A Systematic Review of the Potential Implication of Infectious Agents in Myasthenia Gravis

Abstract

Background: Myasthenia gravis (MG) is an autoimmune disorder of unknown etiology in most patients, in which autoantibodies target components of neuromuscular junctions and impair nerve to muscle transmission. Objective: To provide a synthesis of the evidence examining infectious agents associated with the onset of MG. Hypothesis: We hypothesized that microbes play a pathogenic role in the initiation of MG. For clinical cases, the onset of clinical signs is used as a proxy for the true onset of autoimmunity. Methods: We searched PubMed and Web of Science. Papers captured through database searching (n = 827) were assessed, yielding a total of 42 publications meeting the inclusion and exclusion criteria. An additional 6 papers were retrieved from the reference lists of relevant articles. For each pathogen, an integrated metric of evidence (IME) value, from minus 8 to plus 8, was computed based on study design, quality of data, confidence of infectious disease diagnosis, likelihood of a causal link between the pathogen and MG, confidence of MG diagnosis, and the number of infected patients. Negative IME values corresponded to studies providing evidence against a role for microbes as triggers of MG. Results: One hundred and sixty-nine myasthenic patients infected with 21 different pathogens were documented. Epstein-Barr virus (median = 4.71), human papillomavirus (median = 4.35), and poliovirus (median = 4.29) demonstrated the highest IME values. The total median IME was 2.63 (mean = 2.53; range -3.79-5.25), suggesting a general lack of evidence for a causal link. Conclusions: There was a notable absence of mechanistic studies designed to answer this question directly. The question of the pathogenic contribution of microbes to MG remains open.

Keywords: autoimmunity; etiology; infection; myasthenia gravis; virus.

Figure 1

Curation of records. Papers captured by search algorithms in PubMed and Web of Science (n = 827) were manually curated to remove duplicates (n = 592). Papers were screened to assess whether they met the inclusion and exclusion criteria, yielding a total of 48 publications. An additional 6 papers were retrieved from the reference lists of relevant articles.

Figure 2

Diagnostic algorithm for myasthenia gravis (MG). Having identified muscle weakness in a patient, biomarkers for myasthenia gravis should next be assessed. Detectible levels of antibodies against the acetylcholine receptor (AChR) or the muscle-specific kinase (MuSK) should be present for a firm diagnosis of MG. Seronegative patients demonstrating muscle weakness confirmed through repetitive nerve stimulation and/or single-fiber electromyography would yield a supportive diagnosis.

Figure 3

Agreement of integrated metric of evidence (IME) values for ten randomized publications. Ten research papers from the pool of included records were randomly chosen and individually assessed by three reviewers, blind to each other's scores. The interclass correlation between IME values was 0.97, demonstrating excellent reliability of the scoring system. Horizontal dotted lines indicate the threshold IME values between negligible and low (2.79), low and intermediate (4.14), and intermediate and high (5.58) levels of evidence.

Figure 4

Integrated metric of evidence (IME) values for infectious agents. Horizontal dotted lines indicate the threshold IME values between negligible and low (2.79), low and intermediate (4.14), and intermediate and high (5.58) levels of evidence. The median IME for the dataset is 2.63 (mean = 2.48; range −3.79–5.25).