Low-dose versus standard-dose intravenous immunoglobulin in generalized myasthenia gravis: a prospective single-center cohort study

Background Generalized myasthenia gravis (gMG) is commonly treated with high-dose intravenous immunoglobulin (IVIG), typically at cumulative doses near 2 g/kg, during clinical worsening. However, the minimal effective IVIG exposure remains uncertain. Methods In this prospective single-center cohort study, we enrolled 34 adults with AChR-positive gMG who received either low-dose IVIG (target ≤0.2 g/kg/day; n = 21) or standard-dose IVIG (target 0.4 g/kg/day; n = 13) according to routine clinical practice. Clinical outcomes were assessed using the Myasthenia Gravis Activities of Daily Living (MG-ADL) scale and the Quantitative Myasthenia Gravis (QMG) score at baseline and at weeks 1, 2, 4, 8, and 12. For the index IVIG course, infusion days, cumulative dose, and primary treatment indication were recorded. Oral prednisone use was evaluated as a short-term steroid-sparing outcome. Between-group comparisons were explored using regression models adjusted for age and thymoma status. Results Both regimens were associated with clinically meaningful improvement in MG-ADL and QMG through week 12, and no statistically significant between-group differences were detected at any follow-up visit. Most patients received 5 infusion days. After weight-based conversion, the low-dose group clustered around a cumulative exposure of approximately 1.0 g/kg, whereas the standard-dose group clustered around approximately 2.0 g/kg. Acute exacerbation was the most common indication in both groups, whereas maintenance therapy was more frequent in the standard-dose cohort. Prednisone requirements did not decline over 12 weeks in either group. Conclusion In this prospective observational cohort, low-dose IVIG was associated with short-term clinical improvement comparable to that observed with standard-dose IVIG, without a clear steroid-sparing signal. These findings should be interpreted cautiously given the small sample size, baseline imbalance, and non-randomized design.