Efficacy of repeated immunoadsorption in patients with post-COVID myalgic encephalomyelitis/chronic fatigue syndrome and elevated β2-adrenergic receptor autoantibodies: a prospective cohort study

Abstract

Background: Since the pandemic, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has become the leading trigger for myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). Evidence indicates that autoimmunity plays an important pathophysiological role. We aimed to evaluate the effectiveness of IA treatment in post-COVID ME/CFS patients.

Methods: This pre-post study included 20 post-coronavirus disease 2019 (COVID) ME/CFS patients found to have elevated β2 adrenergic autoantibodies (β2 AR-AB) between October 2022 and October 2023. Patients, with a median disease duration of 22 months (IQR: 15-31), were treated with five immunoadsorption sessions at Charité - Universitätsmedizin Berlin, Germany. Seven were male and 13 female, with a median age of 40 years (IQR: 36-51). The primary end point was the change in the Short Form (36) Health Survey physical functioning domain (SF36 PF) from baseline to four weeks post immunoadsorption. Key symptoms were assessed via questionnaires over six months. Handgrip strength and EndoPAT® measurements were used to evaluate muscle fatigue and vascular dysfunction. Seven patients who worsened after an initial response received a second cycle.

Findings: The treatment was generally well tolerated, reducing total immunoglobulin G by 79% (CI: 73-84%) and β2 AR-AB by 77% (CI: 58-95%). Patients demonstrated a mean increase in the SF36 PF of 17.75 points (CI: 13.41-26.16), with the greatest improvement occurring between months two and three, and significant gains maintained through month six. 14/20 (70%) patients were categorized as responders with an increase in the SF36 PF of ≥ ten points. Further lasting improvements were reported in fatigue, post-exertional malaise, pain, cognitive, autonomic, and immunological symptoms. Female patients had increased repeat handgrip strength at month six.

Interpretation: Immunoadsorption may improve symptoms in post-COVID ME/CFS patients. The beneficial effects of IgG depletion suggest a significant role for autoantibodies and disturbed B-cell function in the condition's pathophysiology.

Funding: Funded by The Federal Ministry of Education and Research and the Weidenhammer Zöbele Research Foundation.

Keywords: Autoantibodies; COVID-19; Chronic fatigue syndrome; Immunoadsorption; Myalgic encephalomyelitis; Pandemic; Plasmapheresis; Post-COVID syndrome; SARS-CoV-2.

Fig. 1

Course of the mean 36-Item Short-Form Survey physical functioning domain (SF36 PF) over the study period. The duration of IA therapy is indicated by the green bar. A higher score indicates less restriction in physical functioning. Error bars represent 95% confidence intervals. Data were analyzed using a linear mixed-effects model fitted by restricted maximum likelihood (REML), with t-tests computed using Satterthwaite's method for degrees of freedom with significance levels indicated as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. From top to bottom, the panels display the trajectories of the SF36 PF for: a) the whole cohort (n = 20), b) responders to the treatment defined by a ≥ ten-point increase in the SF36 PF at month 1 post IA (n = 14), and c) non-responders to the treatment defined by a ≤ ten-point increase in the SF36 PF at month 1 post IA (n = 7).

Fig. 2

Course of immunoglobulin and autoantibody levels over the study period (n = 20), the duration of IA therapy is indicated by the green bar. Error bars represent 95% confidence intervals. Data were analyzed using a linear mixed-effects model fitted by restricted maximum likelihood (REML), with t-tests computed using Satterthwaite's method for degrees of freedom with significance levels indicated as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. From left to right, the panels display the trajectories of: a) immunoglobulin G (IgG), b) immunoglobulin A (IgA), c) immunoglobulin M (IgM), d) β1 adrenergic receptor autoantibodies (β1 AR-AB), e) β2 adrenergic receptor autoantibodies (β2 AR-AB), f) M3 acetylcholine receptor autoantibodies (M3 AchR-AB), and g) M4 acetylcholine receptor autoantibodies (M4 AchR-AB).

Fig. 3

Clinical symptom progression over the study period (n = 20). The duration of IA therapy is indicated by the green bar. Error bars represent 95% confidence intervals. Data were analyzed using a linear mixed-effects model fitted by restricted maximum likelihood (REML), with t-tests computed using Satterthwaite's method for degrees of freedom with significance levels indicated as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. From left to right, the panels display the trajectories of: a) 36-Item Short-Form Survey energy/fatigue domain (SF36 energy/fatigue), where a higher score indicates less fatigue, b) 36-Item Short-Form Survey pain domain (SF36 pain), where a higher score indicates less pain, c) post-exertional malaise (PEM) as assessed by the DePaul Symptom Questionnaire (DSQ-PEM), higher scores indicate more severe PEM, and d) Composite Autonomic Symptom Score (COMPASS31), where a higher score indicates more autonomic symptoms.

Fig. 4

Course of mean handgrip strength (Fmean) in % in female patients (n = 13) measured a) initially and then b) repeated after one hour over the study period. The duration of IA therapy is indicated by the green bar. Error bars represent 95% confidence intervals. Data were analyzed using a linear mixed-effects model fitted by restricted maximum likelihood (REML), with t-tests computed using Satterthwaite's method for degrees of freedom with significance levels indicated as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Fig. 5

Course of 36-Item Short-Form Survey physical functioning domain (SF-36 PF) over the study period in patients who received a second cycle of IA therapy (n = 7). The duration of IA therapy is indicated by the green bar. Error bars represent 95% confidence intervals. Data were analyzed using a linear mixed-effects model fitted by restricted maximum likelihood (REML) for each cycle, with t-tests computed using Satterthwaite's method for degrees of freedom with significance levels indicated as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.