Skip to content

Autoimmunity and inflammation

TL;DR — The strongest single piece of evidence for an immune mechanism in fibromyalgia (FM) is passive transfer: purified IgG from FM patients, injected into mice, produced mechanical and cold hypersensitivity, reduced grip strength, reduced locomotion and loss of intraepidermal innervation, while IgG-depleted patient serum and healthy-control IgG did nothing (Goebel 2021, PMID 34196305). Patient IgG binds satellite glial cells (SGCs) in the dorsal root ganglion, and clinical studies in two cohorts found anti-SGC IgG elevated in FM and associated with more severe symptoms (Krock 2023, PMID 36943275). The phenotype has been re-demonstrated, including in male mice, and extended mechanistically to Aβ low-threshold mechanoreceptors (Berwick 2025, PMID 40408228; Israel 2025, PMID 40898777) — but every passive-transfer paper located in this session shares authors with the original, so there is no fully independent replication. Circulating cytokine differences are real but small and mixed in direction: pooled TNF-α SMD 0.36, IL-6 0.15, IL-8 0.26 and anti-inflammatory IL-10 0.61 (O'Mahony 2021, PMID 33576773). FM is not classified as autoimmune: no defined autoantigen, no HLA association established, no immunotherapy RCT in FM, and the anti-SGC signal is not FM-specific (it appears after COVID-19 recovery) (Berwick 2025, PMID 40408228). The question is a live published debate, not a settled position (Clauw & Sarzi-Puttini 2024, PMID 37634681).

1. The passive-transfer experiments

1.1 The founding study (Goebel 2021, J Clin Invest, PMID 34196305)

Methods. IgG was purified from FM patients and from healthy controls and administered to C57BL/6 mice. Comparators included IgG-depleted patient serum and healthy-control IgG. Readouts spanned behaviour (mechanical and cold sensitivity, locomotor activity, paw grip strength), ex-vivo skin–nerve preparation electrophysiology, skin innervation, and immunolabelling of mouse and human dorsal root ganglion (DRG) tissue.

Findings.

Domain Result
Behaviour Increased sensitivity to noxious mechanical and cold stimulation; reduced locomotor activity; reduced paw grip strength
Peripheral electrophysiology Nociceptive fibres in skin–nerve preparations from FM-IgG-treated mice showed increased responsiveness to cold and mechanical stimulation
Innervation Loss of intraepidermal innervation in FM-IgG-treated mice
Controls IgG-depleted FM serum and healthy-control IgG had no effect
Mechanism Patient IgG did not activate naive sensory neurons directly; it labelled satellite glial cells and neurons in vivo and in vitro, plus myelinated fibre tracts and a small number of macrophages and endothelial cells in mouse DRG — and no cells in the spinal cord. FM IgG also bound human DRG

Why it matters. It satisfies one classical criterion for an antibody-mediated disease (transfer of phenotype with the antibody fraction), it locates the target in the peripheral sensory ganglion rather than the CNS, and it offers a mechanistic explanation for the small-fibre pathology described on pathophysiology-peripheral — denervation as a consequence of a circulating factor.

What it does not show. It is a mouse model of hypersensitivity, not of fibromyalgia; the "fatigue and emotional distress" of the human syndrome are not modelled. The antigen is unidentified. Patient IgG was pooled or used per-patient at high doses over days; the relationship to endogenous antibody titres over years of illness is unknown.

1.2 Replication status — stated honestly

PubMed searching in this session returned four passive-transfer papers in FM, and the author lists overlap substantially:

Paper What it adds Independence
Goebel 2021, PMID 34196305 Original demonstration (Liverpool / Karolinska / King's College London) —
Berwick 2025, PMID 40408228 "We validated FMS-IgG passive transfer": pooled serum from 18 patients per group, IgG 8 mg/day for 3 days, male C57BL/6J mice (n = 6/group). FM-IgG reproduced mechanical and cold hypersensitivity and reduced grip strength in male rodents. Pooled post-acute-COVID-19-syndrome (PACS) IgG was inert Adds a Stanford group (Sahbaie, Guo, Clark) but retains Goebel and Andersson as authors
Israel 2025 (Brain), PMID 40898777 Causal link between circulating FM IgG and sensitization of Aβ low-threshold mechanoreceptors to mechanical and cold stimuli, ex vivo and in vivo; a larger proportion of Aβ-LTMRs responded to cold in people with FM, though (contrary to the mouse result) these fibres showed reduced mechanical responses in humans Same King's College London / Liverpool network
Israel 2026 (J Physiol), PMID 41955337 Retrospective analysis of skin–saphenous nerve recordings: C-mechano and Aδ-mechano afferents from FM-IgG-treated mice fired significantly more after — but not during — mechanical force steps, i.e. after-discharge mirroring patients' lingering sensations Same network

Statement for the record: the passive-transfer phenotype has been reproduced across sexes, laboratories within the consortium, and readouts, but no replication by a group fully independent of the original investigators was located in this session's searches. This is the single most important caveat attached to the autoimmune hypothesis.

2. Autoantibody candidates

2.1 Anti-satellite glial cell (anti-SGC) IgG

Study Design Result
Krock 2023, PMID 36943275 Two cohorts (Sweden, Canada); indirect immunofluorescence on murine SGC cultures; separate assessment of binding to human DRG sections Anti-SGC IgG elevated in FM vs controls in both cohorts. Higher levels associated with higher self-reported pain in both, and with higher Fibromyalgia Impact Questionnaire scores and increased pressure sensitivity in the Swedish cohort. Not associated with disease duration. A "FM-severe" cluster had higher anti-SGC IgG than FM-mild and controls, and higher binding to human SGCs
Fanton 2023, PMID 37683961 Larger FM cohort; ¹H-MRS of thalamus and rostral ACC; fMRI with calibrated pressure pain Confirmed the correlation with clinical severity; high-antibody patients had higher pain intensity and worse disease status. Anti-SGC IgG correlated negatively with scyllo-inositol in thalamus and rACC and with total choline and macromolecule 12 in thalamus. Critically, anti-SGC IgG was not associated with pressure-pain sensitivity or with cerebral processing of evoked pressure pain → the authors infer relevance to spontaneous rather than evoked pain
Jakobsson 2025, PMID 39922554 35 female FM, 33 age/BMI-balanced controls; serum lipidomics FM subjects with higher anti-SGC IgG reported more intense pain. 63 lipids differed between FM and controls or between severe and mild FM; LPC 22:0 and LPC 24:1 correlated positively with both pain intensity and anti-SGC IgG level
Jakobsson 2026, PMID 41735414 35 FM, 32 matched controls; bile acids and short-chain fatty acids Non-conjugated (microbially produced, secondary) bile acids elevated in FM; total bile acids higher in FM subjects with high vs low anti-SGC IgG; specific bile acids associated with severity and poorer mental well-being

The specificity problem. In the PACS study, IgG from patients with post-acute COVID-19 syndrome and from people recently recovered from acute COVID-19 both stained SGC-enriched cultures strongly positive — yet only FM IgG was pronociceptive on transfer (Berwick 2025, PMID 40408228). Anti-SGC reactivity is therefore not a specific marker of FM and does not by itself confer the pain phenotype.

2.2 Other candidates

  • Anti-TS-HDS (trisulfated heparin disaccharide) IgM. In 22 FM patients, 5 were positive; 4 of those had non-length-dependent small-fibre neuropathy (p = 0.0393). But positivity did not differ from a control database (p = 0.41) and the excess of non-length-dependent SFN was a trend only (p = 0.06) (Malik 2019, PMID 30801480). Small, exploratory, and TS-HDS is a marker of dysimmune small-fibre neuropathy generally rather than FM.
  • Dysimmune markers in FM patients with biopsy-proven SFPN. Among FM patients whose skin biopsies were diagnostic for small-fibre polyneuropathy, 8 had dysimmune markers and 2 had hepatitis C serologies on aetiological screening (Oaklander 2013, PMID 23748113).
  • No defined FM autoantigen has been identified. The Goebel work localizes binding to SGCs and neurons without naming the target (PMID 34196305).

3. Cytokines: what actually differs

Two meta-analyses, a decade apart, converge on "small, mixed, low-quality".

Meta-analysis Scope Result
Üçeyler 2011, PMID 22034969 25 articles, 1,255 FM patients and 800 healthy controls; 13/25 studies enterable into meta-analysis Overall methodological quality low; studies largely non-comparable in method and target cytokine. Systematic review: higher serum IL-1 receptor antagonist, IL-6 and IL-8, and higher plasma IL-8. Meta-analysis: higher plasma IL-6 (SMD = −0.34 [−0.64, −0.03], p = 0.03, sign as reported). The majority of investigated cytokines did not differ
O'Mahony 2021, PMID 33576773 29 studies (2,458 participants) systematically reviewed; 22 studies (1,772 participants) meta-analysed TNF-α SMD 0.36 (95% CI 0.12–0.60), p = 0.0034, I² = 71%; IL-6 SMD 0.15 (0.003–0.29), p = 0.045, I² = 39%; IL-8 SMD 0.26 (0.05–0.47), p = 0.01, I² = 61%; IL-10 SMD 0.61 (0.34–0.89), p < 0.001, I² = 10%. The chemokine eotaxin was consistently raised across studies in the systematic (not pooled) analysis

Three readings follow. First, the effect sizes are small (0.15–0.36) except IL-10, with heterogeneity up to I² = 71% — nothing here is diagnostic. Second, the profile is not a classical inflammatory signature: the largest pooled effect is for an anti-inflammatory cytokine (IL-10), which the authors describe as a mixed pro- and anti-inflammatory plus chemokine signature. Third, an isolated eotaxin elevation, if it holds, points toward chemokine/eosinophil biology rather than a canonical autoimmune pathway; it has not been independently confirmed as a pooled estimate.

4. Neuroinflammation versus systemic inflammation

These are distinct claims and are supported by different evidence.

  • Neuroinflammation (glial). [¹¹C]PBR28 TSPO PET in 31 FM and 27 controls across two sites showed widespread cortical elevations, most pronounced in medial and lateral frontal and parietal walls; a smaller [¹¹C]-L-deprenyl-D₂ subsample (11 vs 11) showed no astrocytic-signal difference (p ≥ 0.53, uncorrected), implicating microglia (Albrecht 2019, PMID 30223011). Elevated brain TSPO is not FM-specific — it was shown earlier in chronic low back pain, where thalamic TSPO correlated negatively with clinical pain and with circulating IL-6 (Loggia 2015, PMID 25582579).
  • Peripheral glia. The IgG work implicates satellite glial cells of the DRG — peripheral, not central, glia (Goebel 2021, PMID 34196305; Krock 2023, PMID 36943275). The original transfer study found no IgG binding to spinal cord cells.
  • Systemic inflammation. The cytokine meta-analyses above; no consistent acute-phase response, no synovitis, no tissue destruction.

The IASP nociplastic-pain framing accommodates this split, naming "autoreactive antibodies acting at the level of the dorsal root ganglia" alongside aberrant cerebral pain processing as candidate FM mechanisms (Kosek 2024, PMID 39560415); mechanistic reviews similarly place systemic inflammation and peripheral nervous system change on a continuum with central processing changes (Sluka 2016, PMID 27291641).

The three do not currently form a single coherent mechanism. Notably, cerebral TSPO elevation (central) and anti-SGC IgG action (peripheral ganglion) have not been linked in the same patients; the only bridge attempted so far related anti-SGC IgG to brain metabolite concentrations rather than to glial markers, and found no relationship to evoked-pain processing (Fanton 2023, PMID 37683961).

5. Why FM is still not classified as autoimmune — and what would change that

The question has been staged as an explicit published debate between opposing experts, which is itself the honest summary of the field's state (Clauw & Sarzi-Puttini 2024, PMID 37634681). The classification barriers:

  1. No defined autoantigen. Binding to SGCs is demonstrated; the molecular target is not.
  2. No specificity. Anti-SGC reactivity is present after COVID-19 recovery in people without FM (Berwick 2025, PMID 40408228).
  3. No independent replication of transfer. All four passive-transfer papers share authors (§1.2).
  4. No immunotherapy RCT in FM, and the closest adjacent RCT was negative. See §6 — the FM evidence is one small open-label series and expert opinion, while IVIG failed its primary endpoint in idiopathic SFN (Geerts 2021, PMID 33766992).
  5. Cytokine profile is not inflammatory in the classical sense and includes the anti-inflammatory pole (O'Mahony 2021, PMID 33576773).
  6. No established HLA or immunogenetic association was located in this session's searches.
  7. Only a subset shows the candidate marker, and the marker tracks severity rather than diagnosis (Krock 2023, PMID 36943275; Fanton 2023, PMID 37683961).

What evidence would change the classification:

  • Independent replication of IgG passive transfer by a laboratory with no overlap with the original consortium.
  • Identification of the SGC/neuronal autoantigen and demonstration that affinity-purified anti-antigen IgG (not bulk IgG) transfers the phenotype.
  • A randomized, placebo-controlled immunotherapy trial in FM — antibody-depleting or antibody-blocking — showing benefit, ideally with anti-SGC IgG level as a predictive biomarker.
  • Demonstration that anti-SGC IgG distinguishes FM from other chronic pain and post-infectious states, which it currently does not.
  • Longitudinal data showing antibody titres precede or track symptom onset and remission (current data show no association with disease duration; Krock 2023, PMID 36943275).

6. Therapeutic implications being tested

Approach Status Evidence
IVIG in FM Not established; one small open-label series 58 FM patients screened; electrodiagnostic findings suggestive of chronic inflammatory demyelinating polyneuropathy in 33%; 15 of these treated open-label with IVIg 400 mg/kg/day × 5 days, with improvement in pain (p = 0.01), tenderness (p = 0.001) and strength (p = 0.04), and non-significant trends for fatigue and stiffness (Caro 2008, PMID 18208823). Uncontrolled, unblinded, single-centre, selected subgroup
IgG for chronic pain generally Promising per expert workshop; RCTs called for Liverpool workshop concluded IgG may reduce pain across several nociceptive and neuropathic conditions including FM, is relatively safe but costly, and that randomized controlled trials and predictive tests are needed (Tamburin 2014, PMID 24422915)
Counter-position Explicitly negative A critical review argues IVIG has not been effective for chronic pain and warns against extrapolating autoimmune framing to syndromes lacking objective markers — written primarily about complex regional pain syndrome but naming FM (Chang 2019, PMID 30639650)
IVIG in small-fibre neuropathy (adjacent field) Reported, and negative The 60-patient randomized, double-blind, placebo-controlled trial of IVIG in idiopathic SFN (protocol: de Greef 2016, PMID 27439408; NCT02637700) reported in 2021. Thirty patients received IVIG and 30 placebo; the primary outcome (≥1-point drop in mean weekly peak Pain Intensity NRS at 12 weeks) was reached by 40% vs 30% (p = 0.588, OR 1.56, 95% CI 0.53–4.53), with no significant difference on any prespecified secondary outcome — rated Class I evidence that IVIG does not reduce pain in painful idiopathic SFN (Geerts 2021, PMID 33766992). Autoantibody-selected SFN trials: NCT03401073 (TS-HDS/FGFR3, completed), NCT04153422 (TS-HDS/FGFR-3/Plexin D1, recruiting), NCT06183008 (aggregated n-of-1, not yet recruiting)
Immunoadsorption / apheresis / rituximab in FM No trials located A ClinicalTrials.gov search for FM crossed with immunoglobulin, immunoadsorption, rituximab, plasma exchange or plasmapheresis returns exactly three records, none of them a test of antibody depletion or B-cell therapy in FM: NCT04381793 (a nutritional "Recovery Factors" product, n = 43, completed), NCT03058224 (IGN-ES001 vs acetaminophen in chronic widespread pain with or without FM, n = 230, completed), NCT04158752 (galcanezumab for trigeminal/glossopharyngeal pain, n = 26, completed). Verified 2026-08-28

Practical inference: the therapeutic corollary of the autoimmune hypothesis — reducing patient IgG titres, as the Goebel paper explicitly suggests (PMID 34196305) — has not yet been tested in a controlled FM trial, and the nearest controlled test in an adjacent condition (IVIG in idiopathic SFN) was negative (Geerts 2021, PMID 33766992). Note that IVIG supplements rather than depletes IgG, so a negative IVIG trial does not test antibody removal; it does remove the most readily available immunotherapy from the "promising" column.

Open questions

  • Will IgG passive transfer replicate outside the originating consortium? All four FM transfer papers located share authors (Goebel 2021, PMID 34196305; Berwick 2025, PMID 40408228; Israel 2025, PMID 40898777; Israel 2026, PMID 41955337).
  • What is the autoantigen? Patient IgG binds SGCs, neurons, myelinated fibre tracts and some macrophages/endothelial cells in DRG, and human DRG, without a named target (PMID 34196305).
  • Why is anti-SGC reactivity present in post-COVID and recovered-COVID sera that are not pronociceptive on transfer? (Berwick 2025, PMID 40408228) — this dissociates the assay signal from the pathogenic activity.
  • Does anti-SGC IgG relate to spontaneous but not evoked pain? Fanton found associations with clinical severity but none with pressure-pain sensitivity or evoked-pain brain processing (PMID 37683961), which conflicts with Krock's Swedish-cohort association with pressure sensitivity (PMID 36943275).
  • Is the eotaxin signal real? It was consistent in the systematic review but not pooled as a meta-analytic estimate (O'Mahony 2021, PMID 33576773) and does not appear in the earlier meta-analysis (Üçeyler 2011, PMID 22034969).
  • Are central glial activation and peripheral SGC autoreactivity the same process? TSPO elevation is cortical (Albrecht 2019, PMID 30223011) while IgG binding is confined to peripheral ganglia with none in spinal cord (PMID 34196305); no study links them in the same patients.
  • Does lowering IgG help? No controlled FM immunotherapy trial exists; the only FM clinical data are a 15-patient open-label IVIg series in a selected CIDP-like subgroup (Caro 2008, PMID 18208823), and the one RCT of IVIG in the adjacent condition of idiopathic small-fibre neuropathy was negative (Geerts 2021, PMID 33766992).

References

  1. Goebel A, Krock E, Gentry C, et al. Passive transfer of fibromyalgia symptoms from patients to mice. J Clin Invest. 2021;131:e144201. PMID 34196305
  2. Krock E, Morado-Urbina CE, Menezes J, et al. Fibromyalgia patients with elevated levels of anti-satellite glia cell immunoglobulin G antibodies present with more severe symptoms. Pain. 2023;164:1828-1840. PMID 36943275
  3. Fanton S, Menezes J, Krock E, et al. Anti-satellite glia cell IgG antibodies in fibromyalgia patients are related to symptom severity and to metabolite concentrations in thalamus and rostral anterior cingulate cortex. Brain Behav Immun. 2023;114:371-382. PMID 37683961
  4. Berwick RJ, Sahbaie P, Kenny G, et al. Postacute COVID-19 syndrome and fibromyalgia syndrome are associated with anti-satellite glial cell IgG serum autoantibodies but only fibromyalgia syndrome serum-IgG is pronociceptive. Pain. 2025;166:e397-e408. PMID 40408228
  5. Israel MR, Berwick R, Vastani N, et al. Aβ low threshold mechanoreceptors contribute to sensory abnormalities in fibromyalgia. Brain. 2025;148:4016-4029. PMID 40898777
  6. Israel MR, Vastani N, Gentry C, Goebel A, Bevan S, Andersson DA. Mechanical sensitization of sensory afferents after passive transfer of fibromyalgia IgG. J Physiol. 2026. PMID 41955337
  7. Jakobsson JE, Menezes J, Krock E, et al. Fibromyalgia patients have altered lipid concentrations associated with disease symptom severity and anti-satellite glial cell IgG antibodies. J Pain. 2025;29:105331. PMID 39922554
  8. Jakobsson JE, Carlsson H, Erngren I, et al. Microbially produced bile acids are associated with increased IgG autoantibodies and poorer mental wellbeing in fibromyalgia. Sci Rep. 2026;16(1). PMID 41735414
  9. Üçeyler N, Häuser W, Sommer C. Systematic review with meta-analysis: cytokines in fibromyalgia syndrome. BMC Musculoskelet Disord. 2011;12:245. PMID 22034969
  10. O'Mahony LF, Srivastava A, Mehta P, Ciurtin C. Is fibromyalgia associated with a unique cytokine profile? A systematic review and meta-analysis. Rheumatology (Oxford). 2021;60:2602-2614. PMID 33576773
  11. Albrecht DS, Forsberg A, Sandström A, et al. Brain glial activation in fibromyalgia — a multi-site positron emission tomography investigation. Brain Behav Immun. 2019;75:72-83. PMID 30223011
  12. Loggia ML, Chonde DB, Akeju O, et al. Evidence for brain glial activation in chronic pain patients. Brain. 2015;138(Pt 3):604-15. PMID 25582579
  13. Clauw D, Sarzi-Puttini P, Pellegrino G, Shoenfeld Y. Is fibromyalgia an autoimmune disorder? Autoimmun Rev. 2024;23(1):103424. PMID 37634681
  14. Caro XJ, Winter EF, Dumas AJ. A subset of fibromyalgia patients have findings suggestive of chronic inflammatory demyelinating polyneuropathy and appear to respond to IVIg. Rheumatology (Oxford). 2008;47:208-11. PMID 18208823
  15. Tamburin S, Borg K, Caro XJ, et al. Immunoglobulin G for the treatment of chronic pain: report of an expert workshop. Pain Med. 2014;15:1072-82. PMID 24422915
  16. Chang C, McDonnell P, Gershwin ME. Complex regional pain syndrome — false hopes and miscommunications. Autoimmun Rev. 2019;18:270-278. PMID 30639650
  17. de Greef BTA, Geerts M, Hoeijmakers JGJ, Faber CG, Merkies ISJ. Intravenous immunoglobulin therapy for small fiber neuropathy: study protocol for a randomized controlled trial. Trials. 2016;17:330. PMID 27439408 (NCT02637700)
  18. Geerts M, de Greef BTA, Sopacua M, et al. Intravenous immunoglobulin therapy in patients with painful idiopathic small fiber neuropathy. Neurology. 2021;96:e2534-e2545. PMID 33766992 (results of NCT02637700)
  19. Malik A, Lopate G, Hayat G, et al. Prevalence of axonal sensory neuropathy with IgM binding to trisulfated heparin disaccharide in patients with fibromyalgia. J Clin Neuromuscul Dis. 2019;20:103-110. PMID 30801480
  20. Oaklander AL, Herzog ZD, Downs HM, Klein MM. Objective evidence that small-fiber polyneuropathy underlies some illnesses currently labeled as fibromyalgia. Pain. 2013;154:2310-2316. PMID 23748113
  21. Kosek E. The concept of nociplastic pain — where to from here? Pain. 2024;165(11S):S50-S57. PMID 39560415
  22. Sluka KA, Clauw DJ. Neurobiology of fibromyalgia and chronic widespread pain. Neuroscience. 2016;338:114-129. PMID 27291641