Lecanemab and donanemab are the first drugs shown in phase 3 trials to slow clinical decline in early symptomatic Alzheimer disease by removing cerebral amyloid. Over roughly 18 months, lecanemab reduced progression on the Clinical Dementia Rating Sum of Boxes (CDR-SB) by 0.45 points, and donanemab by 0.67 points in its primary low/medium tau population. These are relative slowings of about a quarter to a third, yet neither difference reaches published thresholds for a minimal clinically important difference. The cost of that benefit is amyloid-related imaging abnormalities (ARIA). Oedema or effusion developed in 12.6% of lecanemab and 24.0% of donanemab recipients, rose to about one in three among APOE ε4 homozygotes, and was occasionally fatal. Safe prescribing therefore depends on biomarker confirmation of amyloid, APOE genotyping, repeated MRI, infusion capacity and clinicians able to recognise ARIA. This review appraises the efficacy and safety evidence and then tests it against conditions in low- and middle-income countries (LMICs), home to more than 60% of people with dementia and to most of the projected rise from 57.4 million cases in 2019 to 152.8 million in 2050. Low-income countries have fewer than one CT scanner and 1.9 radiologists per million people, and one year of donanemab in India costs about 17 times national per capita gross domestic product in drug acquisition alone. At current prices and capacity, population-wide use is neither affordable nor safe in most LMICs. A staged model is proposed: timely diagnosis and risk-factor control for everyone, locally validated blood-biomarker triage at referral hubs, and antibody treatment confined to accredited centres with mandatory registries and prices negotiated against demonstrated value.