Phase 1 trial supports safety and mechanism of action of peptide immunotherapy for peanut allergy

Food allergy is a leading cause of anaphylaxis worldwide. Allergen‐specific immunotherapy is the only treatment shown to modify the natural history of allergic disease, but application to food allergy has been hindered by risk of severe allergic reactions and short‐lived efficacy. Allergen‐derived peptides could provide a solution. PVX108 comprises seven short peptides representing immunodominant T‐cell epitopes of major peanut allergens for treatment of peanut allergy.


| INTRODUC TI ON
Peanut allergy affects 1%-2% of Westernised nations and is a leading cause of food-induced anaphylaxis, placing a significant burden on patients, families and healthcare systems. 1 Allergen-specific immunotherapy (AIT) is the only treatment shown to modify the course of allergic disease.However, administration of whole allergens can trigger IgE-mediated reactions during treatment. 2e only regulatory approved food AIT is a peanut powder based oral immunotherapy (OIT; Palforzia®).However, a metaanalysis of 12 controlled trials revealed that OIT administration significantly increased the risk of anaphylaxis compared to peanut avoidance without treatment. 3][6][7][8][9][10] OIT and EPIT trials for peanut allergy indicate that efficacy wanes rapidly following treatment cessation, 11,12 suggesting requirement for indefinite daily dosing to maintain effects, which increases compliance issues and risk. 5directing the balance of allergen-specific CD4 + T-cell phenotypes towards that associated with a non-allergic status, by reducing T-helper (Th) 2 subsets and/or favouring other Th and Award Number: U19 AI125378-01; The Australian Food Allergy Foundation Phase 2 trials would provide important proof-of-concept for using peptides to treat food allergy.

K E Y W O R D S
food allergy, peptide immunotherapy, phase 1 clinical trial, T cell

G R A P H I C A L A B S T R A C T
PVX108 is a peptide immunotherapy designed to treat peanut allergy.PVX108 induced negligible activation of peanut-sensitized basophils ex vivo.A randomized, placebo-controlled Phase 1 clinical trial showed that PVX108 was safe and well tolerated in peanut-allergic adults and induced a shift in peanut-reactive T-cell phenotype balance to reduce Th2A dominance.Abbreviations: CCR6, C-C chemokine receptor type 6; ISR, injection site reaction; PVX108, a peptide immunotherapy for peanut allergy, comprising seven synthetic peptides representing immunodominant T-cell epitopes from major peanut allergens Ara h 1 and Ara h 2; ST2, interleukin-1 receptor-like 1; Th, T helper.
4][15] Short, synthetic peptides representing T-cell epitopes of major allergies are designed to induce tolerance via CD4 + T-cell modulation, without administering whole allergens.
Peptides (<29 aa) are too small to cross-link IgE, significantly reducing risk of mast cell activation and consequent allergic reactions.In addition to improving safety, AIT without inflammation has potential to improve efficiency of tolerance induction by avoiding upregulation of co-signalling molecules on antigen presenting cells.T-cell receptor (TCR) engagement by its peptide epitope (complexed to MHC II) without simultaneous engagement of co-stimulatory receptors favours tolerogenic or regulatory responses in T cells. 16,179][20] Once a shift is induced such that the net T-cell response favours tolerance, the modified T cells are anticipated to alter other immune cells (indirect downstream effects) to progressively lower immune reactivity to allergen.T cells can directly supress mast cell degranulation, 21 and reduce inflammatory cell retention, activation and/or survival in tissues. 22Collectively, these processes can redirect or suppress both acute and long-term allergic responses.
In contrast to whole allergen AIT, B cells are not a direct target for PVX108 binding and therefore are not anticipated to contribute to early therapeutic effects.8][29] In the case of peanut, exposure would be restricted to accidental ingestion or (low level) environmental exposure which would change the timing and magnitude of resulting antibody changes compared to whole allergen AIT.Where B-cell engagement may differ from traditional AIT, redirecting the net T-cell response to allergen is an anticipated and desired effect for both peptide and whole allergen AIT.Hence, while there will be overlap in the cell types involved in peptide and whole allergen AIT, the order and nature of the changes are anticipated to differ for peptides, as are their relative contribution to efficacy. 26Finally, unlike natural extracts, synthetic peptides are also readily standardized for pharmaceutical production.
PVX108 is a peptide immunotherapy for peanut allergy, comprising seven synthetic peptides representing immunodominant Tcell epitopes from major peanut allergens Ara h 1 and Ara h 2. 30,31 Herein, we report pre-clinical safety data from basophil activation tests, followed by safety and tolerability data from a randomized controlled Phase 1 clinical trial of PVX108 in peanut-allergic adults, the first ever trial of peptide immunotherapy for food allergy.We also report data from exploratory assays of PVX108 biological activity in Phase 1 participants.

| PVX108 design and manufacture
PVX108 comprises seven peptides (11-20 amino acids) representing immunodominant CD4 + T-cell epitopes of Ara h 1 (peptides #1-4) and Ara h 2 (peptides #5-7). 30,31cell and HLA-binding assays used peptides synthesized by Mimotopes or Think Peptides, respectively.Current Good and basophil activation to peanut ≥15%.Food challenges were not performed for safety reasons.The primary objective was to determine safety and tolerability of single and multiple PVX108 doses.
The secondary objective was to guide Phase 2 dose regimens.An exploratory objective was to evaluate potential biomarkers of PVX108 activity.These were measured in Stage 2 participants and included an 18-month follow-up visit (extension study AVX-001-EXT) to evaluate longer-term effects.No treatment was administered in AVX-001-EXT.Participants were instructed to avoid peanut throughout the studies.Being the first-in-human trial with PVX108, no sample size analyses were performed.Cohort sizes were selected using relevant precedents and guidance to meet the primary safety objective, but statistical comparisons were only performed for biomarker investigations.Expiry Volume and Peak Expiratory Flow (PEF)).Participants were contacted approximately 24 h after each dose to assess any AEs since discharge from the study unit.In Stage 2, these parameters were also measured 4 weeks after the last dose.In this study, a hypersensitivity reaction was defined as any form of an allergic hypersensitivity reaction, local or systemic, including but not limited to anaphylaxis.

| Protocol amendments
The original protocol included another secondary objective to investigate safety and tolerability of an adaptive intradermal dose escalation regimen of PVX108 in peanut-allergic adults.However, this became redundant after Stage 1 completion, when the Safety Review Committee recommended that Stage 2 dosing could commence at the highest dose of 150 nmol without dose escalation.The number of doses to be administered was also reduced from 9 to 6 over 16 weeks.

| Skin prick test
SPTs were performed on forearm (flexor) with neat peanut extract (1:20 w/v, Stallergenes Greer), plus three 10-fold dilutions for Stage 2 biomarker evaluation (1 in 10, 100 and 1000).Histamine and 50% glycerine (vehicle) provided positive and negative controls, respectively.Wheal size (mm) was recorded as the mean of the longest and shortest perpendicular wheal axes.A valid test required wheals of <2 and >4 mm for negative and positive controls, respectively.
An endpoint titration approach was used for biomarker analyses, 32 with intra-participant responses followed at the lowest peanut dilution giving a positive wheal (≥3 mm) at pre-dose.

| Peanut-specific T-cell phenotyping
Peanut-specific T-cell phenotyping and RNA-sequencing was performed on cryopreserved PBMC as described previously. 33Briefly, 10 million PBMC/mL were cultured with 10 μg/mL of Ara h 1 and Ara h 2 (Indoor Biotechnologies) and 1 μg/mL of anti-CD40 blocking monoclonal antibody (mAb) (clone HB14, Miltenyi Biotec) to maintain CD154 at the surface.After 18 h cells were harvested and 1/100th saved for cell frequency estimations, calculated by F = n/N, where n = CD154 + cells in the bound fraction after enrichment and N = total CD4 + CD45RA − T cells. 34Remaining cells were labelled with anti-CD154-PE-CF594 and anti-CD137-PE-Cy7 mAb anti-PE magnetic beads, then enriched on magnetic columns (Miltenyi Biotec), labelled with surface marker antibodies of interest (Table S5) and acquired by flow cytometry (FACS Fusion™, BD Biosciences).Specific function-related surface marker combinations identified mutually exclusive CD4 + Th-phenotypes (Figure 5A and Figure S2).RNA-sequencing library preparation and analysis was performed as described previously. 33In total, 307 samples were sequenced.269 (88%) samples passed the quality criteria.

| Statistical analyses
For serum immunoglobulin and SPT data, statistical analysis was not pre-defined.For T-cell data, pre-planned statistical assessments included one-way ANOVA for assessing within group changes over time, with p < .05considered significant.Post hoc, two-way RM ANOVA was also performed with Sidak's and Tukey's multiple comparisons tests for comparisons within and between groups over time.Data were log transformed where appropriate and possible.
Zero values within the T-cell dataset prevented log transformation, so square root transformed data were used for statistical analysis.
Residual plots were generated for two-way RM ANOVAs to determine whether normality and homoscedasticity assumptions were met.Being exploratory assays, p < .1 was taken into consideration to identify potentially relevant effects for future investigation.

| Pre-clinical safety and characterization of PVX108
Prior to pre-clinical safety evaluation, PVX108 peptides were assessed for their ability to bind different human MHC II (HLA) molecules and activate human CD4 + T cells.Each PVX108 peptide bound distinct combinations of 7-19 HLA II molecules of 52 tested, with Ara h 1 and 2 peptides collectively binding 24 and 19 different HLA, respectively (Figure S1A).Each PVX108 peptide also induced concentrationdependent CD4 + T-cell activation within peanut-allergic donor PBMC (Figure S1B).All 39 donors recognized ≥1 peptide, with >70% recognizing ≥5 peptides, and a third recognizing all seven peptides (Table S1).
To prevent treatment-induced anaphylaxis, PVX108 must avoid IgE cross-linking on histamine-releasing cells.To test this, basophil activation to PVX108 was assessed ex vivo in whole blood from 185 peanut-allergic adults with confirmed basophil reactivity to peanut (Figure 1).PVX108 induced no activation across a million-fold concentration range, in all except one donor where low-level activation (20.2% CD63 + basophils) was observed at the highest concentration tested (100 nmol/mL).Further evaluation of this donor replicated the response to 100 nmol/mL PVX108 (18.3% CD63 + basophils) but produced negative results for activation at 200 nmol/mL (11.9% CD63 + basophils).In contrast, peanut-induced basophil activation reached 84% and was still increasing at the highest concentration tested (5 μg/mL), suggesting that ≈20% was the maximum activation possible by PVX108.

| Phase 1 trial of PVX108 in peanut-allergic adults
Safety and tolerability of PVX108 were evaluated in a randomized, double-blind placebo-controlled Phase 1 clinical trial (AVX-001) of 67 peanut-allergic adults comprising a single ascending dose arm (Stage 1) and a repeat-dose arm (Stage 2) (Figure 2).
The CONSORT flow diagram and participant demographics for both stages of the trial are shown in Figure 3 2), composed of 120 mild, 24 moderate and two severe events.Both severe events were considered unrelated to PVX108.Ninety-four TEAEs were assessed as treatment-related, 87 in PVX108-treated participants and seven in placebo.Seventy-two of these events were mild (n = 64) or moderate (n = 8) ISRs, all of which occurred in PVX108-treated participants (11 of 13 (84.6%)participants in total), and primarily comprised erythema (n = 58) (Table 2 and Table S2B).
ISRs were also recorded in placebo participants, but none reached the classification (Grade 1) for an AE.Aside from ISRs, the only treatment-related TEAE reported in more than one participant per group was headache (2 events) in the placebo group (Table 2 and Table S2B).A single asthma event occurred in a PVX108-treated

| Exploratory mechanistic studies in Phase 1 participants
Activation-induced marker assays were used to monitor distinct peanut-reactive CD4 + T-cell populations within PBMC from trial participants, following in vitro stimulation with Ara h 1 and Ara h 2 (Figure 5A and Figure S2).The frequency (number per million) of peanut-reactive (CD154 − /CD137 + /LAP + /GARP + ) Treg cells within the total CD4 + memory T-cell population increased significantly from pre-dose to Month 18 in the active group, but not the placebo group (Figure 5B).The frequency of peanut-reactive (CD154 + ) effector Th cells increased significantly at Week 21 and Month 18 compared to pre-dose in the active group, although similar trends were also observed for placebo (Figure S3A).Frequencies increased in all peanutreactive Th subsets within the active group except for the Th2A subset (associated with pathology in allergy 33,35,36 ; Figure S3B), with the largest and most significant increases seen in Th17 cells (Figure S3C).In placebo, although patterns were similar, changes were consistently less pronounced than in active (Figure S3B,C).
However, the changes over time did not differ between groups for frequency assessments.S3).Peanut-specific IgG4 levels increased subtly from pre-dose to Month 18 in active but not placebo, with increases primarily emerging after dosing from Week 21 to Month 18 (Figure S5).There were no consistent changes in SPT wheal size over time at a group level, although 5/9 active participants showed decreased SPT responses at Month 18 compared to pre-dose, which was not observed for any placebo participant (0/5).

| DISCUSS ION
PVX108 is the first peptide-based immunotherapy administered to food-allergic subjects.In contrast to trials of other specific immunotherapies containing whole peanut allergens, 3,7,8,[43][44][45][46][47][48] there was no evidence of treatment-related hypersensitivity events, and no AEs of clinical concern following administration of 97 intradermal doses of up to 150 nmol (approximately 1.7 mg) PVX108 across 46 peanutallergic adults, including participants with asthma and prior anaphylaxis to peanut.ISRs were the only AE with higher incidence in PVX108-treated participants compared to placebo.These predominantly consisted of transient, mild erythema and did not increase in severity or frequency with increasing PVX108 dose level or exposure, consistent with a non-specific response to the presence of a peptide or protein.0][51] In the Phase 3 REALISE trial, patch site itchiness, redness (erythema) and swelling were not even reported as AEs in the first 6 months of use unless they led to study discontinuation or SAEs.Using these criteria, none of the ISRs in our trial (mild or moderate) would have been reported as AEs.TA B L E 1 Baseline demographics.
In a Phase 3 trial of Palforzia, the only FDA-approved whole peanut OIT, 14.2% of the active group had a systemic allergic reaction related to treatment during the intervention period and required epinephrine administration. 48Over 50% of these reactions was considered moderate, and 1.9% were considered severe.Over a third of the participants that experienced anaphylaxis during the trial experienced more than one episode.The most common adverse events in the active group involved the gastrointestinal tract (including abdominal pain and vomiting) and the respiratory tract.
Overall, 11.6% of the active group withdrew due to treatmentrelated AEs, the majority of which comprised acute or chronic gastrointestinal reactions (6.5%), followed by respiratory tract reactions (3%), systemic allergic reactions (1.9%) and skin reactions (1.3%). 48Although safer than OIT, treatment-related systemic allergic events were still reported in 4.8% of participants in the REALISE trial for peanut EPIT, the majority of which were moderate events, and approximately 50% of which required epinephrine. 51Similar systemic allergic event rates were reported in the first EPIT Phase 3 trial. 10though the above studies consisted of larger cohorts, our study still had sensitivity to detect a systemic allergic reaction rate of 2.2% for all participants receiving PVX108, and 2.9% for those receiving 10 nmol or above (the lowest dose progressed to Phase 2 studies).It is also important to consider that the aforementioned studies were also primarily conducted in children who tend to experience less frequent or severe reactions than adults in OIT. 524][55][56] In addition to its safety profile, pre-clinical testing TA B L E 2 Summary of all adverse events by treatment group.Abbreviations: (%) 1 , percentage of participants experiencing an adverse event; (%) 2 , percentage of adverse events; n 1 , number of participants experiencing an adverse event; n 2 , number of adverse events; TEAE, treatment emergent adverse event.a Indicates moderate severity, all other listed non-injection site treatment-related TEAEs were mild.
showed PVX108 bound a wide range of HLA II molecules, (collectively expressed by >99.5% of the population) and induced doserelated T-cell responses, confirming that these binding properties translate to physiologically relevant responses in target T cells.
While the primary endpoint of the Phase 1 study was safety and tolerability of PVX108, we had the opportunity to assess several immunological parameters as an exploratory endpoint.There was a decrease in the percentage of ST2 + Th2A cells and a corresponding increase in the percentage CCR6 + Th17 cells over time within the peanut-reactive Th cell population in participants receiving PVX108.Frequency data suggested this shift was predominantly driven by expansion of Th17 cells while Th2A numbers remained unchanged.Comparable decreases in the Th2A:Th17 ratio within the peanut-reactive Th pool were recently associated with reduced clinical reactivity to peanut in the Phase 2 OIT IMPACT study 57 and the Phase 3 PALISADE trial of Palforzia. 33In this case, frequency data suggested reductions in Th2A dominated, while Th17 numbers remained constant.The CCR6 + Th17like cells in our study and the IMPACT study co-expressed FOXP3, suggesting they may also have regulatory function.CCR6 + Th17 cells can transition to CCR6 + Foxp3 + Treg via shifts in Foxp3 and RORγt expression, influenced by TGFβ. 14,58,59Interestingly, both murine and clinical studies suggest that delivering immunotherapy via the skin (K) Pre-selected genes associated with distinct Th phenotypes within sorted populations for pre-dose (D1) samples (PVX108-treated (A); placebo-treated (P)).(L, M) Heatmaps of gene expression within peanut-reactive (total CD154 + ) Th cells over time (D1: pre-dose; D141: Week 21; M18: Month 18) for PVX108-treated participants with pre-dose samples (n = 7).(L) Pre-selected genes correlating with Th subsets.(M) Genes selected on differential expression between time points (adjusted p < .1).Gene clusters annotated by changes in expression over time.
1][62][63] These data suggest peptides and whole allergen may achieve similar shifts in the net T-cell response via different mechanisms.
Decreases in allergen-reactive Th2 cells particularly Th2A cells 13,33,35,[64][65][66][67] and increases in regulatory T cells 14,15,54,57,58 are both widely associated with clinical benefit following whole allergen AIT, including for peanut allergy.However, in the PALISADE trial, T-cell changes showed minimal progression after the updosing phase, and in the IMPACT trial, they reverted rapidly following treatment cessation, as did clinical efficacy.In contrast, the T-cell changes observed during our study strengthened posttreatment, becoming most pronounced by Month 18.It is encouraging that the kinetics of these changes align with the kinetics of efficacy outcomes in aero-allergen peptide immunotherapy trials, whereby efficacy improved 6-8 months after a short course of 4-11 injections over 3-4 months, with effects sustained for 1-2 years. 23,24,68,69In contrast, both OIT and EPIT have struggled to achieve sustained effects post-treatment, even after up to 5 years of daily dosing. 12,70e preliminary assessment of B cell (antibody) and SPT re-

2. 2 . 2 |
Randomization, blinding and study procedures Randomization, blinding and study procedures are described in the AVX-001 Protocol Synopsis (Data S1).Participants were randomized between 10 May 2017 and 21 June 2018.Recruitment ceased upon reaching randomization targets.The last participant visit in AVX-001 was 9 November 2018.AVX-001-EXT visits were from 11 September 2019 to 16 January 2020.Primary outcome measures of safety were assessed pre-dose and at regular intervals up to 10 h after each dose and included adverse events (AEs) (including injection site reactions (ISRs)), vital signs, electrocardiogram, physical examination, clinical laboratory tests (haematology, clinical chemistry, liver function, coagulation and urinalysis) and respiratory function tests (spirometric Forced
Pre-clinical studies 2.3.1 | Peptide-HLA binding Peptide binding to 52 HLA II molecules was assessed by ProImmune REVEAL assay (www.proim mune.com).A biologically relevant binding score threshold (1.99) was determined by aligning REVEAL data with T-cell response data.The probability of having ≥1 of the HLA molecules bound by PVX108 was calculated using published HLA allele frequencies (ProImmune).

F
I G U R E 1 Basophil activation in response to PVX108.Percentage activated (CD63 + ), basophils (CCR3 + ) in blood from peanut-allergic donors (n = 185) stimulated with PVX108 or peanut extract (n = 145 for 0.005-0.00005μg/mL peanut).100 nmol PVX108 ≈ 1.2 mg total peptide.Stimulation buffer (No Antigen) and placebo were negative controls.Anti-IgE and Formyl-Methionyl-Leucyl-Phenylalanine (fMLP) were positive controls.Red bars indicate mean values.There were no serious adverse events (SAEs), no clinically significant changes in electrocardiograms or safety laboratory parameters and no adverse events (AE) of clinical concern (Table2and TableS2A,B).The only acute allergic or hypersensitivity reaction was in response to a suspected accidental peanut encounter, 11 days after a PVX108 dose.This was treated with an anti-histamine (Loratadine) and deemed unrelated to treatment.In PVX108-treated participants, all post-dose Peak Expiratory Flow (PEF) assessments were within 20% of the pre-dose PEF values (Figure4A,B).The only AE to occur more frequently in PVX108-treated participants than placebo were transient injection site reactions (ISRs), which all resolved without intervention or sequelae, and predominantly comprised mild (Grade 1) erythema ( 65 were treatment emergent adverse events (TEAEs), composed of 57 mild (Grade 1) and 8 moderate (Grade 2) events.Forty-two TEAEs were considered treatment-related, 31 in PVX108-treated participants and 11 in placebo (Table2 and Table S2A).Twenty-four of these events were mild (n = 23) or moderate (n = 1) ISRs, primarily erythema (n = 19), and 20 of which occurred in PVX108-treated participants.Overall, ISRs were observed in 18 of 33 (54.5%)PVX108-treated participants and in 4 of 16 (25%) placebo-treated participants.Aside from ISRs, the only treatment-related TEAEs to occur in more than one participant per group were throat irritation (2 events) in PVX108treated participants, and asthma (2 events) in placebo-treated participants, one of which was treated with inhaled Salbutamol (

R E 2
Phase I trial design.The trial (AVX-001) was conducted in two stages.Stage 1: single ascending dose study enrolling eight cohorts, randomized 2:1 PVX108:placebo (49 participants total; one participant was withdrawn and replaced).Doses indicated by cohort.Stage 2: repeat-dose study enrolling one cohort randomized 2:1 PVX108:placebo (18 participants total).Blood samples were collected for exploratory immunological analyses at pre-dose, Week 21 and Month 18 (in AVX-001-Extension comprising a single follow-up visit).

participant 9 .
5 h after the fifth dose which was considered possibly related to treatment and treated with inhaled Salbutamol.The 15 participants who completed Stage 2 were also assessed for evidence of early immune changes induced by PVX108 to inform biomarker investigations in future studies.These participants also completed extension study AVX-001-EXT (an 18-month follow-up visit for exploratory biomarker assessments only).No AEs were reported in AVX-001-EXT.
While frequency data can inform mechanisms by which T-cell changes occur, cell numbers are more susceptible to technical variability between timepoints than monitoring subset ratios within the peanut-reactive Th population.These ratios are also a better determinant of the net Th response to peanut.The balance of phenotypes within the total peanut-reactive Th population changed over time following PVX108 treatment compared to placebo.Significant decreases in the percentage of Th2A (Figure5D,E), conventional Th2 (Figure5F,G) and Th1 cells (FigureS3D) and significant increases in the percentage of Th17 cells (Figure5H,I) within the peanut-reactive Th population were observed in the active group over time but not the placebo group (summarized in Figure5J).Similar results were observed for the percentage of cells expressing Th2-associated markers ST2 and CRTH2, and Th17 associated marker CCR6 within this population (FigureS3E-G, respectively).For all these subsets (plus GARP + Treg), the differences between the groups were consistently more distinct by month 18 compared to pre-dose, though none fell below p = .05.However, the most pronounced differences between active and placebo were in the changes over time in Th2A and Th17 percentages within peanut-reactive Th cells, which was significant for Th2A as tested by two-way RM ANOVA.Gene expression analyses within the ST2 + (Th2A marker) and CCR6 + (Th17 marker) peanut-reactive T-cell subsets confirmed that ST2 + cells expressed Th2 related genes (IL4, IL5, IL13, IL31, IL17RB, IL1RL1(ST2)) while CCR6 + cells expressed Th17 related genes (IL17A, IL17F, RORC, CCR6) and also FOXP3 and TGFβ which are expressed by Th17-subsets with regulatory properties (Figure5K).The gene expression profile of these subsets remained consistent over time (FigureS4) supportive of an increase in Th17 or Th17/Treg cells rather than increased CCR6 expression on T cells of another phenotype.The shift between Th2 and Th17 populations within PVX108-treated participants (Figure5J) was also F I G U R E 3 CONSORT flow diagram of Stage 1 and Stage 2. evident from gene expression changes within the total peanutreactive Th-cell population over time (Figure 5L,M), with a gradual decrease in Th2A-related gene expression and increases in Th17-related gene expression (IL17A and RORC by Week 21 then IL17F by Month 18) (Figure 5L).FOXP3 expression also increased from pre-dose to Week 21 for five of the seven participants analysed.There were no consistent changes in IL10 or TGFβ expression (regulatory cytokines) within peanut-reactive Th cells over time.However, TGFβ expression was detected in all samples and timepoints, while IL10 expression was only detected in isolated samples.Further analysis of the full data set of >10,000 genes revealed a combination of genes that were differentially expressed at Week 21 and/or Month 18 following PVX108 treatment (adjusted p < .1)(Figure 5M).These genes broadly clustered into groups with distinct directional and/or kinetic changes (coloured sidebar) and include several genes associated with differentiation of Th17 cells (D2HGDH, 37 PXDC1 38 and PDP2 39 ) or Treg cells (MEF2D, 40 RHEBL1 41 and TFE3 42 ).Peanut-specific antibody responses and skin prick test responses were analysed to look for evidence of downstream effects of T-cell modulation on B cells and mast cells.There were no changes in total IgE, peanut-specific IgE or the ratio of peanut-specific IgG4:IgE in either group (Table

Finally, in contrast to the 11 .
6% withdrawal rate reported for Palforzia, only one (2.2%) of the 46 participants treated with PVX108 was withdrawn due to an AE deemed possibly treatment-related (a Grade 2 syncope event following dose administration).Safety of PVX108 was further supported by the absence of allergen-specific IgE increases following treatment (attributed to the lack of IgE epitopes within PVX108) and the inability of PVX108 to activate peanut-sensitized basophils ex vivo (n = 185).Despite widespread attempts to generate hypo-allergenic peanut-based products, no other approach has shown a comparable inability to activate basophils across an equivalent concentration range as shown for PVX108

F I G U R E 4
Phase 1 safety data in peanut-allergic adults.Peak Expiratory Flow (PEF) over time post-dose as a percentage of pre-dose values for PVX108-treated participants in Stage 2 at dose 1 (A, n = 13) and dose 6 (B, n = 10).Each line represents an individual participant.Blue dots show mean values ± SD.Similar data were obtained following all doses in Stages 1 and 2. PEF did not fall below the threshold for spirometry follow-up (20% of pre-dose) following PVX108 administration in either Stage.Graphical summary of the number and severity of all Treatment Emergent Adverse Events (TEAE) by dose level (Stage 1, C), and dose number (Stage 2, D) in active participants.F I G U R E 5 Changes in peanut-reactive memory CD4 + T cells after PVX108 treatment.(A) Gating of memory (CD45RA − ) CD4 + T cells after (Ara h 1 + 2) stimulation.Peanut-reactive (CD154 + ) effector Th cells were subdivided into distinct phenotypes based on CXCR3, CCR4, ST2 and CCR6 expression.Peanut-reactive Treg were identified as CD137 + GARP + LAP + .Dashed boxes show populations sorted for RNAsequencing.(B-I) Graphs show frequency or percentage of peanut-reactive T-cell populations over time within (line graphs) and between (box plots) groups (active: blue, n = 10; placebo: yellow, n = 5).Y-axis shows square root scale.Boxes indicate median values with interquartile range.Red dots indicate mean values.Lines represent individual participants.Differences within and between groups were assessed by two-way RM-ANOVA (time × group interaction p = .031for Th2A and p = .052for Th17).(J) Pie charts show mean percentage of Th subsets within peanut-reactive Th cells over time by treatment.(K-M) Heatmap represents gene expression within sorted Th populations (see A).
sponses to peanut were inconclusive.While the timing and magnitude of the IgG4 increases would align with T cell-mediated modulation of B-cell responses to low level peanut exposure such as SPT testing or micro levels in the environment, these parameters need reassessing in future trials.While our Phase 1 design was sufficient to assess initial safety of PVX108, the relatively short treatment course, low sample numbers, and unequal group sizes between placebo and active, limited interpretation of comparisons of immune changes between groups, especially where variability within the placebo group was high and pre-dose values differed between groups (Th2A percentages).The lack of oral food challenge-proven allergy presented a limitation in participant characterization in this trial (excluded per local ethics request).However, inclusion of confirmed basophil activation to peanut in addition to other parameters for confirming participant eligibility provided additional validation of allergic status.Phase 2 trials of PVX108 have been designed to help overcome these limitations for future investigations.Despite these limitations, the data obtained from this study align with the intended design and activity of PVX108.These data underpinned regulatory approval to commence ongoing Phase 2 trials to assess PVX108 efficacy in peanut-allergic children, and collectively suggest that peptide immunotherapy has potential to overcome long-standing issues of safety and convenience associated with conventional AIT.AUTH O R CO NTR I B UTI O N S Conceptualization and funding acquisition: SRP, PH, ROH, EW.Design Clinical Trial: SRP, PH, ROH, MH, WS, JB, JR, ML, KN, GL.Design immunological studies: SRP, EW, JR, ROH, PH, KN, GL, ML, KN, GL, AV.Experimental work and data acquisition: JA, TP, SK, HDB.Clinical oversight: MH, WS, JB, ROH, SP, PH.Data analysis collation and interpretation: AV, SRP, EW, MM, PH, JB, ML, SK.Supervision and project administration: SRP, PH.Writing-original draft: AV, SRP.Writing-critical review & editing: ROH, JR, MM, ML, EW, JB, PH, HDB, GL.Writing-review and approval: All authors.

Table 2 and
TableS2A).Mild/moderate asthma was reported in four participants (2 PVX108-treated, 2 placebo).One mild asthma event ered unrelated to treatment.All 18 participants reported at least one AE.Of the 150 AEs reported, 146 events were TEAEs (Table