Objective: To test the hypothesis that the epsilon 4 allele of APOE is asso
ciated with a region-specific pattern of brain atrophy in AD. Methods: Volu
mes of the hippocampi, entorhinal cortices, and anterior temporal and front
al lobes were measured in 28 mild to moderate AD patients and 30 controls u
sing MRI, Within the AD group, 14 patients were noncarriers (-/-), 9 were h
eterozygous (epsilon 4/-), and 5 were homozygous (epsilon 4/4) for the epsi
lon 4 allele, Dementia severity was similar across the three AD groups. Res
ults: Smaller volumes were found with increasing dose of the epsilon 4 alle
le in the hippocampus, entorhinal cortex, and anterior temporal lobes in AD
patients. When compared with controls, the volume loss in the right and le
ft temporal regions ranged from -15.3 to -22.7% in the -/- AD group, from -
26.2 to -36.0% in the epsilon 4/- group, and from -24.0 to -48.0% in the ep
silon 4/4 group (p < 0.0005). In contrast, larger volumes were found in the
frontal lobes with increasing epsilon 4 gene dose. When compared with cont
rols, volume differences of the right: frontal lobe were -11.8% in the -/-
AD group, -8.5 in the epsilon 4/- group, and -1.4% in the epsilon 4/4 group
(p = 0.03). Conclusions: We found smaller volumes in the temporal lobe reg
ions but larger volumes in the frontal lobes with increasing APOE-epsilon 4
gene dose in AD patients. These data suggest a region-specific biological
effect of the epsilon 4 allele in the brains of AD patients.