Skip to content

IgA Nephropathy Second-Pass Review

Date: 2026-04-14

Scope

Second-pass review of the existing IgA_Nephropathy disorder entry with explicit lump/split assessment against IgA vasculitis.

Initial audit of the pre-existing entry

Before revision, the live YAML had several issues relative to the requested bar:

  • The pathograph had only three nodes and no explicit downstream edges.
  • The entry did not represent the IgA vasculitis boundary anywhere in disease content.
  • The MONDO anchor was present in disease_term but not mirrored in mappings.
  • The genetics section was too thin and included a likely misleading claim that the CFHR1/CFHR3 deletion is a risk factor; current literature instead treats the common deletion as protective in IgAN.
  • Evidence typing was inconsistent with study type; many mechanistic statements were supported only by reviews but were not explicitly treated that way in the overall framing.

MONDO / Monarch cross-check

Repo-local OAK check plus current external MONDO/Monarch-facing resources were aligned on the key disease identifiers:

  • MONDO:0005342 has label IgA glomerulonephritis with exact synonyms including IgA nephropathy and Berger's disease.
  • MONDO:0019167 has label immunoglobulin A vasculitis with exact synonyms including IgA vasculitis and Henoch-Schonlein purpura.

Implication for curation:

  • The current ontology does not collapse IgAN into IgA vasculitis or vice versa.
  • A subtype fold-in of IgA vasculitis under IgA nephropathy would work against the existing MONDO split and against common clinical naming.

Useful links:

  • MONDO IgA glomerulonephritis: https://bioportal.bioontology.org/ontologies/MONDO/?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FMONDO_0005342
  • MONDO immunoglobulin A vasculitis: https://bioportal.bioontology.org/ontologies/MONDO/?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FMONDO_0019167

KDIGO / CureGN / literature consistency check

Three external framing sources were especially important:

  1. KDIGO 2025 IgAN/IgAV guideline
  2. KDIGO publishes a combined guideline for IgAN/IgAV, which acknowledges deep mechanistic overlap.
  3. At the same time, KDIGO's evaluation chapter for IgAN still instructs curators/clinicians to consider secondary causes including IgA vasculitis.

  4. This supports a "jointly discussed but not ontologically collapsed" view.

  5. CureGN

  6. CureGN operationalizes IgAN and IgA vasculitis nephritis together in the same research program but still names them as distinct study diseases/populations.

  7. This is consistent with a shared-mechanism, separate-diagnosis approach.

  8. Pillebout review (PMID:40069065)

  9. This review states that IgA vasculitis can be considered a systemic form of IgAN, that the two diseases share the four-hit model, and that kidney biopsy can be indistinguishable.
  10. The same abstract also states that chronic lesions are more frequent in IgAN, proliferative lesions are more frequent in IgAVN, and the characteristic rash of IgAVN drives earlier diagnosis.
  11. That combination argues against treating IgA vasculitis as a mere subtype of IgAN inside a kidney-focused disorder file.

Useful links:

  • KDIGO guideline PDF: https://kdigo.org/wp-content/uploads/2024/08/KDIGO-2025-IgAN-IgAV-Guideline.pdf
  • CureGN about page: https://www.curegn.org/about
  • CureGN study summary: https://www.curegn.org/studies/study-participation

ClinGen cross-check

ClinGen was checked specifically to avoid drifting into false monogenic framing.

Findings:

  • No dedicated ClinGen gene-disease validity curation surfaced for IgA nephropathy or immunoglobulin A vasculitis as primary monogenic disease entities.
  • The disease does appear in genetics literature and in pathway-oriented discussion, but the ClinGen landscape did not justify CAUSATIVE framing for any single gene in this disorder entry.

Implication for curation:

  • Keep pathway mediators like TNFSF13 and TNFSF13B on mechanism nodes.
  • Keep the genetic section limited to susceptibility/protective loci and explicitly polygenic language.
  • Do not narrate IgAN as a monogenic disease.

ClinGen search entrypoint:

  • https://search.clinicalgenome.org/kb/gene-validity

Lump / split decision

Decision: keep related-but-separate, not subtype structure, and do not introduce a paired root in this PR.

Reasoning:

  • IgA vasculitis is a systemic small-vessel vasculitis in MONDO, not a child of IgA glomerulonephritis.
  • The kidney lesion is strongly overlapping and often histologically indistinguishable, but the systemic phenotype boundary matters clinically and ontologically.
  • A subtype representation would wrongly imply that IgA vasculitis is just a renal subtype of IgAN.
  • A new paired-root umbrella strategy might be defensible in a future broader curation effort, but it would require introducing a new umbrella disease concept and matching ontology strategy. That is out of scope for a second-pass fix of the existing IgAN file.

Practical encoding in the YAML:

  • The main disease remains anchored to MONDO:0005342.
  • IgA vasculitis is represented in differential_diagnoses with MONDO:0019167.
  • The top-level notes make the split explicit.

Mechanism curation choices

The revised pathograph was rebuilt around a seven-step kidney mechanism chain:

  1. Mucosal immune dysregulation and BAFF/APRIL-driven B-cell activation
  2. Galactose-deficient IgA1 overproduction
  3. Anti-Gd-IgA1 autoantibody formation and immune complex assembly
  4. Mesangial immune complex deposition and complement activation
  5. Mesangial proliferation and inflammatory amplification
  6. Podocyte damage and filtration barrier failure
  7. Chronic complement-linked progression and kidney failure risk

To make the graph fully connected and browser/demo-safe without inventing shortcut biology:

  • the renal phenotypes were hung off the kidney injury chain rather than left as isolated symptom cards
  • the treatment nodes were linked to explicit target phenotypes or mechanisms
  • the MHC, glycosylation, and complement-locus susceptibility nodes were aligned to matching mechanism steps
  • the old standalone environmental node was dropped because the current graph extractor cannot represent an upstream exposure cleanly without creating an orphaned node
  • the protective complement signal was encoded as a CFH-anchored protective locus with notes about the CFHR1-CFHR3 deletion-tagging haplotype, which is more honest than treating the deletion itself as a standalone gene node

This avoids:

  • dangling nodes
  • direct "IgA deposits -> ESRD" shortcuts
  • monogenic overstatement
  • dishonest evidence-source tagging

Key references used for the second pass

  • PMID:38362118 BAFF/APRIL and mucosal hyper-responsiveness
  • PMID:39095059 Gd-IgA1 glycosylation and immune complexes
  • PMID:22904352 anti-Gd-IgA1 autoantibodies in human disease
  • PMID:39188719 mesangial proliferation, podocyte damage, and progression framing
  • PMID:38053977 complement pathway framing
  • PMID:38182298 mesangial C3 deposition and alternative/lectin pathway context
  • PMID:35781866 C3 deposition and severity/progression
  • PMID:39003309 complement-pathic prognosis subset
  • PMID:38438966 biopsy-level IgA deposits / hypercellularity / C3 deposits
  • PMID:40069065 IgAN vs IgAV split boundary
  • PMID:21399633 polygenic susceptibility architecture