# Preventing the Atopic March and the Initiation of EoE — Literature Synthesis & Prophylactic Strategy

*Turning the pMHC-anergy platform toward prevention: can a short, antigen-directed intervention in an at-risk, pre-symptomatic child delay or block progression toward EoE — the way teplizumab delays type 1 diabetes? A literature-grounded synthesis. Hypothesis-generating, not clinical guidance (see disclaimer).*

---

## 1. The opportunity: EoE risk is familial but environment-dominated

EoE clusters in families, which is what makes a prevention program conceivable — you can identify at-risk children before disease. The definitive quantification is **Alexander et al. 2014** (*JACI*, doi:10.1016/j.jaci.2014.07.021), a 914-proband nuclear-family cohort (2,192 first-degree relatives) plus a monozygotic/dizygotic twin registry:

- EoE in **first-degree relatives = 1.8% (unadjusted), 2.3% (sex-adjusted)**; risk in **siblings = 2.4%** — i.e. roughly the **1:40** figure. (Brothers and males carried the highest recurrence risk ratios.)
- **Recurrence risk ratios 10–64×** over the general population (general-population prevalence ~0.05%).
- **Combined gene + shared-environment heritability = 72.0%**, but the paper's central conclusion is that **environmental cues dominate over genetic ones** (twin concordance patterns implicate strong shared-environment effects).

The strategic read: **familial risk is high enough to define an enrichable at-risk cohort, and the environment-dominated architecture means the disease is modifiable** — the two preconditions for a prevention program. HLA type and the atopic-march trajectory (below) further stratify which children are climbing toward EoE.

## 2. The atopic march is a staged, interceptable trajectory

Children at risk typically progress: **skin-barrier dysfunction / atopic dermatitis (infancy) → food sensitization (1–3 yr) → allergic rhinitis + asthma (preschool–school age) → EoE (childhood/adult)**. Each transition is a candidate interception point (Sicherer & Sampson 2017, doi:10.1016/j.jaci.2017.11.003; Tsakok 2016 on AD→food allergy, doi:10.1016/j.jaci.2015.10.049; Brough 2020 epicutaneous sensitization, doi:10.1111/all.14304). The mechanistic through-line is **epicutaneous sensitization through a breached skin barrier priming type-2 immunity**, later manifesting at mucosal surfaces — so early barrier and antigen-exposure decisions shape the whole march.

## 3. Prevention already works in allergy — three proven paradigms

Prophylaxis of atopic disease is **not speculative**; three interventions already prevent downstream disease and set the precedent for an EoE program:

1. **Early oral antigen introduction (tolerance by exposure).** **LEAP** (Du Toit 2015, *NEJM*, doi:10.1056/nejmoa1414850) — early peanut introduction in high-risk infants cut peanut allergy by ~80%. **EAT** (Perkin 2016, *NEJM*, doi:10.1056/nejmoa1514210) — early introduction of multiple allergens reduced food allergy in per-protocol analysis. Establishes that **timing and route of first antigen encounter determine tolerance vs sensitization** (oral/tolerogenic vs epicutaneous/sensitizing — the "dual-allergen-exposure" model).
2. **Allergen immunotherapy (AIT) prevents progression, not just symptoms.** Pajno 2001 (doi:10.1046/j.1365-2222.2001.01161.x) and the EAACI prevention guidance (Roberts 2017, doi:10.1111/all.13317) show AIT in monosensitized children **prevents new sensitizations and reduces asthma onset**. This is the closest allergy analogue to disease-modifying prevention.
3. **Skin-barrier support.** Simpson 2014 (doi:10.1016/j.jaci.2014.08.005) — prophylactic emollient from birth reduced AD incidence in high-risk infants, plausibly damping the epicutaneous-sensitization step upstream of the whole march.

## 4. The teplizumab lesson (Type 1 diabetes) — the disease-modifying template

The strongest proof that **a brief immune-directed course in a pre-symptomatic at-risk person delays disease onset** comes from T1D:

- **Herold et al. 2019** (*NEJM* TN-10 trial, doi:10.1056/nejmoa1902226): a **single 14-day course of teplizumab (anti-CD3)** in autoantibody-positive, still-normoglycemic relatives **delayed clinical T1D onset by a median ~2 years** — leading to the first FDA approval of a drug that *delays* an autoimmune disease.
- **Mechanism** (Lledó-Delgado 2024, doi:10.1172/jci177492; Herold 2024 review, doi:10.1038/s41577-023-00985-4): anti-CD3 delivers a **partial-agonist TCR signal without full costimulation**, driving autoantigen-specific effector T cells into a **hyporesponsive / exhausted / anergic** state and shifting the regulatory balance — persistently.

**Why this transfers to EoE.** The pMHC-anergy therapeutic we prioritized uses the *same immunological principle* — engage the antigen-specific TCR to induce anergy rather than activation — but with a crucial advantage: **antigen-specificity**. Teplizumab is a pan-T-cell anti-CD3 (systemic, with a cytokine-release and transient-immunosuppression liability); a **food-peptide-loaded pMHC reagent targets only the food-reactive clones**, sparing the rest of the repertoire. The T1D staging framework (Sims 2022 general-population screening, doi:10.2337/dbi20-0054; Primavera 2020, doi:10.3389/fendo.2020.00248) — genetic + autoantibody staging to select who gets pre-symptomatic therapy — is the operational model for staging atopic-march children by HLA + sensitization status.

## 5. Mechanistic substrate for an EoE prophylactic

Two tolerance mechanisms are available to a prophylactic, both antigen-specific:
- **Oral / mucosal tolerance** (Pabst 2012, doi:10.1038/mi.2012.4; Chehade 2005, doi:10.1016/j.jaci.2004.11.008): the gut's default response to fed protein is Treg-mediated tolerance; the LEAP/EAT result is this pathway used prophylactically. An EoE program could **combine guided early oral introduction with monitoring** in high-risk infants.
- **Engineered antigen-specific anergy/tolerance** (Passerini 2020, doi:10.3389/fimmu.2020.02194): pMHC multimers, tolerogenic antigen-presentation, or Treg-directed approaches that delete/anergize a defined specificity — the teplizumab principle made antigen-specific. This is where our platform's personalized peptide engine plugs in.

## 6. A staged, risk-matched prevention strategy for EoE

Mapping interventions to the march (see figure `atopic_march_prevention.png`), from least to most targeted:

| Window | At-risk stage | Intervention | Precedent |
|---|---|---|---|
| ① Infancy | family history + AD / barrier defect | skin-barrier support + **guided early oral introduction** of key allergens | Simpson 2014; LEAP; EAT |
| ② Toddler–preschool | early sensitization, climbing march | **allergen immunotherapy** to block new sensitization & asthma onset | Pajno 2001; EAACI 2017 |
| ③ Pre-EoE, high-risk | HLA-permissive + multi-food sensitized + strong family history | **antigen-specific pMHC anergy** — teplizumab-style, but food-peptide-targeted | Herold 2019 (principle) |

**Risk stratification to select who gets window ③** (the most interventional): the same tools this project already built — **HLA class-II typing** (which food peptides a child can present; e.g. DR7-casein), **personalized epitope burden**, **the eosinophil/allergic GWAS load**, and **atopic-march stage**. A child with high-presentation HLA + heavy eosinophil-axis genetics + advancing march is the analogue of the autoantibody-positive T1D relative — the pre-symptomatic individual in whom a brief, antigen-specific course is justified.

## 7. Why an antigen-specific prophylactic could beat teplizumab's profile

- **Specificity → safety.** Targeting only food-reactive clones avoids the global T-cell modulation, cytokine release, and transient immunosuppression of anti-CD3 — a far more favorable risk-benefit for dosing *healthy at-risk children*, where the safety bar is highest.
- **The diagnostic and the prophylactic share one engine.** The same personalized peptide-pool platform that identifies a child's presentable food antigens (diagnostic arm) defines the tolerizing reagent (prophylactic arm) — a theranostic that stratifies risk and delivers the intervention.
- **A measurable surrogate endpoint exists.** Food-antigen-specific peTH2 frequency (the diagnostic readout; Hill/Spergel platform) is a candidate pharmacodynamic marker — does a prophylactic course reduce the pathogenic-Th2 clone before EoE manifests? — mirroring how autoantibody/C-peptide track teplizumab's effect.

## 8. Open questions & honest limits

- **When does the "point of no return" pass?** The march's interception windows are inferred from separate trials; the optimal age to intervene for *EoE specifically* is unknown.
- **Ethical bar for dosing healthy children.** Any window-③ intervention in a pre-symptomatic child demands teplizumab-grade evidence of favorable risk-benefit and a validated high-risk-selection rule; over-treatment of children who would never progress is the central risk.
- **EoE-specific prevention data are essentially absent.** The paradigms above are borrowed from peanut allergy, asthma, AD, and T1D; no trial has yet shown prevention of *EoE onset*. This synthesis argues the hypothesis is well-motivated, not that it is proven.
- **Environment-dominated heritability cuts both ways:** modifiable, but the causal environmental exposures (diet timing, microbiome, barrier) are not yet pinned down for EoE (Çelebi Sözener 2022 epithelial-barrier hypothesis, doi:10.1111/all.15240).

## 9. Disclaimer

Research/hackathon synthesis, not medical advice and not a clinical protocol. No prophylactic described here is validated for preventing EoE; dosing pre-symptomatic children would require formal trials and regulatory review. Familial-risk figures are from Alexander 2014 as cited. Decisions for any child must be made by qualified clinicians.

---

## Key references (DOI-verified, retrieved this session)
- Alexander ES et al. 2014, *JACI* — twin/family study; EoE heritability, familial risk (FDR 1.8–2.3%, sibling 2.4%, RRR 10–64×, h²gc 72%). doi:10.1016/j.jaci.2014.07.021
- Herold KC et al. 2019, *NEJM* — teplizumab delays clinical T1D in at-risk relatives (TN-10). doi:10.1056/nejmoa1902226
- Lledó-Delgado A et al. 2024, *J Clin Invest* — persistent antigen-specific T-cell changes after teplizumab. doi:10.1172/jci177492
- Herold KC et al. 2024, *Nat Rev Immunol* — the immunology of type 1 diabetes. doi:10.1038/s41577-023-00985-4
- Sims EK et al. 2022, *Diabetes* — screening for T1D in the general population (staging). doi:10.2337/dbi20-0054
- Du Toit G et al. 2015, *NEJM* — LEAP: early peanut introduction prevents peanut allergy. doi:10.1056/nejmoa1414850
- Perkin MR et al. 2016, *NEJM* — EAT: early allergenic-food introduction. doi:10.1056/nejmoa1514210
- Pajno GB et al. 2001, *Clin Exp Allergy* — AIT prevents new sensitizations in asthmatic children. doi:10.1046/j.1365-2222.2001.01161.x
- Roberts G et al. 2017, *Allergy* — EAACI allergen-immunotherapy guidelines (prevention). doi:10.1111/all.13317
- Simpson EL et al. 2014, *JACI* — emollient skin-barrier protection reduces AD incidence. doi:10.1016/j.jaci.2014.08.005
- Pabst O & Mowat AM 2012, *Mucosal Immunol* — oral tolerance to food protein. doi:10.1038/mi.2012.4
- Passerini L & Bacchetta R 2020, *Front Immunol* — antigen-specific tolerance induction in T-cell-mediated disease. doi:10.3389/fimmu.2020.02194
- Sicherer SH & Sampson HA 2017, *JACI* — food allergy epidemiology/pathogenesis (atopic march). doi:10.1016/j.jaci.2017.11.003
- Tsakok T et al. 2016, *JACI* — does atopic dermatitis cause food allergy? doi:10.1016/j.jaci.2015.10.049
- Brough HA et al. 2020, *Allergy* — epicutaneous sensitization in food-allergy development. doi:10.1111/all.14304
- Çelebi Sözener Z et al. 2022, *Allergy* — epithelial-barrier hypothesis. doi:10.1111/all.15240