вход по аккаунту


Ependymal changes in the human fetal brain.

код для вставкиСкачать
Ependymal Changes in
the Human Fetal Brain
Elizabeth C. Dooling, MD, Je G. Chi, MD, and Floyd H. Gilles, M D
Focal ependymal loss, increasing in extent with advancing gestational age, was found in 11 1 human fetal brains. Its
regular occurrence in specific sites, viz, the lateral walls of the lateral ventricles, over sector CA, of the hippocampus,
and on the ventral surface of the corpus callosum and lateral walls of the septum pellucidum, without prominent
subependymal gliosis, suggests that restricted ependymal loss may be related to growth and development of the fetal
brain. More extensive loss should be considered a marker of a neuropathological process.
Dooling EC, Chi JG, Gilles FH: Ependymal changes in the human fetal brain.
Ann Neurol 1:535-541, 1977
I n surveying a large series of human fetal brains of
different gestational ages, we have consistently observed discontinuity of the ependyma lining the
walls of the lateral ventricles, leadingus to believe this
process represents a maturational event. Friede [ 11
has stated that focal defects in the ependyma of the
lateral ventricles are common in unselected newborns, citing an incidence of 38 out of 5 1 consecutive
autopsies of babies and hypothesizing that these
ependymal defects may result from deformation of
the ventricles and compression of the skull at birth.
The purpose of this report is to describe the age of
onset and topographical distribution of these morphological changes in the ventricular lining of brains of
fetuses at various gestational ages and to discuss possible mechanisms of origin. An abstract has been published [la].
Materials and Findings
One hundred eleven brains of fetuses ranging in
gestational age from 13 to 42 weeks, which had been
serially sectioned according to a protocol [2], were
selected because of absence of gross defects or lesions
from a group of 1,155 normal and abnormal fetal
brains collected under the auspices of the Collaborative Perinatal Project. At least one coronally sectioned brain at each of these weeks of gestational age
(except 15 and 19 weeks) was examined, and for twothirds of the sampled ages, o n e or more brains cut in
either the horizontal or the sagittal plane were also
studied. Approximately one-third of the brains were
from stillborns, and few neonates survived more than
24 hours.
In brains of less than 20 weeks’ gestational age the
From the Department of Pathology (Neuropathology), Children’s
Med‘cai Center’ the Department Of
Neuropathology, Harvard Medical School, Boston, MA.
primitive ventricles are lined by closely packed, multilayered ependymal cells that are distinct from the
underlying densely cellular matrix. Between 2 1 and
25 weeks of gestation the ependyma lining the lateral
wall of the occipital lateral ventricle gradually becomes focally attenuated, the cells becoming flatter
and losing the cuboidal o r columnar appearance of the
intact adjacent ependymal cells (Fig 1). At 26 to 27
weeks of gestation, early degenerative changes
characterized by pyknosis, hyperchromatism or vesiculation of the nuclei, and occasionally ballooning of
the cytoplasm may be seen in the ependymal cells
lining the occipital horns. The dorsolateral aspect of
the occipital horn caudad to the level of the atrium
may show ependymal cell loss as early as 2 3 weeks and
consistently demonstrates focal loss by 27 to 28
weeks’ gestational age. There may be multifocal
losses, usually just anterior to the occipital tip. The
ependymal defects are associated with reduced numbers or absence of underlying residual matrix cells,
and ependymal tubules or rosettes in the subependymal zone are not uncommon (Fig 2). Glial proliferation o r hypercellularity in the areas of ependymal loss
is not pronounced. Further changes in the ependyma
of the occipital horns are evident after 32 weeks’
gestation and consist of irregularity of the lining covering the medial wall and extensive loss of ependyma
lining the lateral wall (Fig 3).
Progressive thinning, blunting, and attenuation of
the ependyma overlying sector CA2 of the hippocampus is initially focal after 28 weeks of gestation (Fig 4).
It is more extensive in most brains after 32 weeks’
gestation, when focal glial proliferation may occur (Fig
5 ) . Generally, gliosis is not prominent in the subepen-
Accepted for publication Jan 3, 1977.
Address reprint requests to Dr Gilles, Children’s Hospital Medical
Center, 300 Longwood Ave, Boston, MA 02 115 .
Fig 1. Attenuated ependymal cells of the lateral ufallof the
occipital horn and absence of underlying matrix cells. (HGE;
before 15 % reduction.)
F i g 2. Muhifocal loss of ependymal lining of the lateral wall
of the occipital horn, with buried ependymal tzcbules. ( H 6 E :
~ 1 0 before
10% reduction.)
536 Annals of Neurology Vol 1 No 6 June 1977
Fig.3. Occipital horn, J hocuingdenudation of ependjma
of the lateral wall and focal loss along the medial umall
(Med). IHG-E: x45 before 23% reduction.)
F i g 4 . Focallm ofependyma overthealveus. I H 6 E ; XI00
before 10% reduction.)
dymal zone of the denuded CA, sector. Except for an
occasional small focus of attenuated epend yma, the
lining of the temporal horn remains unbroken; the
ependyma always is intact near the medial angle of the
temporal horn and over Sommers’ sector in these
otherwise normal brains.
In the rostra1 part of the ventricular system there is
patchy loss of ependyma lining the ventral surface of
t h e corpus callosum and the lateral walls of the septum
pellucidum (Fig 6). It may be seen as early as 25 weeks
of gestation but is especially marked after 32 weeks.
Although the dorsolateral and ventrolateral angles of
the anterior horns often lack a continuous ependymal
lining, these are areas subject to mechanical distortion. Abrupt disruption of tissue without any associated reactive change such as cellular influx of macrophages o r glia and with intactness of the underlying
matrix usually can be distinguished from blunting of
the angles of the ventricles and mild o r moderate
ventricular dilatation accompanied by ependymal loss
and buried ependymal tubules. In the latter cases, it is
common to find widespread loss of ependyma at all
levels of the lateral ventricles on both medial and
lateral walls.
The ependymal lining of the third ventricle is continuous in all cases except those with massive hydrocephalus. The ependyma of the aqueduct is well
preserved, although its most dorsal aspect is highly
specialized in the subcommissural region (Chi and
Dooling, unpublished observation, 1976). Infrequently seen are buried ependymal tubules, ependymal gaps, or subependymal gliosis. The ependymal
lining of the fourth ventricle, except that overlying the
area postrema, remains unbroken at all levels.
In summary, we have found that ependymal loss
Dooling, Chi, and Gilles: Ependymal Changes in the Human Fetal Brain 537
begins in the occipital horns toward the end of the
second trimester. Between 28 and 35 weeks, a period
when the sulcal and gyral pattern of the fetal brain
becomes increasingly complex [3], there is more
widespread loss in specific sites: along both walls of
the occipital horns, over sector CA2 of the hippocampus, along the lateral walls of the septum pellucidum,
and o n the ventral surface of t h e corpus callosum,
more so rostrally (Fig 7, Table). The denuded areas
are not characterized by marked subeyendymal gliosis
or other reactive changes.
Is ependymal loss in the lateral ventricles unique to
the fetal brain? Can these focal ependymal losses be
attributed to traumatic changes in the course of processing postmortem material? O r d o these lesions
indicate an intrauterine infectious or postnatal anoxic
Shuangshoti and Netsky [4]have stated that ependymal and choroid plexus ceIls are found in the cerebrospinal fluid of patients of varying ages and that the
desquamation of cells is not related to improper handling of tissue; rather, they believe that desquamation
of epithelial cells into the ventricular system normally
occurs throughout life. Other authors [ 5 ] who have
noted shedding of epithelial cells in cerebrospinal
fluid have commented on the high numbers of these
cells in young infants (and also in children with hy-
Annals of Neurology Vol 1 No 6 June 1977
F i g 5 . Ependymul loss ouer sector CA, of the rostral
hippocampus. Note focalgliosis (arrow). ( H 6 E ; x4.5 before
10 % redaction .)
drocephalus, in whom there is stretching, thinning,
and then loss of ependyma). Thus, the ependymal
gaps in our histological material may reflect discrete
desquamation during fetal life.
The consistency of our finding of ependymal defects is supported by Friede [ ll, who has noted their
occurrence near the corners of the lateral ventricles
and on the ventral surface of the corpus callosum,
although he did not specify the gestational ages of the
infants he studied. He stated that either slight glial
proliferation or rarefaction of the matrix may occur,
but the glial response does not become more marked
with increasing gestational age. Pedunculated transependymal excrescences of matrix in the absence of
ependymal gaps indicated a malformative or degenerative process.
Other investigators have interpreted any ependymal loss as a pathological event. In a study of 80
cases of kernicterus and 7 cases of posticteric encephalopathy, Haymaker et a1 [6] found, in “occasional” cases, denudation of ependymal cells “here
and there.” In such areas the underlying germinal cells
were sometimes much sparser. They thought these
changes may have occurred neonatally, possibly in
Fig 6. Ependjmd loss along the ventral surface of the
corpus rallosum (CC). Note intuctness of ependymal
linirzg overtbe residual matrix zone (Mj, zuhirb overlies the
raudate ni~cle~s
( c a d ) . IHGE; x45 before3 % reduction.)
relationship to hemorrhage or plasma transudation
into the cerebrospinal fluid.
Among 5 1 infants (of whom 74% were premature)
who had sustained severe anoxia and usually required
resuscitation, Banker and Larroche [71 found that the
ependyma of the lateral wall of the occipital horns
adjacent to a zone of periventricular necrosis frequently disappeared. Subependymal gliosis was present in the more chronic cases they studied, but the
ependyrnal lining of the medial ventricular wall was
intact. Their material did not include serial sections
but rather representative sections of coronally sectioned brains, so it is possible that the more extensive
ependymal losses we noted were not present in the
samples available to them.
In a study of 100 brains of adults 20 to 60 or more
years of age who died of diverse neurological diseases,
Johnson and Johnson [83 found that 65 showed a
stereotyped lesion characterized by multifocal loss of
ependymal cells together with either subependymal
gliosis or loosening of the subependyrnal glia. They
considered this a premortem lesion and postulated
that the changes dated from “preadult” life. Russell [91
has stated that ependymal cells appear inert and incapable of regenerative activity in all forms of progressive damage to the ventricular lining. Denudation of
extensive areas results in exposure of the subependymal zone, which is capable of active proliferation.
Adjacent ependymal granulations may fuse to produce a gliofibrillary feltwork, deep to which may be
ependymal rosettes, “like fossils in a rock.” Without
information about the location or extent of the ependymal defects in the 65 adult brains, one cannot be
sure whether the ependymal granulations were persistent fetal defects similar to what we have described,
were acquired postnatally, o r were related to the patients’ primary neurological disorders, as varied in
origin as they were.
Experimental and naturally occurring viral infections, particularly mumps, have produced severe
granular epcndymitis [ 81. No evidence of intrauterine
infection was present elsewhere in the brains we examined, as one would expect if there were a
generalized infectious process, so it is unlikely that the
ependymal breaks reflect an intrauterine viral infection.
We speculate that the ependymal defects in the fetal
brain, occurring more frequently in the late second
and early third trimesters of gestation, stereotyped in
Dooling, Chi, and Gillrs: Ependynial Changes in the Human Fetal Brain
Distribution of Focal Ependymal Changes in Normal Fetal Brains
Percent of Brains Showing Ependymal Loss
Gestational Age
<28 ( N = 31)
28-35 (N = 49)
>35 (N = 31)
Occipital Horns
Frontal Horns
Hippocampus: CA2
F i g 7 . Schematic drawing of areas of ependymal loss, based
on tracing of enlargedphotographs of coronal sections from
the brain of a 34-uieekfetus.
location, and relatively consistent in extent, result
from a modeling process necessitated by the increasing size and complexity of the growing brain; furthermore, that the ventricular walls, subject to pulsatile
hydrostatic pressure of varying force, have less underlying support in those regions adjacent to incompletely myelinated tracts rather than to densely cellular neuronal aggregates, thus allowing the lining
ependyma to be more easily lost in these sites and
explaining the predilection for ependymal loss of the
occipital horns. It is possible that transient perinatal
ischemia-hypoxia can aggravate the vulnerability of a
normally defective or interrupted ependymal zone. If
extensive ependymal loss then occurs, changing hydrostatic pressure can cause dilatation of mild or
moderate degree in the ventricular system of the
cerebral hemispheres and so have untoward effects on
the underlying gray matter (eg, hippocampus) and
white matter (eg, the optic radiations, corpus callosum, and superior frontooccipital bundle). Additional studies of the ependyma, similar to those of the
choroid plexus in material from therapeutic abortions
[ 101 and embryos and fetuses ranging from 6 weeks’
gestational age onward [ 11, 121 as well as in animals
made hydrocephalic by vitamin-deficient diets [ 13,
141, would supplement our knowledge of the biological evolution of the ependymal lining of the ventricles during gestation.
This study was supported by the Collaborative Perinatal Project
(Contract N01-NS-3-2312) of the National Institute of Neurological and Communicative Disorders and Stroke.
1. Friede RL: Developmental Neuropathology. New York,
Springer-Verlag, 1975, pp 12-13
1a.Dooling EC, Chi JG, Gilles FH: Ependymal modifications in
the human fetal brain (abstract). Am J Pathol 86:56A, 1977
2. Yakovlev PI: Whole brain serial histological sections, in
Tedeschi CG (ed): Neuropathology: Methods and Diagnosis.
Boston, Little, Brown and Company, 1970, pp 371-378
3. Chi JG, Dooling EC, Gilles FH: Gyral development of the
human brain. Ann Neurol 1:86-93, 1977
4. Shuangshoti S, Netsky M G Human choroid plexus:
540 Annals of Neurology Vol 1 No 6 June 1977
morphologic and histochemical alterations with age. Am J Anat
128:73-93, 1970
Wilkins RH, Odom G L Cytological changes in cerebrospinal
fluid associated with resections of intracranial neoplasms. J
Neurosurg 25:24-34, 1966
Haymaker W, Margoles C, Pentschew A, et al: Pathology of
kernicterus and posticteric encephalopathy, in Swinyard CA
(ed): Kernicterus and Its Importance in Cerebral Palsy.
Springfield, IL, Charles C Thomas, 1961, pp 21-228
Banker BQ, Larroche J-C: Periventricular leucomalacia of
infancy. Arch Neurol7:386-410, 1962
Johnson KP, Johnson RT: Granular ependymitis. Am J Pathol
67:511-521, 1972
Russell DS: Hydrocephalus, in McIntosh R, Hare CC ( 4 s ) :
Neurology and Psychiatry in Childhood. Baltimore, The
Williams & Wilkins Company, 1954, pp 160-175
10 Bohr V, Mdllgird K Tight junctions in human fetal choroid
plexus visualized by freeze-etching Brain Res 81 314-318,
11 Kappers JA Structural and functional changes in the telencephalic choroid plexus during human ontogenesis, in
Wolstenholme GEW, 0 Connor CM (eds) The Cerebrospinal
Fluid Boston, Little, Brown and Company, 1958, pp 3-31
12 Dunn J, Kemohan JW Histological changes within the choroid
plexus of the lateral ventricle their relation to age Proc Staff
Meet Mayo Clin 30 607-615, 1955
13 Newberne PM, O’Dell BL Histopathology of hydrocephalus
resulting from a deficiency of vitamin BI2 Proc SOCExp Biol
Med 97 62-65, 1958
14 Witzel EW, Hunt GM The ultrastructure of the choroid
plexus in hydrocephalic offspring from vitamin A deficient
rabbits J Neuropathol Exp Neurol 2 1 250-262, 1962
Dooling, Chi, and Gilles: Ependymal Changes in the Human Fetal Brain 541
Без категории
Размер файла
674 Кб
ependymal, change, brain, human, fetal
Пожаловаться на содержимое документа