Clinical features of early-onset Alzheimer disease in a large kindred with an E280A presenilin-1 mutation - Early-Onset Familial Alzheimer's Disease (eFAD)

OBJECTIVES: To characterize clinical features of a very large pedigree with early-onset Alzheimer disease (AD) in which all affected individuals carry the identical glutamic acid-to-alanine mutation at codon 280 in the presenilin-1 gene. DESIGN: Clinical histories were obtained by patient and family interviews and through medical or civil records. Using standard diagnostic criteria, a case series of 128 individuals was identified, of which 6 have definitive (autopsy-proven) early-onset AD, 93 have probable early-onset AD, and 29 have possible early-onset AD. SETTING: Community based in Antioquia, Colombia. PATIENTS: A population-based sample in which all members of 5 extended families (nearly 3000 individuals) were surveyed. Criteria for inclusion required obtaining sufficient information to categorize the individual as affected. MAIN OUTCOME MEASURES: Age at onset, neuropsychological profile, neurologic history, and examination. RESULTS: The patients had a mean age at onset of 46.8 years (range, 34-62 years). The average interval until death was 8 years. Headache was noted in affected individuals significantly more frequently than in those not affected. The most frequent presentation was memory loss followed by behavior and personality changes and progressive loss of language ability. In the final stages, gait disturbances, seizures, and myoclonus were frequent. CONCLUSIONS: Other than the early onset, this clinical phenotype is indistinguishable from sporadic AD except that affected individuals frequently complained of headache preceding and during the disease. Despite the uniform genetic basis for the disease, there was significant variability in the age at onset, suggesting an important role for environmental factors or genetic modifiers in determining the age at onset.

Lopera F, Ardilla A, MartÃnez A, Madrigal L, Arango-Viana JC, Lemere CA, Arango-Lasprilla JC, HincapÃe L, Arcos-Burgos M, Ossa JE, Behrens IM, Norton J, Lendon C, Goate AM, Ruiz-Linares A, Rosselli M, Kosik KS. Clinical features of early-onset Alzheimer disease in a large kindred with an E280A presenilin-1 mutation. JAMA. 1997 Mar 12;277(10):793-9.

THE GENETIC basis for many cases of
early-onset familial Alzheimer disease
(AD) has now been established. Remark-
ably, mutations in at least 3 different
genes lead to a disease similar both clini-
cally and neuropathologically. The mu-
tated genes are the amyloid precursor
protein (APP) gene on chromosome 21,1
presenilin-1 (PSl) on chromosome 14,
and presenilin-2 (PS2) on chromosome
l.2-6 Most of the described mutations oc-
at various loci within the PSl gene.2,6Â·7

One aim in genetically defining the many
mutated loci capable of causing AD is to
understand how different mutations give
rise to the disease phenotype. A detailed
description of the phenotype is a pre-
requisite to tracing the mechanism by
which specific mutations lead to the pa-
thology and the clinical picture of the
disease. In some cases, small differences
in the specific mutated codon of a single
gene can lead to significant phenotypic
differences. For example, mutations
within APP lie on either side of the
sequence that corresponds to the ex-
cised ÃŸ-amyloid (or AÃŸ) peptide, the ma-
jor protein found in senile plaques. These
mutations lead to a phenotype that is
typical of AD, except for the early on-
set. However, a mutation within the AÃŸ
sequence gives rise to a distinct pheno-
type -- designated hereditary cerebral
hemorrhage with amyloidosis-Dutch
type.8 This single-base mutation, which
causes a glutamic acid-to-glutamine sub-
stitution in codon 693, shifts the most
prominent site of amyloid deposition from
the neuropil to the cerebral vascula-
ture and shifts the clinical presentation
from dementia to cerebral hemorrhage.
Families with a mutation at the adja-
cent codon, 692, have a phenotype with
features of both presenile dementia and
cerebral hemorrhage attributable to ce-
rebral amyloid angiopathy.9

In the case of the presenilins, patients
with mutations at various loci develop a
slowly progressive dementia that be-
gins in the third to fifth decade. Several
pedigrees with PSl or PS2 mutations
have been reported, and some of the
reports have described distinctive clini-
cal features. Haltia and colleagues10 de-
scribed a family with an M146V muta-
tion in PSl that has an unusually early
onset, ranging from age 35 to 39 years.
Although early myoclonus was noted,
only 1 patient is described in detail, and
the myoclonus was present at a late stage
of the disease when that patient became
bedridden. Lampe and colleagues11 de-
scribed the "L family" in whom they
noted early myoclonus, seizures, and pro-
gressive aphasia. Of the 16 reported
cases, the meanÂ±SD age at onset was
41.6Â±4.7 years, and the meanÂ±SD age
at death was 47.2Â±3.8 years. In only 6
patients was a mean Â± SD age docu-
mented for the onset of myoclonus
(44.7Â±2.8 years); likewise the meanÂ±SD
age for the onset of seizures (44.8 Â±2.9
years) was documented in 6 patients.
From these ages, it seems that, relative
to the rapid course of the disease in this
pedigree, the onset of myoclonus and
seizures occurred well after the onset of
cognitive symptoms when the disease
was advanced. Kennedy and colleagues12
also noted myoclonus in the affected
members of the pedigree they reported,
but no details are given regarding the
onset of the myoclonus relative to the
disease course. Among 50 individuals
from the Volga German families who
harbor a PS2 mutation, myoclonus was
noted in 12%, and seizures were noted
in 24%.13 Findings did not occur early in
the disease course for any of those cases.
Myoclonus and seizures were also a late
but frequent occurrence in a large fam-
ily aggregate from Normandy in whom
the disease was linked to chromosome
14.14

The first suggestion that an early-
onset dementing disorder was present
in a large kindred from the Colombian
state of Antioquia was presented by
Cornejo et al15 and was followed by the
identification of several additional fami-
lies from the same geographical region.16
Members of the 5 Colombian kindreds
described in this article all have a point
mutation in codon 280 that results in a
glutamic acid-to-alanine substitution in
PSl .6 A haplotype analysis demonstrates
that the families are likely to have a
common founder.17 The large size of these
kindreds allows a detailed study of the
range of disease phenotypes associated
with this mutation.

METHODS

Ascertainment of the Pedigrees

Five pedigrees from Antioquia with
an index case of early-onset AD were
assembled from extensive community-
based searches for affected families. The
index cases for each family were selected
from the Neurology Department, Uni-
versity Hospital San Vincente de Paul,
Medellin, Antioquia. These cases were
evaluated by neurologic and neuropsy-
chological testing. The clinical diagnos-
tic criteria used were the National
Institute of Neurological and Commu-
nication Disorders and Stroke-Alzhei-
mer's Disease and Related Disorders
Association18 and the Diagnostic and
Statistical Manual of Mental Disorders,
Fourth Edition; 6 cases had autopsy con-
firmation of their diagnosis and met the
established criteria for AD.19,20 At least
1 case in each family was analyzed for
the presence of a mutation in the PSl
gene. The diagnosis of probable AD re-
quired that the patient be examined by
1 of the participating neurologists. When
it was not possible to evaluate an indi-
vidual directly, the case was considered
as probable AD if histories obtained from
2 family members both fulfilled the di-
agnostic criteria. The case was consid-
ered as possible AD if it was not pos-
sible to obtain more than 1 reliable
history, and it was considered definite
AD if there was postmortem confirma-
tion.

Beginning with the index cases, we
assembled a pedigree for each extended
family. To construct the family tree and
to determine the status of individual
cases, several strategies were used: door-
to-door search for all family members
and examination of baptismal records,
notary registries, and clinical records of
the Mental Hospital of Antioquia and
the University Hospital San Vincente
de Paul. Consent for evaluation was ob-
tained from all subjects according to a
protocol approved by the human sub-
jects committee of the University of
Antioquia.

Clinical and
Neuropsychological Analysis

In addition to routine neurologic ex-
amination, a random subset of 15 indi-
viduals mildly to moderately affected
by AD and 34 normal controls under-
went neuropsychological testing. Con-
trols were relatives taken from the same
families as the familial AD patients and
were therefore at risk, but they matched
the familial AD group in sociocultural
conditions, including educational level.
At the time of the examination, they
were active, functionally normal, and
with no significant complaints or memory
impairments.

The neuropsychological test battery in-
cluded Spanish-language versions of in-
struments used by the Consortium to Es-
tablish a Registry for Alzheimer's Disease
(CE RAD)21 and several additional tests
adapted to the cultural and linguistic id-
iosyncrasies of the target population. The
battery included the following sections:
verbal fluency, naming, Mini-Mental State
Examination (MMSE),22 memory of
words, constructional praxis, recall of
words, recognition of words, recall of line
drawings, and trail-making tests.23 Some
additional neuropsychological tests were
also administered: the Raven Test,24 the
Wechsler Memory Scale,25 the praxis abil-
ity test, the Boston Naming Test, Span-
ish version,26 the Boston Diagnostic Apha-
sia Examination, Spanish version,27 the
Rey-Osterrieth Complex Figure with the
scoring system proposed by Taylor,28 and
Serial Verbal Learning.29 The Wechsler
Memory Scale, Boston Diagnostic Apha-
sia Examination, Rey-Osterrieth Com-
plex Figure, and Serial Verbal Learning
tests have been previously normalized in
Colombia in different age and educational
groups.29-33

Analysis of Postmortem Tissue

Autopsy material was available from
6 individuals in 2 pedigrees. The brain
was fixed with formalin, and sections
were taken from the frontal, parietal,
temporal, and occipital cortices, and the
hippocampus, caudate, thalamus, mid-
brain, pons, medulla, and cerebellum.
Sections were stained with hematoxylin-
eosin, a modified Bielschowsky method,
and antibodies to the AÃŸ peptide, glial
fibrillary acidic protein, and tau. The
neuropathological diagnosis of AD was
based on quantitative criteria suggested
by CERAD34 and by Khachaturian.19
Some specific aspects of the neuropa-
thology were presented as part of an
immunocytochemical study.36

RESULTS

Pedigrees and Sample
Characteristics

Five extended families with early-on-
set AD were identified in Antioquia. The
families live in towns located in a rela-
tively small altipiano area within a ra-
dius centered at 6Â°53.04' N, 75Â°20.13' W
at a mean altitude of 1700 m above sea
level. These families trace their gene-
alogies over 5 to 7 generations (Figure
1). A common origin of all 5 families is
likely based on the presence of the same
PSl mutation (E280A) in all of the af-
fected patients, the presence of a rare
haplotype,17 the limited geographic dis-
tribution of the family members, and
the sharing of 4 last names among a high
proportion of the family members. The
most common of these last names can be
traced to 1783, when the name was reg-
istered for the first time in the town of
Yarumal, Colombia (founded in 1780).

The combined sample size among all
of the pedigrees, including healthy and
at-risk individuals, is approximately 3000
individuals. From this sample, a diag-
nosis of probable AD was made in 93
individuals, possibleADin29individu-
als, and definite AD (autopsy-confirmed)
in 6 individuals. The sex distribution
among the cases was 66 men (52%) and
62 women (48%). Nearly all of the pa-
tients live in a rural setting, except for
a few who live in the city of Medellin,
Colombia. The mean educational level
attained was the second grade. The ma-
jority of the male patients were coffee
farmers, and the majority of the female
patients were housewives.The meanÂ±SD
age at onset among 88 individuals for
whom this information was available and
considered reliable(41 women and 47
men) was 46.8Â±6.4 years (range, 34-62
years). There was no significant differ-
ence in the age at onset between men
and women. Among the patients for
whom there was an adequate history,
27% were smokers, 12% consumed al-
cohol excessively, and 7% had a history
of head trauma that resulted in uncon-
sciousness. At the time of the analysis,
69 of the 93 probable AD patients and
14 of the 29 possible AD patients had
died. The meanÂ±SD age of death, based
on 54 observations, was 54.8Â±7.3 years
(range, 38-65 years), with no significant
differences between men and women.
When possible AD patients were ex-
cluded, the mean age at onset was 47.7
years (range, 34-62 years),and the mean
age at death was 60 years (range, 48-65
years). The average interval between on-
set and death was 8 years (range, 3-18
years). All of the patients were taken
care of in their homes throughout the
duration of the disease, and none re-
ceived nursing home care.

Clinical Symptomsand
NeurologicExamination

The most common initial complaints
were progressive memory loss and
changes in personality and behavior. Ev-
ery history included the complaint of
memory disturbance. Language difficul-
ties such as naming were also frequent
in the early stage of the disease. Most of
the patients did not have neurologic
problems other than impairment of men-
tal status until late in the disease course.
Gait difficulty, seizures, and myoclonus
were among the latest changes observed.
A total of 73% of those affected com-
plained of severe headache, often as a
prodromal feature, up to several years
before the onset of dementia. Only 2
(17%) of 12 nonaffected subjects from
the largest families reported headache.
Table1 presents the frequency of neu-
rologic findings among the sample of
affected patients for whom complete neu-
rological information was available(118
patients).

Neuropsychological Evaluation

Patients with a diagnosis of possible
AD and severely demented patients were
excluded from the compiled mental sta-
tus data.Two subjects in the control group
(1 man, 1 woman), and 3 subjects in the
affected group(1 man, 2 women) were
completely illiterate, and the rest had a
limited education. The highest level of
school attendance was 13 years in the
control group and 11 years in the familial
AD group. In the CERAD neuropsycho-
logical test battery, statistically signifi-
cant differences between both groups
were observed in the following tests
(Table 2): verbal fluency (except the 46-
to 60-second interval), naming (except for
high-frequency words), MMSE, memory
of words (except reading and intrusions),
constructional praxis (except for the circle
and the rectangles), recall of words (but
not intrusions), recognition ofwords (cor-
rect "no" response), recall of drawings
(except circle and cube), and trail making
(part A, errors). Verbal fluency and nam-
ing were both impaired in the affected
patients, but verbal fluency defects were
more evident than naming deficits. The
mean MMSE score in the affected group
was 14.1 of 30, and the mean MMSE score
among nonaffected family members was
27.1 (range, 24-30). The MMSE had been
normalized in Antioquia, and the normal
score has a lower limit of 23. Among the
literate subjects, no differences in read-
ing words were found; however, memory
of words differed significantly. The fa-
milial AD group had more difficulty than
the control group with construction of
drawing complex figures (eg, a cube) but
not simple figures (eg, a circle). The cube,
the only 3-dimensional figure in the test,
presented the greatest difficulty for the
AD patients. In the recall of figures, the
most difficult were the rectangle and the
diamond, and the easiest were the circle
and the cube.

Statistically significant differences
were observed in the following addi-
tional neuropsychological tests not in-
cluded in CERAD: Wechsler Memory
Scale (orientation, mental control, logi-
cal memory, associative learning, total
score, and memory quotient), praxis
(right hand and left hand but not buc-
cofacial), Boston Naming Test, Rey-
Osterrieth Complex Figure (copy and
immediate recall), and Serial Verbal
Learning (first trial and delayed recall)
(Table 3).

Results in the Wechsler Memory Scale
were particularly notable. Logical
memory and associative learning (both
verbal memory tests) were signifi-
cantly impaired in the AD patients,
whereas visual reproduction (nonver-
bal memory) scores were remarkably
similar in both groups. The memory quo-
tient, using Colombian norms,29 was more
than 2 SDs lower in the familial AD
group. In the praxis ability test, diffi-
culties in performing movements under
verbal command were impaired for both
hands. Buccofacial movement scores, al-
though decreased in the affected pa-
tients, were better preserved than hand
movements. In the Boston Naming Test,
scores in the familial AD group were
roughly half of the scores observed in
the controls; however, the control group
performance was also low, as expected
because of the limited education level
of the population. Naming difficulties
were observed in the confrontation nam-
ing subtest, but not in responsive nam-
ing and body-part naming. High-
probability repetition (but not low-
probability repetition), word reading (but
not oral reading), and writing (mechan-
ics but not primer-level writing) were
relatively well preserved in the demen-
tia patients. Those with dementia had
extremely low scores in the Rey-
Osterrieth Complex Figure-copy con-
dition, a constructional ability, and the
score for the Rey-Osterrieth Complex
Figure recall condition, a test using non-
verbal memory, was only 1 point of 36.
The Serial Verbal Learning test was es-
pecially difficult for the patients with AD,
and statistically significant differences
were observed in delayed recall.

Neuropathological Features

In the 6 autopsies performed, the brain
was atrophie. Three of the cases were
genotyped and had the expected glutamic
acid-to-alanine substitution at position 280
in the PSl gene. Histologically, all the
cases met the criteria for AD in terms of
plaque and tangle density. A description
of the pathologic characteristics in 1 case
is representative. The brain weighed 700
g, and there was severe and extensive
cortical atrophy and symmetrical central
atrophy. Atrophy of the frontal and tem-
poral lobes was most severe, with the
gyri forming thin projections into the sub-
arachnoid space. The hippocampus was
markedly shrunken bilaterally, and the
striatum, particularly the head of the cau-
date, was atrophie. The susbtantia nigra
was depigmented.

Histologically the case satisfied the cri-
teria for AD. In the cerebral cortex the
neuritic plaque density often exceeded
30/mm2. By digital image quantitation of
the inferior frontal gyrus, there was a
total plaque density of 150/mm2 on cross
section; however, only a portion of these
plaques had adjacent neurites or dense
amyloid cores at the center. Of interest
was the presence of neuritic plaques and
dense core plaques in the cerebellum as
well as abundant diffuse plaques often
located within the depths of the tertiary
cerebellar folia. Within the hippocampus
andcerebral cortex, there were promi-
nent neurofibrillary tangles and "ghost
tangles." The neurofibrillary tangles
stained with tau antibodies and the dif-
fuse and compact plaques stained with
AÃŸ antibodies (Figure2, top and bottom).
The detailed characteristics of the AÃŸ
staining with antibodies against various
forms of theAÃŸ peptide have been re-
ported.35 Other stigmata of AD such as
granulovacuolar degeneration were ob-
served in the hippocampus, but Hirano
bodies were not observed. The entire ce-
rebral cortex showed a prominent vacu-
olation of the neuropil in the superficial
layers, accompanied by severe gliosis of-
ten seen in Creutzfeldt-Jakob disease
(Figure 2, middle); however, the tissue
tested negative with antibodies against
prion protein. There was significant pal-
lor within the central region of the cen-
trum semiovale.

Clinical Cases

Case1. -- A 47-year-old man present-
ed with the chief complaint of gradual
progressive memory loss beginning 1-1/2
years earlier. He had run a farm for
many years, but had gone bankrupt 4
years earlier. From that time onward
he worked irregularly. When he was
seen in the clinic, his family reported
that he often forgot their names. He had
reduced verbal fluency, did not initiate
conversation, and had difficulty follow-
ing the thread of a conversation. Most
recently he had lost interest in his work
and in his family responsibilities. He be-
gan to wander and sometimes did not
seem to recognize his own house. How-
ever, he remained competent to care for
himself in his daily living activities. His
MMSE score was 4. He could not recall
any digits or recall a sentence. In cat-
egories, he named 8 animals in 1 minute,
2 fruits in 1 minute, and 1 body part in
1 minute. He had gone to elementary
school for only 2 years. Laboratory tests
included nonreactive VDRL, negative
human immunodeficiency virus test, and
normal levels of triiodothyronine, thy-
roxine, thyrotropin, vitamin B12, serum
urea nitrogen, and creatinine. He car-
ried the glutamic acid-to-alanine sub-
stitution at position 280 of PSl. Over
the next 2 years, he developed gait dif-
ficulty, total loss of language ability, and
frequent seizures. He died at age 50
years after a 5-year course of dementia.
His autopsy confirmed AD,

Case 2. -- A 47-year-old right-handed
woman was first seen 8 years after the
onset of progressive memory loss. At
the onset of the disease, she often re-
peated the same questions or stories,
and she often forgot where she had left
objects. She had no formal education
and worked as a housewife. Over the
ensuing years, she developed difficulty
in preparing meals, neglected her fam-
ily obligations, and appeared apathetic.
She did not complete activities that she
began and became isolated from others.
Four years after the onset of memory
loss, she developed agitation, became
aggressive at times, and began wander-
ing. She neglected her personal hygiene,
and her language function deteriorated.
By 6 years after the onset of her illness
and 2 years before being examined, she
developed urinary and occasional fecal
incontinence. She developed myoclonic
jerks of the trunk. She had a history of
smoking 20 cigarettes per day, and she
rarely drank alcohol. On examination at
the time of presentation, she had echo-
lalia and hypophonia and often mumbled
the same words over and over again.
Her MMSE score was 0. She had gen-
eralized hyperreflexia, sucking, and pal-
momental reflexes. She carried the glu-
tamic acid-to-alanine substitution at
position 280 of PSl. She died at age 47
years shortly after being examined. Her
autopsy confirmed AD.

COMMENT

The first reported kindred with patho-
logically documented familial AD was
published in 1932.36 Since then, more
than 100 large families with inherited
AD have been reported.37 Among these
families, the multiplex family described
in this article is undoubtedly the largest
familial AD kindred in the literature.
By virtue of its size and the frequency
of consanguinity, the extended family
could provide an important investiga-
tive resource for many current ques-
tions from ethical to genetic issues.
Among the latter, a search for homozy-
gotes in this population represents an
important goal. That all of the families
live within the same geographic region,
have family names in common, have the
same mutation, and share a rare allele 17
suggests a founder effect. Anthropologi-
cal, genetic, and historical studies of the
Antioquian community show that the
population originated from the Basque
region of Spain, with a small genetic
input from the amerindian and black
populations.38-39 Genotypic studies as well
as a search for familial AD among fami-
lies on the Iberian peninsula may reveal
a founder for this PSl mutation.

One common complaint among the af-
fected patients was headache. This
symptom has not been reported in as-
sociation with sporadic AD, but it was
noted as a possible prodromal feature in
patients with V717I mutations in the
APP gene.40 Many of the patients in our
study reported chronic headache with
occipital predominance, and this com-
plaint often preceded the onset of AD
by several years.

The neuropsychological characteris-
tics were typical of AD. The CERAD
and Wechsler tests indicated a severe
progressive alteration of verbal and non-
verbal memory, as well as logical and
associative memory. There were impair-
ments of language comprehension and
verbal fluency and difficulty in naming,
particularly for low-frequency words.
There were frequent severe impair-
ments in tests of constructional apraxia,
as demonstrated by CERAD and Rey-
Osterrieth Complex Figure results. Ex-
ecutive functions dependent on frontal
lobes were severely affected, as indi-
cated by the low phonologic and seman-
tic categorical fluency. The testing para-
digms were normalized for the Hispanic
population and for individuals with lim-
ited formal education because neuropsy-
chological tests are highly sensitive to
educational variables,41 and illiteracy
correlates with an extremely low per-
formance in most neuropsychological
tests.42,43 Although there were a few pa-
tients who had attained up to 11 years
of schooling, the size of this group was
too small to determine the effect of edu-
cation on age at disease onset.

In at least 2 reports describing smaller
families with mutations in PSl,mi clini-
cal features appeared to distinguish fa-
milial AD families from sporadic AD. The
"L family"11 was reported to have early
progressive aphasia, early-appearing my-
oclonus, generalized seizures, and para-
tonia. A family of Finnish descent10 was
reported to have early and prominent my-
oclonus. The large sample size available
within the collected members of the An-
tioquian pedigree allow us to conclude
that the E280A mutation causes a form of
AD that is clinically identical to the range
of signs and symptoms seen in sporadic
AD, except for the early age at onset.
Findings such as myoclonus, gait diffi-
culty, and seizures occur late. Although
seizures4"7 and aberrant motor activity48
are well established as late manifesta-
tions of sporadic AD, all of the published
families, including those described here,
suggest that seizures and myoclonus arise
more frequently with PSl mutations, but
always after the onset of the cognitive
impairment. More likely, the frequent ob-
servation of myoclonus and seizures rep-
resents the rapid progression of the dis-
ease. Indeed, a younger age at onset
predicts a significantly faster rate of dis-
ease progression.49 The clinical phenotypic
features described for other PSl muta-
tions, including early myoclonus and sei-
zures, may reflect the effects of a muta-
tion at sites other than position 280. In
this regard, a clinical description of the
family with the Glu280Gly mutation7
would be of particular interest because
they have a mutation in the identical codon
as the Antioquian families. Alternatively,
as the sample size grows for families with
other PSl mutations, the clinical pheno-
type may appear increasingly similar to
sporadic AD.

The postmortem findings were diag-
nostic of AD -- senile plaques and neu-
rofibrillary tangles were abundant. The
cerebellum was also significantly af-
fected with diffuse plaques as well as
neuritic plaques and dense core plaques.
Although cerebellar senile plaques are
well recognized to occur in sporadic AD,
their presence is variable, and only rarely
are they associated with neuritic changes
or do they form cores.50 Neuritic changes
around plaques were positive with ubiq-
uitin antibodies and negative with tau
antibodies.35 Despite the significant cer-
ebellar involvment, gait difficulty was a
late occurrence.

Within this very large pedigree, a
number of commonly assessed clinical
features show phenotypic variability.
For example, the age at onset ranged
from 34 to 62 years. If the 2 patients at
the extreme of each range were dis-
carded as outliers, the onset would still
span 21 years (ages 39 to 60 years). Al-
though both of the outliers were cat-
egorized as probable AD because they
lack autopsy confirmation, they both
carry the E280A mutation and, there-
fore, provide strong evidence for a wide
age range at disease onset. The age range
of disease onsets reported with various
mutations in PSl is from 35 to 55 years,7
a range that falls within the ages ob-
served in our pedigrees. The wide age
range of disease onsets among patients
that carry an identical mutation strongly
suggests an effect of genetic or envi-
ronmental modifiers. One genetic de-
terminant believed to affect age at on-
set is the number of apolipoprotein E
(APOE) e4 alleles51; however, the APOE
e4 allele distribution in this population
does not correlate with age at onset.17
Besides APOE e4, other possible ge-
netic disease modifiers include the
CYP2D6B allele,52 and polymorphisms
in NACP/alpha-synuclein,53 possibly alpha1-an-
tichymotrypsin,5456 and the intron 3' to
exon 8 of PSl.57 Environmental expo-
sures, which may affect the incidence of
AD,58 are also a possible explanation for
the wide range of age at onset; however,
at this juncture, there are no obvious
clues concerning the identity of such a
factor. Clearly, these families provide
an ideal clinical setting in which to search
for genetic and environmental modifi-
ers of disease expression.

This study was supported by Colciencias grant
1115-04-040-95 (Drs Lopera and Ossa).

The authors gratefully acknowledge the collabo-
ration of the families and civil authorities of the
municipalities that participated in this study, the
University Hospital San Vincente de Paul, and the
Mental Hospital of Antioquia.

References
1. Goate A, Chartier-Harlin M-C, Mullan M, et al.
Segregation of a missense mutation in the amyloid
precursor protein gene with familial Alzheimer's
disease. Nature. 1991;349:704-706.
2. Sherrington R, Rogaev EI, Liang Y, et al. Clon-
ing of a novel gene bearing missense mutations in
early onset familial Alzheimer disease. Nature. 1995;
375:754-760.
3. Rogaev EI, Sherrington R, Rogaeva EA, et al.
Familial Alzheimer's disease in kindreds with mis-
sense mutations in a
gene on chromosome 1 related
to the Alzheimer's disease type 3 gene. Nature.
1995;376:775-778.
4. Li J, Ma J, Potter H. Identification and expres-
sion analysis of a potential familial Alzheimer dis-
ease gene on chromosome 1 related to AD3. Proc
Natl Acad Sci U S A. 1995;92:12180-12184.
5. Levy-Lahad E, Wasco W, Poorkaj P, et al. Can-
didate gene for the chromosome 1 familial Alzhei-
mer's disease locus. Science. 1995;269:973-977.
6. Clark RF, Hutton M, Fuldner RA, et al. The
structure of the presenilin 1 (S182) gene and iden-
tification of six novel mutations in early onset AD
families. Nat Genet. 1995;11:219-222.
7. Van Broeckhoven C. Presenilins and Alzheimer's
disease. Nat Genet. 1995;11:230-232.
8. Levy E, Carman MD, Fernandez-Madrid IJ, et
al. Mutation of the Alzheimer's disease amyloid
gene in hereditary cerebral hemorrhage, Dutch-
type. Science. 1990;248:1124-1126.
9. Hendriks L, van Duijn CM, Cras P, et al. Pre-
senile dementia and cerebral haemorrhage linked
to a mutation at codon 692 of the \g=b\-amyloidpre-
cursor protein
gene. Nat Genet. 1992;1:218-221.
10. Haltia M, Viitanen M, Sulkava R, et al. Chro-
mosome 14-encoded Alzheimer's disease: genetic
and clinicopathological description. Ann Neurol.
1994;36:362-367.
11. Lampe TH, Bird TD, Nochlin D, et al. Pheno-
type of chromosome 14-linked familial Alzheimer's
disease in a large kindred. Ann Neurol. 1994;36:
368-378.
12. Kennedy AM, Newman SK, Frackowiak RS, et
al. Chromosome 14 linked familial Alzheimer's dis-
ease: a clinico-pathological study of a single pedi-
gree. Brain. 1995;118:185-205.
13. Bird TD, Lampe TH, Nemens EJ, et al. Char-$
acteristics of familial Alzheimer's disease in nine
kindreds of Volga Germany ancestry. Prog Clin
Biol Res. 1989;317:229-234.
14. Campion D, Brice A, Hannequin D, et al. A
large pedigree with early-onset Alzheimer's dis-
ease: clinical,
neuropathologic, and genetic charac-
terization. Neurology. 1995;45:80-85.
15. Cornejo W, Lopera F, Uribe CS, Salinas M.
Descripci\l=o'\nde una familia con demencia presenil
tipo Alzheimer. Acta Med Colombiana. 1987;12:55\x=req-\
61.
16. Lopera F, Arcos M, Madrigal L, et al. Demencia
tipo Alzheimer con agregacion familiar en An-
tioquia, Colombia. Acta Neurol Colombiana. 1994;
10:173-187.
17. Lendon CL, Martinez A, Behrens IM, et al. The
E280A mutation causes Alzheimer's, but age of
disease onset is not determined by ApoE alleles.
Hum Mutat. In press.
18. McKhann G, Drachman D, Folstein M, Katz-
man R, Price DL, Stadlan EM. Clinical diagnosis of
Alzheimer's disease: a report of the NINCDS-$
ADRDA Work Group, Department of Health and
Human Services Task Force on Alzheimer's Dis-
ease. Neurology. 1984;34:939-944.
19. Khachaturian ZS. Diagnosis of Alzheimer's dis-
ease. Arch Neurol. 1985;42:1097-1105.
20. Rogers JD, Brogan D, Mirra SS. The nucleus
basalis of Meynert in neurological disease: a quan-
titative morphological study. Ann Neurol. 1985;17:
163-170.
21. Morris JC, Heyman A, Mohs RC, et al. The
Consortium to Establish a Registry for Alzheimer's
Disease (CERAD), part I: clinical and neuropsy-$
chological assessment of Alzheimer's disease. Neu-
rology. 1989;39:1159-1165.
22. Folstein MF, Folstein SE, McHugh PR. 'Mini\x=req-\
Mental State': a practical method for grading the
cognitive state of patients for the clinician. J Psy-
chiatr Res. 1975;12:189-198.
23. Reitan RM, Wolfson D. The Halsted-Reitan
Neuropsychological Test Battery: Theory and Clini-
cal Interpretation. Tucson, Ariz: Neuropsychology
Press; 1993.
24. Raven JC. Revised Manual for Raven's Pro-
gressive Matrices and Vocabulary Scale. Oxford,
England: Oxford Psychologists Press Ltd; 1976.
25. Wechsler D. Manual for the Wechsler Memory
Scale-Revised. San Antonio, Tex: The Psychologi-
cal Corp; 1987.
26. Kaplan E, Goodglass H, Weintraub S. The Bos-
ton Naming Test. Philadelphia, Pa: Lea & Febiger;
1983.
27. Goodglass H, Kaplan H. Evaluaci\l=o'\nde las Afa-
sias y de Transtornos Similares. Buenos Aires,
Argentina: Editorial Medica Panamericana; 1983.
28. Taylor EM. The Appraisal of Children With
Cerebral Deficits. Cambridge, Mass: Harvard Uni-
versity Press; 1959.
29. Ardila A, Rosselli M, Puente A. Neuropsycho-
logical Evaluation of the Spanish Speaker. New
York, NY: Plenum Press; 1994.
30. Ardila A, Rosselli M. Neuropsychological char-$
acteristics of normal aging. Dev Neuropsychol. 1989;
5:307-320.
31. Ardila A, Rosselli M. Development of language,
memory and visuospatial abilities in 5- to 12-year\x=req-\
old children using a neuropsychological battery.
Dev Neuropsychol. 1994;10:97-120.
32. Rossselli M, Ardila A, Florez A, Castro C. Nor-
mative data on the Boston Diagnostic Aphasia Ex-
amination in a Spanish-speaking population. J Clin
Exp Neuropsychol. 1990;12:313-322.
33. Rosselli M, Ardila A. Effects of age, education
and gender on the Rey-Osterrieth Complex Fig-
ure. Clin Neuropsychologist. 1991;5:370-376.
34. Mirra SS, Heyman A, McKeel D, et al. The
Consortium to Establish a Registry for Alzheimer's
Disease (CERAD), part II: standardization of the
neuropathologic assessment of Alzheimer's disease.
Neurology. 1991;41:479-486.
35. Lemere CA, Lopera F, Kosik KS, et al. The
E280A presenilin 1 Alzheimer mutation produces
increased A\g=b\42deposition and severe cerebellar
pathology. Nat Med. 1996;2:1146-1148.
36. Schottyky J. Uber prasenile verblodungen.
Z Gesamte Neurol Psychiatr. 1932;140:333-387.
37. Bird TD. Familial Alzheimer's disease. Ann
Neurol. 1994;36:335-336.
38. Gomez JR, Bravo ML. Estudio de la Estruc-
tura Genetica de la Poblacion del Santuario An-
tioquia. Medellin, Colombia: Universidad de An-
tioquia; 1985:93H.
39. Arcos OM. Epidemiologia Genetica de Hendi-
dura Facial no Sindromatica en la Poblacion de
Antioquia. Medellin, Colombia: Universidad de An-
tioquia; 1992.
40. Mullan M, Tsuji S, Miki T, et al. Clinical com-
parison of Alzheimer's disease in pedigrees with
the codon 717 Val->Ile mutation in the amyloid
precursor protein gene. Neurobiol Aging. 1993;14:
407-419.
41. Ardila A. Directions of research in cross-cul-
tural neuropsychology. J Clin Exp Neuropsychol.
1995;17:143-150.
42. Ardila A, Rosselli M, Rosas P. Neuropsycho-
logical assessment in illiterates: visuospatial and
memory abilities. Brain Cogn. 1989;11:147-166.
43. Rosselli M, Ardila A, Rosas P. Neuropsycho-
logical assessment in illiterates, II: language and
praxic abilities. Brain Cogn. 1990;12:281-296.
44. Erkinjuntti T. Differential diagnosis between
Alzheimer's disease and vascular dementia: evalu-
ation of common clinical methods. Acta Neurol
Scand. 1987;76:433-442.
45. Mendez MF, Catanzaro P, Doss RC, Arguello
R, Frey WH. Seizures in Alzheimer's disease: clini-
copathologic study. J Geriatr Psychiatry Neurol.
1994;7:230-233.
46. McAreavey MJ, Ballinger BR, Fenton GW. Epi-
leptic seizures in elderly patients with dementia.
Epilepsia. 1992;33:657-660.
47. Hauser WA, Morris ML, Heston LL, Anderson
VE. Seizures and myoclonus in patients with Alz-
heimer's disease. Neurology. 1986;36:1226-1230.
48. Mega MS, Cummings JL, Fiorello T, Gornbein
J. The spectrum of behavioral changes in Alzhei-
mer's disease. Neurology. 1996;46:130-135.
49. Koss E, Edland S, Fillenbaum G, et al. Clinical
and neuropsychological differences between patients
with earlier and later onset of Alzheimer's disease:
a CERAD analysis, part XII. Neurology. 1996;46:
136-141.
50. Mann DM, Iwatsubo T, Cairns NJ, et al. Amy-
loid beta protein (A\g=b\) deposition in chromosome
14\p=m-\linked Alzheimer's disease: predominance of
A\g=b\42(43). Ann Neurol. 1996;40:149-156.
51. Corder EH, Saunders AM, Strittmatter WJ, et
al. Gene dose of apolipoprotein E type 4 allele and
the risk of Alzheimer's disease in late onset fami-
lies. Science. 1993;261:921-923.
52. Chen X, Xia Y, Alford M, et al. The CYP2D6B
allele is associated with a milder synaptic pathology
in Alzheimer's disease. Ann Neurol. 1995;38:653-658.
53. Xia Y, Rohan de Silva HA, Rosi BL, et al.
Genetic studies in Alzheimer's disease with an
NACP/\g=a\-synuclein polymorphism. Ann Neurol.
1996;40:50-58.
54. Kamboh MI, Sanghera DK, Ferrell RE,
DeKosky ST. APOE*4-associated Alzheimer's dis-
ease risk is modified by \g=a\l-antichymotrypsin poly-
morphism. Nat Genet. 1995;10:486-488.
55. Nacmias B, Tedde A, Latorraca S, et al. Apo-
lipoprotein E and \g=a\l-antichymotrypsin polymor-
phism in Alzheimer's disease. Ann Neurol. 1996;
140:678-680.
56. Muller U, Bodeker R-H, Gerundt I, Kurz A.
Lack of association between \g=a\l-antichymotrypsin
polymorphism, Alzheimer's disease, and allele \r=ie\4of
apolipoprotein E. Neurology. 1996;47:1575-1577.
57. Wragg M, Hutton M, Talbot C, Alzheimer's
Disease Collaborative Group. Genetic association
between intronic polymorphism in presenilin-1 gene
and late-onset Alzheimer's disease. Lancet. 1996;
347:509-512.
58. White L, Petrovitch H, Ross GW, et al. Preva-
lence of dementia in older Japanese-American men
in Hawaii: the Honolulu-Asia Aging Study. JAMA.
1996;276:955-960.