Introduction/Aims: We have encountered non-myasthenic patients being given a diagnosis of myasthenia gravis (MG). This study aims to investigate the frequency of, and factors contributing to, overdiagnosis of MG. Methods: This is a retrospective analysis of patients referred to our tertiary neuromuscular center for evaluation due to a previously suspected/confirmed MG diagnosis during a 6-year span. Results: A total of 531 patients sought a second opinion regarding their MG diagnosis, and 77 (14.5%) were found to have non-myasthenic conditions. A total of 11 patients tested positive for acetylcholine receptor (AChR) antibodies. Repeated AChR antibodies became negative in five patients while in four patients, AChR binding antibody titers were persistently low. In seven patients, striational antibody was the only positive antibody identified. In 25 patients, a prior electrodiagnostic (EDX) study was deemed positive, including 14 patients with abnormal repetitive nerve stimulation (RNS) and 12 with abnormal single fiber electromyography (SFEMG). Technical issues were noted on prior RNS studies in 8 patients, and repeat RNS was negative in 10 patients. In eight patients with previously abnormal SFEMG, results showed minimal or equivocal abnormalities. In two patients, a repeat SFEMG was normal. Further analysis revealed atypical clinical presentation, deceptively positive ice pack test, clinically insignificant antibody result and misleading EDX finding as main contributors to MG overdiagnosis. Discussion: Overdiagnosis of MG is not uncommon, and occurs more frequently in seronegative patients. To make an accurate diagnosis of MG, there is a need to recognize atypical presentations, and avoid overreliance on minor or non-specific serological and electrodiagnostic findings.
CITATION STYLE
Andrapalliyal, N., Claytor, B., & Li, Y. (2023). Incidence and causes of overdiagnosis of myasthenia gravis. Muscle and Nerve, 67(6), 464–468. https://doi.org/10.1002/mus.27774
Mendeley helps you to discover research relevant for your work.