera may be influenced by many factors including inhibitory Fc receptor engagement¹⁴ as well as antigen sequestration from IgE-crosslinking on the surface of basophils. The further study of the relative contributions of these diverse mechanisms may shed light on other regulatory mechanisms that may be important to clinical efficacy, including the affinity of immunoglobulin Fc for inhibitory receptors and modulation of inhibitory receptor signaling pathways.

In summary, the decrease in basophil sensitivity to the immunodominant antigen Ara h 2 is persistent in OIT-treated patients with sustained unresponsiveness and can be used as an early biomarker to identify those most likely to attain durable clinical efficacy. On the other hand, basophil AUC to Ara h 2 post-OIT closely mirrors clinical reactivity to peanut in OIT-treated individuals. The OIT-induced suppression of basophil activation is influenced by qualitative characteristics of the induced immunoglobulin repertoire. As humoral responses are likely an important component of long-lived tolerance to allergen induced by immunotherapy, refined measures of induced immunoglobulin responses will help better understand, differentiate, and promote more effective clinical responses to immunotherapy.

Supplementary Material

E1. Fig E1: Serum antigen-specific IgE during baseline and peanut OIT.

A. Serum peanut specific IgE and Ara h 2 IgE in SU (blue) and TD (dashed, red) subjects at baseline.

B. Serum peanut specific IgE and total IgE in SU (blue) and TD (dashed, red) subjects over time using a logistical regression model with random intercepts. Lines indicate means and shaded regions indicate standard deviation.

E2. Fig E2: Direct ex vivo basophil sensitivity to Ara h 6.

Direct ex vivo basophil sensitivity, as measured by ED50, and basophil AUC, measured as AUC, of the dose-response curve to Ara h 6 using a generalized additive mixed model (with Laplace approximate smoothing) in SU (blue) and TD (dashed, red). The y-axis is logarithmically scaled in the AUC graph and the period of peanut avoidance is shaded in grey. Lines indicate means and shaded regions indicate standard deviation.

E3. Fig E3: Suppression of basophil activation by serum.

Basophils from normal donors were sensitized with SU serum (A) or TD serum (B), activated with various doses of Ara h 2, and suppressed with early, post-OIT and post-avoidance sera from the same patient or with medium only. Experimental controls included sensitized basophil stimulation with medium or anti-IgE. Basophil activation was assessed by flow cytometry including CD203c (y-axis) and CD63 (x-axis). The percentage of CD63^hi activated basophils were quantified (right-sided quadrants) for calculation of AUC or ED50.

E4. Fig E4: Plasma from post-OIT subjects with sustained unresponsiveness decreases basophil sensitivity to Ara h 2.

A. Basophil sensitivity, measured as ED50 of the dose-response curve, of basophils suppressed with SU (blue, n=5) or TD (red, n=5) Plasma from early, post-OIT, and post-avoidance time points of basophils sensitized with SU (top panel) or TD (bottom panel) IgE on activation with Ara h 2. Comparison used paired Wilcox rank sum tests.

Key messages:

• The decrease in post-OIT basophil sensitivity to Ara h 2 in those with sustained unresponsiveness has utility as early as 3 months during OIT.

• Basophil AUC to Arah2 is a biomarker of clinical reactivity post-OIT in patients with transient desensitization.

Abbreviations:

OIT

oral immunotherapy

OFC

oral food challenge

DBFC

double blind food challenge

AUC

Area under the curve

ED50

dose that induces 50% of the maximal basophil response