Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a complex disease associated with high symptom burden and low quality of life (1–3). The societal costs are massive, and in the US the annual cost was estimated at 18–51 billion dollars (4).

The prevalence was estimated to 0.1–0.8% of the population (5, 6). ME/CFS is often preceded by an infection or other immunological trigger, is three to four times more prevalent in females than males, and is influenced by genetic predisposition (7).

The public awareness of ME/CFS has increased after the COVID-19 pandemic, with a proportion of patients struggling with brain fog, post-exertional malaise (PEM) and fatigue after the initial SARS-CoV-2 infection, denoted a post COVID-19 condition, or Long COVID (LC). Many patients recover, but among those with persistent LC for more than 2 years, approximately half meet the diagnostic criteria for ME/CFS (8, 9), indicating that this subgroup represents a post-infectious fatigue syndrome. Evidently, including this group will increase the prevalence, and in a recent large cross-sectional US survey, out of the 1.67% of adults who reported ME/CFS-like illness, 14% of cases were preceded by COVID-19 infection (10).

ME/CFS is presently a disease with unknown etiology, with no validated biomarker and no standard approved effective treatment. Patients experience less than consistent health services (11), and the lack of consensus on pathomechanisms may provide the ground for questionable and unscientifically founded advice and treatment of patients. There is an urgent need for elucidation of disease mechanisms and for development of effective treatment.

In 2021, we published a viewpoint article discussing pathomechanisms and possible intervention in ME/CFS (12). In brief, we suggested that ME/CFS could involve a variant of an autoimmune mechanism, with a role for B cells, plasma cells and autoantibodies. After systemic infections such as COVID-19 or Epstein–Barr virus (EBV), functional autoantibodies to a broad range of extracellular proteins emerge in otherwise healthy individuals as part of a normal immune response, including receptors and soluble molecules such as interferons, interleukins, and cytokines (13, 14).

Thus, it is plausible that the immune response after a systemic infection may induce a broad array of functional autoantibodies that affect biological systems and could be involved in the different symptoms experienced by ME/CFS patients. A role for G-protein coupled receptor (GPCR) autoantibodies has been suggested (15), including antibodies to muscarinic and adrenergic receptors, with subsequent dysregulation of autonomic function.

Such functional autoantibodies occur in previously healthy individuals after infections and usually subside over time, but in ME/CFS (or LC) some autoantibodies may persist. This model is compatible with a study of CFS following infectious mononucleosis, in which 13% of the infected fulfilled the criteria for CFS after 6 months, 7% after 12 months, and 4% after 2 years (16). However, although specific functional autoantibodies were elevated during acute COVID-19 infection (14), the number and magnitude of autoreactivities to the exoproteome were similar in individuals with or without ME/CFS (17), suggesting that the pattern of functional autoantibody responses, rather than single pathogenic autoantibodies, may be relevant.

Postinfectious functional autoantibody responses with disturbed regulation of blood flow may be a central pathomechanistic factor in ME/CFS, resulting in tissue hypoxia upon exertion. Endothelial dysfunction in small and large arteries has been shown (18–21). Studies have reported preload failure with reduced venous tone and venous return with low right atrial pressure during exercise tests, possibly due to autonomic dysregulation, followed by decreased cardiac output on exertion, and also reduced peripheral oxygen extraction with increased SvO₂ in mixed venous blood (22, 23). Such physiological alterations with resultant tissue hypoxia on exertion, would be expected to induce autonomic compensation with increased sympathetic tone, which has been repeatedly observed in ME/CFS for decades (24, 25), as well as metabolic adaptations (26–28). Likewise, physiological functions like neuronal signaling may be affected.

Further, a role for autoantibodies in the ME/CFS pathomechanism is supported by data from immunoadsorption studies with clinical responses in ME/CFS patients (29, 30), and varying results in LC (31, 32). This technique extracts antibodies in general from the blood and transient symptomatic recovery can be seen within hours or days. However, the source of the antibody production is not affected.

From 2007 and onwards, we have observed patients in our cancer ward with long-standing ME/CFS who got cancer, and who reported that immunomodulatory cancer drug treatment they received had beneficial effects on the ME/CFS symptoms. We hypothesized that ME/CFS in a subgroup is associated with a variant of an autoimmune pathomechanism. Based on this hypothesis, we have performed several interventional clinical trials using the B-cell depleting anti-CD20 antibody rituximab in ME/CFS (33–37). Our early rituximab trials indicated that a subgroup of patients, and approximately half of those given rituximab maintenance treatment, showed signs of clinical improvement from B-cell depletion (36). However, the results from a larger randomized, double-blinded and placebo-controlled trial showed no difference between rituximab and placebo treatment (37). We also performed an open-label clinical trial with six infusions of the cytotoxic and immune system modulating drug cyclophosphamide in patients with moderate to severe ME/CFS, with clinical responses in 55%, and apparent benefit on the long-term clinical outcomes (33, 38). A recent review article has summarized ongoing trials and drugs of possible interest in ME/CFS and Post-COVID-19 condition (39).

Taking into account our observations from the clinical trials, we suggest that in ME/CFS patients the main autoantibody production occurs in long-lived plasma cells which, being CD20 negative, are not affected by the anti-CD20 antibody rituximab (40). This is a well-known mechanism for lack of response to rituximab in established autoimmune diseases, unless the treatment period is extended until the recruitment of new plasma cells is affected. We suspect this may have been the case in our previous open-label rituximab maintenance study (12, 36). Local observations of two ME/CFS patients with transient responses to the proteasome inhibitor bortezomib further supports the concept of plasma cell-directed treatment, however not pursued to a trial because of potential toxicity.

The known functions of long-lived plasma cells are well compatible with disease mechanisms we believe to be in ME/CFS. Once they have been programmed to produce autoantibodies, these plasma cells can survive in niches in the bone marrow or gut wall for decades and send a continuous stream of antibodies into the blood and further into different tissues.

Thus, we believe there is a rationale for therapeutic intervention targeting plasmablasts, short- and long-lived plasma cells in ME/CFS. An unselective plasma-cell depleting agent targeting CD38 will subsequently reduce both protective antibodies and autoantibodies, and could induce an altered balance of the plasma cell pool in some patients, potentially resulting in long-lasting changes in the disease phenotype upon repopulation and increasing IgG levels after intervention.

Herein we report the results of the KTS9 pilot study of 10 female patients with moderate to severe ME/CFS, treated with subcutaneous administration of the monoclonal anti-CD38 antibody daratumumab, targeting plasmablasts and long-lived plasma cells. The study is exploratory, and the main objectives were to assess safety and feasibility, and to explore possible beneficial or negative clinical effects on the course of ME/CFS through at least 12 months’, and up to 24 months of follow-up.

As there are no previous reports on plasma-cell depletion therapy in ME/CFS, the primary objective for this pilot study was to investigate the feasibility and toxicity of subcutaneous daratumumab (Darzalex®) in patients with moderate to severe disease. There were mild adverse events of CTCAE grade 1–2, and no SAE. All patients experienced mild erythema at the injection site, and there were no injection-related reactions (IRR) or allergic reactions. There were, as described above, uncomplicated episodes of lower urinary tract infections in one patient and herpes zoster in another, but no episodes of severe infections and no hospitalization. All patients had episodes of viral upper airway infection, including COVID-19, but uncomplicated. All patients completed the planned daratumumab injections and follow-up.

Six out of 10 patients experienced clinical improvement during follow-up. The pattern of improvement was quite homogenous, starting from 6 to 8 weeks after the first injection, with gradual improvements of all ME/CFS related symptoms through follow-up. During intermittent upper airway viral infections including COVID-19, the improving patients experienced transient symptom worsening, typically of 1–2 weeks’ duration. To us, the improvement or elimination of all ME/CFS symptoms suggests that in this subgroup, central pathomechanistic factors are affected by anti-CD38 intervention. Among the patients with clinical improvement, five out of six experienced pronounced and sustained symptom alleviation until end of follow-up after 12–24 months from inclusion. Thus, in these five, the mean SF-36 Physical Function score at the end of follow-up was 88 (range 80–95) which is close to or within the general population range (44, 45), and the mean maximum increases in SF-36 PF from baseline was 53 (range 39–70) during follow-up. Similar changes were detected through follow-up for DSQ-SF total score and were supported by increases in physical activity level. Toward the end of follow-up, these five patients with sustained clinical improvement described improved health and a major positive impact on their daily lives. One patient with moderate–severe ME/CFS had a transient and moderate clinical improvement, but experienced a gradual relapse toward the end of follow-up (green line in figures). She was among the first six patients included, received four daratumumab injections, and had a relatively early increase in serum IgG levels during follow-up. At extended follow-up at 66 and 92 weeks after start of intervention, she again experienced some improvement in physical functioning, reporting SF-36 PF of 35 at week 66, and SF-36 PF of 50 at week 92, compared to 21.4 at baseline. After the end of 24 months’ follow-up, one patient (blue line) experienced a gradual symptom worsening with partial relapse of ME/CFS, starting from 26 months after inclusion.

The experiences of the six patients with clinical improvement are in sharp contrast to the four patients with no significant clinical impact on the disease course. These had very stable values for SF-36 PF, DSQ-SF total score, Function level and steps per 24 h during follow-up. Thus, except for one transient and moderate response, the data indicate a dichotomous pattern for clinical improvement, either pronounced and sustained through follow-up, or no change at all.

The purpose of daratumumab intervention was to achieve a transient plasma-cell depletion affecting especially the long-lived plasma-cell compartment, with moderate decline in serum IgG levels. The mean relative reduction in serum IgG levels during follow-up was 48% (range 22–62%) and was higher among the six patients experiencing clinical improvement (mean 54%) compared to the four patients with no clinical benefit (mean 40%). At follow-up at 5 and 9 months, the mean serum IgG levels were around the lower level of the normal range, increasing from then on. Interestingly, except for the patient with gradual relapse from 26 months (i.e., after the end of trial follow-up), the gradually increasing serum IgG levels have so far not been accompanied by increasing ME/CFS symptoms. However, longer follow-up will be needed to indicate whether some patients have actually had a lasting effect from transient depletion and subsequent repopulation accompanied by increasing IgG levels, and perhaps an altered balance in the pattern of functional autoantibodies which we believe are important in the pathomechanism.

The IgG subclasses, measured at baseline and at 9 months, showed some differences in the relative reductions during follow-up. The most evident relative reduction of overall 60% was seen for IgG4, with 65% among patients with clinical improvement, versus 29% among patients with no clinical benefit. IgG4 has special characteristics, being the least abundant subclass, with reduced ability to activate Fc-dependent effector functions such as ADCC and complement-mediated cell killing, with stochastic fragment antigen binding (FAB) arm exchange and often bispecific design, although monovalent in antigen binding. IgG4 usually has blocking and inhibitory effects on the target molecule or the immune response, and autoimmune diseases associated with IgG4 autoantibodies are often more responsive to anti-CD20 B-cell depletion intervention (46). Whether the present IgG4 data has pathomechanistic implications for ME/CFS, remains to be elucidated.

As expected, due to their expression of CD38, NK cells in peripheral blood were depleted after the first daratumumab injection, to values either below or close to the lower normal range (59–533 × 10⁶/L) depending on the baseline level. The number of NK cells slowly increased at 5 and 9 months follow-up, depending on whether the patient had the last daratumumab injection at 6 weeks or had additional daratumumab injections until week 30. All patients who completed 15–21 months’ follow-up so far had NK-cell numbers within the normal range.

Interestingly, there was a significant correlation between baseline number of NK cells in peripheral blood and clinical improvement during follow-up, assessed by maximum increase in SF-36 Physical Function, or by maximum decrease in DSQ-SF total score. The four patients without clinical benefit had baseline NK-cell numbers in the lower normal range (mean 97, range 80–115), while those who experienced clinical improvement had higher NK cell numbers at baseline (mean 238, range 136–383). NK cells mediate antibody-dependent cellular cytotoxicity (ADCC) by binding to the IgG Fc region, to ensure adequate plasma-cell depletion when daratumumab binds to CD38. Thus, higher number of baseline NK cells would be expected to enhance plasma-cell depletion, and result in a more marked decline in serum IgG levels. Similarly, one would expect depletion of NK cells when daratumumab binds to CD38 on the NK cells themselves. The patients with no clinical benefit from daratumumab intervention had lower numbers of NK cells in peripheral blood both at baseline and through follow-up until 40 weeks, and also higher serum IgG levels with lower relative reductions from baseline. Whether the observed associations between low baseline NK-cell numbers, serum IgG reductions and clinical improvement during follow-up are biologically determined or coincidental, cannot as yet be determined for certain due to the low number of observations.

According to the literature, in multiple myeloma NK-cell numbers were found to decline with increasing daratumumab exposure, but with little impact on efficacy or safety due to the remaining peripheral blood mononuclear cells (PBMC) fractions maintaining the ability to carry out ADCC, at least ex vivo (47). Also, even though multiple myeloma patients receive more daratumumab injections than in the present study, the NK cells are not completely depleted by daratumumab.

In addition to a reduced number of NK cells in peripheral blood, NK-cell dysfunction in ME/CFS has been described for decades (48). In a recent meta-analysis, reduced NK-cell cytotoxicity was interpreted as the most consistent immune finding in ME/CFS (49), but with high heterogeneity among studies and partly ascribed to different experimental methods. Reduced NK-cell cytotoxicity has also been reported in autoimmune diseases such as systemic lupus erythematosus and rheumatoid arthritis (50). A reduced NK-cell cytotoxicity could perhaps contribute to subtle reactivation of latent infections, such as by herpes virus, thereby maintaining a dysregulated immune response. However, our clinical impression when assessing patients with ME/CFS, is that a subgroup of approximately half of patients report that they have very rarely suffered common viral infections after acquiring ME/CFS, which could rather indicate a hyperactivated innate immune system.

In recent years, several case series have been published demonstrating high efficacy and good tolerability for daratumumab intervention in several treatment-refractory autoimmune diseases, with a role emerging for anti-CD38 antibody therapy (51, 52). In autoimmune haemolytic anaemia (53), anti-phospholipid syndrome (54), treatment-refractory autoimmune cytopenias, cold agglutinin disease, and other autoimmune diseases (ITP, systemic lupus, anti-CASPR2 encephalitis, rheumatoid arthritis, ANCA-associated vasculitis) daratumumab has shown high efficacy in case series (55–57). In systemic lupus erythematosus refractory to B cell depletion, cyclophosphamide and several immunosuppressive drugs, daratumumab was effective and clinical responses sustained by anti-BAFF (belimumab) maintenance treatment (58). These case series reported good tolerability to daratumumab.

A recent review summarized experiences with daratumumab in different refractory autoimmune diseases (59). A total of 38 publications reporting the clinical course of 83 patients were included. The summarized data showed a response rate of remission or improvement in 81%. Among the included studies the median number of daratumumab infusions/injections used was four (range 1–24 injections). Interestingly, the authors found no significant correlation between number of daratumumab doses and either clinical response or hypogammaglobulinemia (59).

Daratumumab is approved for use with chemotherapy and other immune modulatory drugs, or as monotherapy for patients with multiple myeloma or amyloidosis, with formulations for intravenous and subcutaneous administration. The optimal therapeutic daratumumab dosing regimen in ME/CFS is of course not known, and in the present pilot study six patients received four injections, while the next four had in total seven injections. The low number of patients precludes transferable conclusions. However, taking into account the lack of correlation between number of daratumumab injections and clinical response in refractory autoimmune diseases (59), and the observed (possible) association between baseline NK-cell number and clinical improvement during follow-up, one could speculate that the first daratumumab injection may be the most important provided that the NK-cell number at baseline is relatively high.

In a new amendment to the present study, we aim to test this by including four new patients with ME/CFS following a SARS-CoV2 infection who fulfil the trial inclusion criteria, and who have baseline NK-cell numbers above 125 (x10⁶/L). In this amendment, the daratumumab dosing schedule has been reduced. These data will be published later as a case series. Possible interventions for antibody-targeting therapies in Long COVID have recently been summarized (60).

A recent study evaluated the presence of autoantibodies in LC patients with high neurological symptom burden (61), using a human autoantigen microarray, immunoprecipitation, immunofluorescence with patient IgG and a broad range of tissues, and also passive transfer of purified patient IgG to mice that developed neurological symptoms. Diversity of detected autoantibodies was evident (60), and staining seemed to co-localize with pericytes and to some extent with endothelial cells. These data strengthen the case for autoreactivity and functional autoantibodies, induced during the active COVID-19 infection, but persisting and disturbing biological function, as important pathomechanistic factors for neurological symptoms in LC (61).

However, a recent article using platforms for extensive autoantibody screening to extracellular proteins including Rapid Extracellular Antigen Profiling (REAP), reported no significant difference in autoantibody reactivities between ME/CFS patients and healthy controls (17). Lack of reproducibility in reports of autoantibodies in ME/CFS could partly be explained by different techniques for expression of structurally complex proteins, which can make proper autoantibody binding to the epitope problematic.

It is possible that the pattern and balance of functional autoantibodies emerging after systemic infections, which may be agonistic or antagonistic and disturb homeostasis, could be more important than the quantity of specific antibodies. In ME/CFS patients, such patterns of autoantibodies may persist over time with symptoms that to some extent mimic an acute viral infection host response with fatigue, malaise, exercise intolerance, feeling feverish, cognitive symptoms, pain, sore throat, tender lymph nodes and sensory hypersensitivity.

CD38 is an ectoenzyme with multiple and high functional complexity (62). CD38 modulates cell differentiation and signal transduction, and has effector functions during inflammation, including cell recruitment, regulation of NAD availability, and cytokine patterns (63). CD38 is expressed on the cell surface and may also be expressed in intracellular compartments. It is present in many different cell types, but mainly in hematopoietic cells. The highest expression levels are found in plasmablasts, short- and long-lived plasma cells, but also NK cells. Weaker CD38 protein expression is found on subsets of macrophages, B cells and T cells.

Due to the high complexity of the CD38 molecule, the exact mechanisms for the observed beneficial effects using daratumumab in ME/CFS are presently not known. Important functions associated with B-cell subsets could be relevant, such as antigen-presentation and cytokine production (64, 65).

However, we believe our pathomechanistic model, with a role for B cells, plasma cells and functional autoantibodies (12) is supported by the observed clinical responses in a subgroup of ME/CFS patients from daratumumab intervention and concurrent reductions of serum IgG levels, and also taking into account the observed transient effects reported from immunoadsorption experiments (29, 30). In addition, there are several other indications that a variant of an autoimmune mechanism is relevant in ME/CFS, such as a marked female preponderance, associations with autoimmunity-related gene variants (66), and a skewed B-cell receptor gene usage (67). Elderly ME/CFS patients have an increased risk of B-cell lymphoma, especially the low-grade marginal zone lymphomas often associated with autoimmunity or chronic infections (68). Serum B-cell activating factor (BAFF) increase and antigen driven B-cell clonality (69), altered B-cell differentiation (70), and presence of autoantibodies to adrenergic and cholinergic receptors (71) have been reported in ME/CFS. In LC, several studies support an autoimmune pathogenesis (14, 72–74). Although plasma-cell induced depletion by daratumumab with a subsequent decline in autoantibodies was the main goal in this study, we cannot exclude that other induced alterations in the immune balance can be involved in the clinical effects.

The main limitations of this pilot study are the small number of patients, exclusively female participants, and the open label design. Although data are promising, no firm conclusions should be drawn before additional studies have been performed. In June 2025, we launched a randomized, double-blind and placebo-controlled study, including patients with moderate to severe ME/CFS patients according to Canadian consensus criteria, and at least 2 years disease duration. Long COVID patients fulfilling the inclusion criteria may qualify for inclusion.

In this new randomized study, we aim to optimize the design, taking into account our previous experiences from the double-blind study using the anti-CD20 antibody rituximab versus placebo in ME/CFS (37), which came out negative despite promising data from the preceding phase II trials (35, 36). Factors we believe may have influenced that trial outcome, and thus should be evaluated carefully and optimized in a new study, are heterogeneity of the patient sample, natural symptom variation over time interfering with the endpoints especially among patients with milder degree of ME/CFS, outcome measures that proved not stringent enough, recall bias, and placebo mechanisms. Also, although the double-blind design is meant to prevent information bias, uncertainty about group allocation may cause some participants to favor more neutral responses when self-reporting changes over time.

In conclusion, in this pilot study, 10 female ME/CFS patients with high symptom burden and long disease duration received the anti-CD38 antibody daratumumab to target long-lived plasma cells, with subsequent reductions of serum IgG levels. The feasibility was very good with no serious adverse events. Six patients experienced clinical responses, five of whom had major and durable symptom improvements through follow-up. However, definite conclusions should not be drawn before a randomized, placebo-controlled study has been performed.