Seefried 2025 — IgG autoantibodies in 184-FM-patient cohort (replication & extension)

One-paragraph summary

A 2025 study assessed the frequency of pathogenic IgG autoantibodies in fibromyalgia patient sera at scale: 184 FM patients vs. matched controls. Sera from 68/184 (37%) FM patients showed binding to dorsal root ganglia tissue, vs. 0/controls. Importantly, the binding patterns clustered into nine distinct profiles — including binding to neurons and to satellite glial cells — suggesting heterogeneity within the autoimmune subtype rather than a single antigen. This is the largest replication of the Goebel 2021 paradigm to date and provides the first quantitative prevalence estimate for the autoimmune subtype.

Claims as triples

• IgG_fm — present_in → dorsal_root_ganglion [evidence: 68/184 FM sera; 0/control sera; confidence: emerging]
• IgG_fm — present_in → satellite_glia [evidence: 9 binding clusters incl. SGCs; confidence: emerging]
• autoantibody_mediated_pain — present_in → fm_autoimmune [evidence: 37% prevalence in 184-patient cohort; confidence: emerging]

Methods note

Patient sera (n=184 FM, controls per the study) screened against rodent (likely murine) DRG tissue using immunohistochemistry. Binding profiles classified into 9 clusters by pattern recognition. Authors: Seefried (Würzburg, Germany) + Karolinska affiliates (Bardic, Grijalva Yepez, et al). Methodological details require verification against the primary source.

Limitations

• Citation source is a forum reference, not a verified primary publication. This paper was discovered via a Science for ME (s4me.info) discussion thread linking to it — the actual journal/preprint citation has not been independently confirmed by the project. Treat the 68/184 figure and the 9-cluster classification as provisional until the primary source is located. Action: locate the primary citation before relying on this for tier promotion.
• Prevalence (37%) is for "binding to DRG tissue" — not the same as "pathogenic in passive transfer." The Goebel 2021 result links binding to pathology, but this study by itself doesn't repeat passive transfer.
• 9 binding clusters suggests antigen heterogeneity — not a single FM autoantigen, but a family of related antigens. This complicates the "identify the antigen" path (Q1).

Open questions raised

• What are the 9 distinct binding clusters? Antigen identification per cluster would generate 9 candidate diagnostic markers and 9 candidate therapeutic targets.
• Do the 68 antibody-positive patients differ clinically from the 116 antibody-negative ones? (Subtype validation.)
• Is "antibody-positive" stable across time or does it fluctuate? (Trait vs. state.)
• Does each cluster have a different clinical phenotype? (Subtype-of-subtype.)

Triangulation notes

• This is the prevalence anchor for fm_autoimmune. The 37% figure suggests the autoimmune subtype is large enough to have major clinical and trial-design implications.
• 9 binding clusters is consistent with the heterogeneity flagged in synthesis/subtypes.md — the autoimmune subtype itself may be a 9-fold polymorphism.
• Combined with Goebel 2025 (mechanism review) and Hanani 2026 (multi-ganglia activation), this completes a three-paper anchor for the autoimmune thread.

Bridges

• B3 (CRPS ↔ FM via pathogenic IgG) — analogous binding-profile clustering in CRPS literature would be a productive comparison.