The clinical-pathological correlations that were prospectively obtaine
d in a cohort of old patients (> 75 years of age) are reviewed The pat
hological data were obtained in 31 cases, either normal or affected by
Alzheimer disease of various degrees of severity. The density of the
AP peptide deposits was pool ly linked with the intellectual status. O
ne patient had a very high density of deposits, although she was consi
dered intellectually normal. When present in a patient, the A beta dep
osits usually involved all the cortical samples; the samples devoid of
deposits most often belonged to the limbic system. The distribution o
f the neurofibrillary tangles was highly selective : the primary areas
(such as the visual cortex) were lesioned only in a few cases, invari
ably the most severely affected ones. Neurofibrillary tangles involved
the associative cortices (sparing the primary areas) in the cases of
intermediate severity. The hippocampal-parahippocampal areas contained
at least a few neurofibrillary tangles in all the cases. The prevalen
ce of the neurofibrillary lesions in that cohort of cases probably ind
icated the chronological (and hierarchical) order of involvement : fro
m limbic to associative, from associative to primary areas. There was
a linear relationship between the density of the neurofibrillary tangl
es and the intellectual deficit in the hippocampal-parahippocampal gyr
us. The relationship was stepwise rather than linear in the isocortica
l samples, suggesting that the neurofibrillary tangles were a late phe
nomenon in those types of cortices. An accumulation of SNAP 25 immunor
eactivity was found in some of the most severely affected cases, point
ing to a deficit in axonal transport. The density and the total number
of neurons were evaluated in a sample of the supramarginal gyrus. The
neuronal loss was found to be severe, but only in the most affected c
ases, when the density of neurofibrillary tangles was higher than 5/mm
(2).