Real-world evidence for lecanemab is showing a clearer safety picture as multiple retrospective studies are synthesized in a meta-analysis of 37 reports. Across those datasets, ARIA occurred in 15.1% of patients, with higher rates in APOE ε4 homozygotes and in people who entered treatment with baseline microhemorrhages. The same synthesis concluded that available real-world data are not strong enough to confirm cognitive efficacy, given the lack of controlled comparisons and differences in how outcomes were captured. The findings are consistent with the idea that APOE and baseline cerebrovascular findings remain key stratification variables for Alzheimer’s therapies that carry permeability risks. For developers and clinicians, the analysis strengthens the case for tighter real-world monitoring frameworks—especially around ARIA surveillance protocols and baseline MRI risk tagging—before broader uptake.