Fibromyalgia syndrome—am I an autoimmune condition?

Either the majority or all patients with severe fibromyalgia syndrome harbour proalgetic serum-immunoglobulin G autoantibodies; their possible relevance for patients must now be established through clinical trials.

Abstract

Assessments of serum-autoantibodies in fibromyalgia syndrome (FMS) date back to the 1980s and have yielded inconsistent results. Based on a new passive transfer paradigm, since 2021 causative involvement of immunoglobulin G–mediated autoimmunity in severe FMS has been demonstrated in several studies, which have included UK, Swedish, and Canadian patients. These findings open the path to the development of novel diagnostic and immune-therapeutic approaches. Autoantibody targets and downstream mechanisms and the molecular processes that translate infection-, toxicity-, or stress-triggers into the FMS immune response in genetically or otherwise vulnerable individuals require study. These results in FMS also suggest that other chronic pain conditions or nonpainful symptom-based disorders may similarly be caused by noninflammatory minimally destructive autoantibody-mediated autoimmunity, thus offering hope for large groups of patients.

1. Background

Abnormal serum autoantibody titres in fibromyalgia syndrome (FMS) have been investigated at least as far back as the 1980s when Dinerman et al.¹⁶ reported the possible existence of a FMS subgroup without rheumatological comorbidity but with seropositivity for antinuclear antibodies (ANAs). The authors suggested that this finding may indicate involvement of autoimmune mechanisms. Since then, small studies have identified raised titres of various autoantibodies, potentially indicating abnormal immune activation or autoimmunity, although other studies were negative, and no consistent patterns have emerged.^{1,4,5,7,10,13,17,20,28,29,31,33,37–41,45–47,49–51,57,60} This evidence indicates that FMS is sometimes associated with mildly raised titres of specific immunoglobulin G (IgG) autoantibodies; furthermore, antibody-positivity is occasionally associated with an retrospectively identified FMS subgroup. The underpinning conceptual framework for these studies has in most cases been that antibody-induced structural or inflammatory change in tissue will induce activation of nociceptive (damage-sensing) nerve fibres, causing pain.

2. Witebsky–Rose postulates

In 1957, Witebsky et al.⁵⁸ outlined criteria that can be used to identify a disorder as autoimmune, modelled on the Koch's postulates for the identifications of infectious microbes. A revised version was published 35 years later by one of the original authors, Noel Rose, together with Constantin Bona⁴⁸ reflecting on advanced understanding of autoimmune mechanisms. Rose and Bona proposed to categorise evidence for autoimmune involvement in disease into (1) direct proof, (2) indirect evidence, and (3) circumstantial evidence.

Very rare direct proof derives from human-to-human transfer of disease such as when a scientist self-injects with plasma from a sick patient, or when pathogenic IgG from a mother is transported across the blood–placental barrier into the unborn child's circulation during the third trimester. Any transfer of serum-IgG purified from patients to experimental animals, which causes the patient phenotype in the animals is considered an alternative form of direct evidence; the authors specifically highlighted the example of serum-IgG transfer from patients with pemphigus to neonatal mice causing typical skin blisters. Sometimes, direct evidence can also derive from in vitro techniques, such as when serum-antibodies purified from patients with paroxysmal cold hemoglobinemia lyse erythrocytes taken from suitable patients.

Methods producing indirect evidence include the immunisation of experimental animals with the pertinent autoantigen, ie, the molecular structure against which the autoimmune response is directed in the human. As with direct proof approaches, these indirect methods are considered successful if they trigger the development of the human phenotype in the animals. Examples for indirect evidence include classical experimental results in myasthenia gravis where immunisation of rabbits with acetylcholine receptors derived from eel causes profound muscle weakness.⁴² Such methods obviously require that the antigen is known.

Finally, circumstantial evidence involves a favourable patient response to immune suppression. Rose and Bona thought that any such evidence would not provide proof for autoimmune disease, however, should incentivise further research. Several novel immune treatment methods might, however, instead be classed as direct evidence if the Witebsky–Rose criteria were to be revised in the future. Immunoadsorption, which removes serum-antibodies by filtration,⁶ and treatment with FcRn receptor antagonists,¹⁴ a novel class of drugs which reduces serum-antibodies by increasing their metabolism, both specifically target antibodies rather than causing general immune suppression. Such novel treatment technologies have been shown effective in several autoantibody-associated disorders, including myasthenia gravis.²⁷ They differ from immune-suppressive approaches with very broad immune modifying effects known at the time of publication of the Witebsky–Rose criteria.

3. Summary of findings in FMS since 2021

3.1. Passive immunoglobulin G transfer of FMS immunoglobulin G—direct evidence for autoimmunity

We hypothesised that, mirroring our previous findings which had demonstrated that IgG antibodies may be responsible for causing persistent complex regional pain syndrome, FMS may similarly be caused by function-modifying IgG autoantibodies that do not elicit tissue destruction or systemic inflammation. Further that their effects can be identified and characterised upon passive immunoglobulin G transfer to rodents; if successful, then such experiments would provide direct evidence for an autoimmune contribution to FMS. Based on our method first developed for Complex Regional Pain Syndrome,^23,52 which in itself was adapted from classical studies in myasthenia gravis (MG),⁵³ we transferred 8 mg/kg/d of affinity-purified human IgG from patients with severe FMS to rodents on each of 4 days and assessed their behaviour. The daily mass of IgG injected was similar to that used in MG studies where initially 10–11 mg were injected/day producing human IgG concentrations in the mouse similar to the human body. This was reduced to 8 mg/d in later MG transfer studies. The injection period was longer (10–14 days) in the classical MG studies; however, first pathogenic IgG effects were observed as early as 12 hours after the first injection with some preparation, and marked weakness had started from 2 to 7 days.⁵⁴ We individually investigated serum-IgG from 8 consecutive UK (Liverpool) patients.

Each IgG preparation caused significant mechanical hyperalgesia in the animals, which was often observed as early as on the second experimental day, ie, after only one IgG injection, and was in all cases observed by the third experimental day, ie, after 2 daily IgG injections. Hyperalgesia was demonstrated by testing mouse hind paws using standard methods (Randall–Selitto). Separate assessment of mouse thighs confirmed widespread hypersensitivity, resembling the FMS condition. Seven out of these 8 preparations also elicited cold hyperalgesia.

Healthy control preparations and the injection of FMS patient serum depleted of IgG (“flow through”) were without effect. These results in UK patients were reinforced by results from testing of 2 serum pools derived from Swedish patients with FMS (8–14 patients), recruited through a rheumatology outpatient clinic (all testing: Andersson-Bevan lab, King's college London).

In these FMS transfer experiments, additional phenotypical elements resembling FMS included significantly reduced grip strength (Andersson-Bevan lab), reduced locomotion during peak activity times at night, and skin mild small nerve fibre pathology (Svensson lab, Karolinska, Stockholm). Skin-nerve preparations comprising of hind paw skin and fibres of the saphenous nerve were then dissected from the injected animals after 4 days of IgG injection and probed for sensory nerve fibre excitability ex vivo. Skin-hypersensitivity to mechanical and cold stimulation was demonstrated (Andersson-Bevan lab). There was no indication for a systemic inflammatory response in the rodents (serum cytokine levels were normal).

The specific cellular binding targets of the transferred IgG were probed in tissues after dissection from the injected rodents after humane killing (ex vivo). These experiments identified IgG deposition in dorsal root ganglia, with no deposits in the spinal cord and brain. Within the DRGs, binding appeared specific to satellite glial cells (SGCs) surrounding sensory nerves; SGCs in the FMS-IgG injected rodents had become activated as evidenced by increased glial fibrillary acidic proteine expression.

In cultures of primary rodent dissociated SGCs from noninjected animals, stained in vitro with either FMS or healthy IgG, increased SGC binding vs healthy was observed. Similarly, in human cadaveric DRGs from deceased tissue donors who did not have FMS stained in vitro, increased staining to SGCs vs healthy was demonstrated (all IgG deposition, staining and cellular experiments Svensson lab).²²

In later studies (Svensson lab) utilising an optimised protocol and assessing serum from larger Canadian and Swedish FMS patient cohorts, the degree of IgG binding in vitro to primary dissociated cultured SGCs again correlated moderately with the patients' pain intensities, and inversely correlated weakly to their pressure pain threshold, but importantly there was no correlation with body mass index, FMS duration (rendering it unlikely that these antibodies are the unspecific result of having chronic disease), or conditioned pain modulation (a marker of central pain inhibition). Increased SGC staining was largely restricted to patients with severe FMS, whereas serum-IgG from donors with a pain intensity of <6/10 rarely stained SGCs. Staining of mouse SGCs also correlated moderately with FMS-IgG SGC staining of cadaveric human DRGs obtained from organ donors.³² Binding to murine neuronal cells was almost always normal; differential binding between SGCs and neurons indicates that increased SGC binding by FMS IgG is unlikely entirely caused by any unspecific “stickiness” of the patient preparations. Several healthy control samples strongly stained SCGs suggesting that SCG-binding antibodies are not necessarily proalgetic.

The association of anti-SGC antibody titre with FMS severity was then confirmed in a most recent large Swedish cohort and was again shown to relate to spontaneous pain intensity, but, in contrast to the earlier study, not to evoked FMS pain.¹⁹

With the described exception of SGC binding, and very recent findings concerning mast cell binding https://www.biorxiv.org/content/10.1101/2025.05.15.652596v1 cellular binding targets of pathogenic FMS autoantibodies, molecular binding structures, or effects downstream from antibody binding are yet unknown.

Patients in the tested cohorts had “primary” FMS without diagnosed rheumatological comorbidities such as rheumatoid arthritis. We currently investigate whether “secondary” cases may similarly be autoimmune. FMS is surprisingly common in patients with certain established autoimmune disorders, eg, 20% to 30% prevalence in rheumatoid arthritis; while this phenomenon has previously been explained by sensitisation effects related to the chronically painful inflamed joints, it is alternatively possible that these patients have developed an additional autoimmune reaction leading to FMS.

3.2. Circumstantial evidence for autoimmunity in FMS

Patients with FMS are predominantly female and mid-aged, features typical in autoimmune conditions. As outlined, patients with rheumatological conditions are at 5 to 30 times higher risk to develop FMS compared with the general population.⁴³

No reports of FMS treatment with either immune-absorption or the recently developed FcRn drugs (which reduce IgG autoantibodies) have been published. A randomised controlled trial of an FcRn drug has completed recruitment in late 2023 with results expected in late 2025 (NCT05643794). Biologics trials in rheumatological conditions, using anticytokines or rituximab, a CD20 antagonist that depletes B cells are often considered a potential source of information when data about pain relief in patients with concomitant (“secondary”) FMS is extracted. However, no reliable data has emerged so far to either refute or confirm any beneficial effects of these drugs in FMS associated with rheumatological disorders. Anecdotally, established medications for rheumatoid arthritis do not improve FMS pain, although very preliminary evidence for a possible, mildly beneficial effect of tumor necrosis factor alpha inhibitors in axial spondylarthritis associated FMS is available.⁴⁴ New therapies potentially addressing both peripheral and central immune targets are being scrutinised for any FMS pain-relieving effects in patients with rheumatoid arthritis and comorbid FMS.³⁵ It should be noted that autoantibody-mediated autoimmunity caused by cell function-modifying autoantibodies would not necessarily respond to anti-inflammatory agents. In addition, inconsistent effects of rituximab have been reported in the largest subgroup of myasthenia gravis, an autoantibody-mediated condition indicating that a negative response to this drug does not preclude autoantibody causation.⁵⁶ This latter finding may be attributable to the persistence of pathogenic antibody production by plasma cells, which are not affected by CD-19/20 antagonist treatment.

3.3. Immune cell changes possibly consistent with FMS autoimmunity

A reduction in the number of circulating natural killer (NK) cells in FMS, particularly the subset “CD56^BRI,” was identified.⁵⁵ Cell surface markers indicated that these cells were on average “chronically activated and exhausted.” In vitro, these cells were hyper-responsive. These changes were associated with evidence for a redistribution of these cells to human skin tissue. The cause for these changes is unknown. The authors speculate that these NK changes might be downstream to an autoimmune trigger, where autoimmune-induced dermal nerve fibre damage might serve as attractant factor.

4. Debate on incomplete transfer phenotype, overlapping pain conditions, inflammation, triggers, and the biopsychosocial dimension

4.1. Incomplete transfer phenotype

It is currently unknown whether all FMS features can be transferred to rodents by passive IgG transfer. For example, we have not yet attained persuasive evidence for the transfer of severe fatigue, or of cognitive dysfunction. Whereas it is possible that novel experimental paradigms will elicit these features in rodents after FMS IgG transfer, alternatively, some phenotype aspects may be caused by other mechanisms. How might it be possible that both IgG-mediated and putative non–IgG-mediated phenotypical disease elements exist in parallel in the same patient, which often have their onset around the same time? More research is needed, but one speculative option is that an overarching pathology causes downstream biological abnormalities, including but not restricted to the loss of B-cell tolerance with production of abnormal IgG. Recently, an abnormal microbiome composition specific to FMS has been reported; fecal transplant from patients to mice causes FMS-like features in these animals, and healthy microbiota transfer may be therapeutic in patients with FMS.^8,18,34 An abnormal microbiome, perhaps formed as a result of triggers such as described below, might prime various immune cell types (such as B cells and neutrophiles)⁹ towards dysfunction, and additionally cause nonimmune pathology.²¹

4.2. Overlapping pain conditions

FMS shares many systemic features with non-FMS pain conditions such as pelvic pain or back pain, including fatigue, short-term memory problems, poor sleep, and others. It was recently postulated that if the FMS symptoms are caused by autoantibodies, then FMS-overlapping conditions such as chronic back pain should be caused by similar mechanisms.¹¹ Since these conditions are common, this was then taken as evidence that FMS is unlikely autoimmune. Of course, whether or not overlapping conditions may have an autoimmune contribution is currently unknown and more research is needed. A substantial, IgM-mediated proalgetic effect has recently been demonstrated in a rodent model of chronically painful disc injury, suggesting that this field is rapidly expanding.²⁵

4.3. Systemic inflammation

The fact that FMS is not associated with a systemic inflammatory response or an obvious responsiveness to anti-inflammatory agents was cited as evidence against autoimmune involvement. It will remain essential to understand mechanisms downstream from binding of the described proalgetic antibodies. Current evidence suggests that similar as in several neurological conditions but possibly not in rheumatological conditions, FMS-associated disease-eliciting autoantibodies indeed may elicit neither systemic inflammation, nor an overt inflammatory reaction locally at the site of antibody binding. To provide an example, neuromyotonia is a rare neurological muscle-nerve autoimmune condition caused by autoantibodies against molecular structures within potassium channel complexes on peripheral neurons. In this condition after antibody binding, potassium channels are being complexed and internalised, subsequently changing nerve membrane potentials, causing typical clinical features, including pain.¹⁵ Ouvert inflammation does not feature and is not required to elicite these changes in neuromyotonia. The burning and cramp-like pain of neuromyotonia is often responsive to immune therapy that reduces serum antibody levels.³⁰ The absence of overt inflammatory elements in the FMS autoimmune pathophysiology, whilst perhaps counterintuitive and unusual, would correspond to the possible lack of a response to classical anti-inflammatory agents.

4.4. Triggers and the biopsychosocial dimension

If FMS is autoimmune, then it has “special” features. The recognition of biopsychosocial triggers to the onset of FMS⁵⁹ implies that a substantial group of patients, although not all and perhaps not the majority, have experienced severe distress before FMS onset. The sensitive nature of such a link, which can and commonly does lead to misunderstanding and stigmatisation, has been discussed.¹² Clearly, FMS does not develop in everybody who experiences severe distress, so that other factors must also play a role. At the same time, more fundamentally the recognition of a likely trauma/distress-trigger in a subgroup of patients with FMS, together with the described autoantibody findings now positions any distress-FMS relation into a new context. This implies that experience of trauma and distress might elicit a specific immune-biological response in genetically or otherwise (eg, past infection, toxicity, trauma) vulnerable individuals, which includes the production of harmful noninflammatory autoantibodies. More studies are needed to confirm this. Similar sequences from distress experience to disease phenotype are of course known from other autoimmune disorders,^2,24 but in FMS, this phenomenon might be particularly common; in addition, FMS is also a much more common condition than “classical” autoimmune disorders; hence, any such relation should perhaps more acutely lead to considerations of preventative approaches.

Alternative primary triggers for the immune reaction leading to FMS might include infection, such as after COVID19 exposure^26,36 or toxicity, such as after fluroquinolone medication.³ Other patients may develop the condition spontaneously, without any trigger.

Understanding autoimmunity in FMS should help us to better understand our human nature and how profoundly, truly biopsychosocial we in fact are. A debate about the additional importance of preventing toxic distress and how early preventative interventions might be designed and implemented would be welcome.

5. Summary

There are strong indicators for a prominent role of autoimmune mechanisms, acting through production of function-modifying proalgetic IgG in the majority, or all patients with severe FMS. Antisatellite glial cell antibodies and anti mast-cell antibodies may contribute to the proalgetic effect. The clinical relevance of pathogenic IgG antibodies for patients is yet to be determined in clinical trials.

Disclosures

A.G. has received consultancy fees from UCB, a company that produces an FcRn receptor blocker, and Clerkenwell Health.