Autoimmune Disease Research Hypotheses

Seven testable predictions for optimizing treatment selection, predicting drug response, and understanding cross-disease connections in autoimmunity.

300-500M
People affected globally
7
Testable hypotheses

Overview

Cross-system analysis of autoimmune biology has generated several testable hypotheses about drug response, disease connections, and treatment optimization. Approximately 30% of autoimmune patients fail first-line therapy—better prediction of response could transform care for hundreds of millions.

The Challenge

Autoimmune diseases share common mechanisms yet are treated as separate conditions. Patterns in drug response across diseases may reveal opportunities for optimization that single-disease studies miss.

The Hypotheses

1

Cross-Biologic Response Prediction

Observation: Etanercept, bimekizumab, and anifrolumab show correlated response patterns despite targeting different pathways (TNF, IL-17, Type I IFN).
Prediction: Patients who achieve excellent response to one of these drugs are more likely to respond to the others if switching is needed.
Testable: In biologic-switch cohorts, correlate prior drug response with subsequent drug response across these three agents.
2

PTPN22 as Universal Predictor

Observation: PTPN22 variants affect risk across multiple autoimmune diseases.
Prediction: PTPN22 genotype may predict response to immunomodulatory therapies across diseases.
Testable: Correlate PTPN22 R620W status with treatment response in RA, lupus, and T1D cohorts.
3

Interleukin Pathway Convergence

Observation: All interleukins share fundamental signaling characteristics.
Prediction: Blocking one interleukin pathway may partially compensate effects of blocking another.
Testable: In patients failing IL-17 inhibitors, test whether IL-23 or IL-12 blockade shows reduced efficacy compared to biologic-naive patients.
4

JAK Inhibitor Selectivity Clustering

Observation: Different JAK inhibitors show distinct response patterns despite similar mechanisms.
Prediction: Filgotinib (selective JAK1) may show a distinct responder population from tofacitinib (JAK1/3).
Testable: Compare responder characteristics between selective and non-selective JAK inhibitors.
5

B-Cell Depletion Hierarchy

Observation: Different anti-CD20 antibodies show varying potency.
Prediction: Obinutuzumab may show superior efficacy in autoimmune disease compared to rituximab.
Testable: Head-to-head trials in lupus nephritis or ANCA vasculitis.
6

Cross-Disease Drug Repurposing

Observation: Drugs effective in one autoimmune disease often work in others.
Prediction: Anifrolumab (approved for lupus) may show efficacy in Sjögren's syndrome and dermatomyositis.
Testable: Clinical trials of anifrolumab in interferon-driven autoimmune diseases.
7

Sequential Biologic Optimization

Observation: Certain biologics may share response predictors.
Prediction: The optimal sequence of biologics can be predicted by baseline characteristics.
Testable: Develop and validate a biomarker panel predicting optimal first-line biologic.

Research Priority Matrix

Hypothesis Data Required Feasibility Impact
H1: Cross-biologic prediction Registry data High High
H2: PTPN22 predictor Pharmacogenomics Moderate Very High
H3: IL convergence Switch studies Moderate Moderate
H4: JAK clustering Trial comparison High Moderate
H5: B-cell hierarchy Head-to-head trials Low High
H6: Drug repurposing Phase 2 trials Moderate High
H7: Sequential optimization Registry + biomarkers Moderate Very High

Potential Impact

300-500 million people have autoimmune diseases. Approximately 30% fail first-line therapy.

If these hypotheses improve treatment selection by 10%:

  • 30-50 million patients with better outcomes
  • Reduced time to effective therapy
  • $10+ billion in reduced healthcare costs
Related: Microbiome & Autoimmunity

Dysbiosis, barrier dysfunction, and reduced regulatory T cells represent the gut-autoimmune axis. See our Microbiome & Immunity hypotheses for testable predictions on butyrate-Treg induction, early-life antibiotics, and fiber interventions.

Collaboration Invitation

We seek research partnerships with:

  • Biologic registries (CORRONA, RABBIT, etc.)
  • Pharmacogenomics databases
  • Clinical trial networks
  • Academic rheumatology centers