[Pubmed] Very late-onset myasthenia gravis: a systematic review and meta-analysis of clinical severity and long-term out

[RSS-Bot]

Eur Geriatr Med. 2026 Oct 7. doi: 10.1007/s41999-026-01621-9. Online ahead of print.

ABSTRACT

PURPOSE: Very late-onset myasthenia gravis (vloMG; onset ≥ 65 years) may represent a clinically distinct subgroup, but its severity and prognosis remain uncertain. We systematically reviewed studies comparing vloMG with earlier-onset MG (non-vloMG) and examined whether findings differed from the conventional early-onset (EOMG) versus late-onset dichotomy (LOMG; ≥ 50 years).

METHODS: MEDLINE, Embase, CENTRAL, and Google Scholar were searched from inception to May 2026. Observational studies of adult MG comparing age-at-onset groups were eligible. The main quantitative outcome was severe early presentation, defined as Myasthenia Gravis Foundation of America (MGFA) class IV-V early at the disease course (at onset, diagnosis, or the earliest pretreatment assessment). Other baseline and longitudinal outcomes were summarized qualitatively because outcome definitions and follow-up duration were heterogeneous.

RESULTS: Ten studies were included. Four studies including 647 patients with vloMG and 1,620 with non-vloMG contributed to the primary meta-analysis. Severe early presentation was more frequent in vloMG than in non-vloMG (10.2% vs 3.8%; Odds Ratio 2.19, 95%CI 1.49-3.24; p < 0.001; I2 = 0%). In contrast, the broader LOMG group showed no significant difference versus EOMG (5.6% vs 4.2%; Odds Ratio 1.12, 95%CI 0.74-1.71; p = 0.60). Qualitative synthesis suggested more common early bulbar and respiratory involvement, as well as need for intubation in vloMG, whereas long-term outcomes were inconsistent though often comparable among different age-at-onset groups, particularly when immunotherapy was used.

CONCLUSION: vloMG is associated with more severe early presentation, but long-term disease control may not differ from non-vloMG. Future studies should standardize age strata, follow-up, treatment exposure, and confounder adjustment.

PMID:42842206 | DOI:10.1007/s41999-026-01621-9


Source: https://pubmed.ncbi.nlm.nih.gov/4284220 ... 2&v=2.20.1